The Taurus

The Taurus wasn’t a major project, but a handy companion piece to the Gemini, an earlier Stylophone modification.

The problem with the Gemini is that it has two voices, output in stereo, but, typical of the Stylophone, it has only the one speaker.  This means that some of its effects are only available via the stereo line out.

As a result of much past experimentation, I have many Stylophone bits left over.  To make the Taurus, which was to be a very simple external amplifier, I used an empty case, some spare grille material, two amplifier circuit boards and two speakers, all from S1 reissue versions of the instrument.

The Stylophone grille isn’t glued down, and can be removed from the inside by pushing out half a dozen lugs which hold it in place.  I did this first, cut a hole in the top of the case for the second speaker – vaguely matching the hole through which the original sounds – and refixed the grille.

The picture shows the second speaker glued in place, and the two amplifier boards connected to a new stereo input socket, the battery box and the speakers:

I wasn’t using any of the original keyboard, switch and socket parts, so I glued some spare grille sections inside the switch and socket holes and outside over the hole through which the keyboard is normally accessed.

A small tripod was attached to the base to enable the speakers to be angled for better distribution of the sound.  Decoration consisted of astrological symbols, in the style of the Gemini, and matching black and white bulls, front and back.

It works well with the Gemini, which has its own volume and mix controls, but is a very basic unit indeed – no volume control, no on/off switch and no external power socket: it uses three AA batteries like the Gemini itself, and could be useful with other instruments needing a slight volume boost and not connected via a line out socket.

 

 

UCreate Music, Part 3 – the software

I may as well admit at the start that I never quite cracked the problem of the UCreate software – nearly, but not quite, as I will explain!

The issue of the UCreate software is a contentious one, as the software to allow users to swap around the sounds and effects in their UCreate could only be used by connecting to the UCreate website; but after a couple of years Mattel took down the website and it was no longer possible to do this.

There is a long and very interesting discussion on the subject – from which many of the ideas below originate – at http://nomeist.com/ucreate-music/199.

The way it originally worked was this: the first time you plugged in your UCreate it automatically connected to the website, and downloaded a program, which – on the Mac at least, looked like this:

This confirmed that your device was connected, and gave you the Firmware version (I think version 9 was the last one) and the individual number of your device.  Each time you plugged in your UCreate and opened the program, this is what you would see.  As the instruction at the top suggests, if you rested your cursor on various parts of the picture, you were given an explanation of the function of the various controls, for example:

However, more importantly, if you clicked the tab on the right, ‘Music portal’, you were able to log onto the website:

As you can’t do this now, I’ve had to rely on an old screenshot from a Windows PC.  I don’t have any screenshots of what you would find when you got there, and even the Wayback Machine only shows a background image, presumably because the site operated via a Flash app; but you were able to download some alternative sounds directly to the UCreate device, and move the effects around.  What users anticipated was, as time went on, more sounds and new effects would appear, perhaps even a community of users sharing their own creations.

Instead, the site was closed, and you now see only this:

Not only was access to the existing sounds and effects withdrawn, but, worse still, also the means of loading them into the UCreate devices; and everyone who used one was stuck with whatever sounds they currently had installed.

For most, this was the original sounds the UCreate came with, or one of the extra packs that had been available on the website.  I have the extra packs, but can’t use them because there’s no separate software available to install them on the machine.

Some astute users had observed that the downloading process occurred in two parts: first the sounds were downloaded from the website to the host computer; then they were transferred from the computer to the UCreate.  The time-lag between these two operations allowed their own sound files, suitably renamed to match those being downloaded, to be quickly substituted between downloading and transferring.  The timing of this was crucial, and could be a bit hit-and-miss, so one proficient user, ‘scragz’, even wrote a routine for Macs that automated the task (https://github.com/scragz/ucreate-loader); but if these were still in the machine when the site was taken down, these were the sounds you were left with.

*

Fortunately, that isn’t the end of the story.  Another user by the name of  ’marbs’ created a program a couple of years ago (http://marbs-online.blogspot.co.uk/) which does the job without the use of the website.  It will be of inestimable value for any UCreate users out there.

First of all, this is by no means simple, but there are 3 steps to getting your UCreate to accept new sounds.  These are:

Step 1. Get the computer to recognise your UCreate device;

Step 2. Create the right kind of sound files for the device to use; and

Step 3. Load the sounds into the device.

I’ll deal with these steps one by one.  First of all, though, download marbs’ programs from here: https://gitorious.org/u-load.  I say ‘programs’, because there are two: one to create the correct sound files (‘LopGenerator’ – they have the extension ‘.lop’) and one to group them and load them into the UCreate device (‘u-load’).

I have to admit that I couldn’t get the pre-compiled Mac version of u-load to work, nor was I able to recompile it with XCode, so I did the following on a Windows computer, which I’m not at all familiar with, but which worked perfectly:

Step 1. Getting the computer to recognise your UCreate device.

If you just plug your UCreate into the computer and run u-load, you’ll get this:

which is no help at all.  marbs’ README file explains that what you need to do is install a ‘libusb driver’ to communicate with the device.

marbs’ recommended way to do this is as follows:

a.  Download the driver installer program ‘Zadig’ from here: http://sourceforge.net/projects/libwdi/files/zadig/, (choose a version according to whether you’re using XP, Vista or Windows 7), and unzip it with a program such as ‘7-Zip‘.

b.  Start Zadig, go to the ‘Options’ menu, and select ‘List All devices’.  marbs says ‘choose GENERALPLUS-MSDC from the drop-down list’.  I didn’t have this on the list, so I used the ‘Edit’ function to rename the one item that was there – I don’t know if this was necessary or not, but as there was only one file in the list, and only one USB device plugged in – the UCreate – I knew it must be the right  one.

c. Then click the ‘Install Driver’ button.

This picture shows the Zadig window, with the 4 actions to take:

1) Choose ‘List All Devices’ from the ‘Options’ menu.

2) Only one item appeared on the list, which is the long box partly hidden under the drop down menu.  Probably a good idea anyway to make sure the UCreate is the only USB device connected, to avoid confusion. I checked the ‘Edit Name’ box so I could rename it ’GENERALPLUS-MSDC’.

3) Actually, I can’t remember what was selected here.  I didn’t change what it already said, which is probably what is showing here.

4) Clicking this button starts the installation.

d.  The installation wasn’t instant: it took a couple of minutes, and some messages came up about restarting devices.  I switched the UCreate off, restarted the computer, switched the UCreate back on, and started u-load.

This time I got the following message:

which was a whole lot better, indicating that u-load now recognised the UCreate, and was ready to load sounds and effects into it.

You only have to do this once, then your device is set up.

One final thing: I think it only fair to point out what marbs says at the end of the README file: ‘To use the original application delivered with your ucreate, you have to deinstall the libusb driver in the device manager.’

In other words, you won’t be able to use the UCreate’s original software now, unless you can do this deinstall. I’m not sure whether Zadig does this or not; but I believe it’s designed not to remove existing drivers, so it would be possible, I presume, to restore the UCreate to its original state with this or with the Windows Device Manager.  It’s a moot point whether it would worth doing it, of course, since the website on which the original software depends no longer exists.

(Doing this on the Windows computer, incidentally, didn’t stop the original software from running on the Mac when I later connected the UCreate to that).

This is what you see when you press 1, then ENTER for ‘Help’:

This tells you exactly what u-load can do: ‘Make’ = group your sound files together into a big file which can be used by the UCreate; ‘Write’ = load these file groups onto the UCreate.

*

Step 2.  You need to create some sound files of the type that u-load and the UCreate can use (with the extension ‘.lop’).  There are some restrictions on this – principally making sure the files are not more than 5 seconds long.  There were reports in the past of some UCreate devices being irreparably damaged by not doing this, so fortunately marbs’ ‘LopGenerator’ program – version 0.5, at least – which creates the files for you, will refuse to do so if they’re too big.

I was able to use both Mac and Windows versions of the LopGenerator which were downloaded together with u-load, although I recompiled the Mac version from the .c file included, as this was version 0.5, which I prefer because it has the oversize file warning.  If you’d like to use this in preference to the included version 0.4, you can download it from here:

LopGenerator doesn’t require the UCreate to be plugged in, it just works on sound files stored in a particular location on your computer.  For some reason – possibly of my own doing, I don’t know – my 0.5 version will only work if it and the sound files it’s working on are in my ‘Users’ folder, in the folder for the main user.  You may or may not have the same problem.  The Windows version merely requires the sound files to be in the same directory as LopGenerator.

All you have to is create ‘.wav’ sound files with the correct attributes, and then have LopGenerator turn them into files the UCreate can recognise.  marbs has written an excellent tutorial on how to do this, which you can find at http://marbs-online.blogspot.co.uk/2011/03/tutorial-how-to-make-custom-samples.html.

Essentially, that’s the way I’ve done it.  I’ve edited the files in Audacity, to ensure:

a. they’re mono and 5 seconds long:

b. The project rate is 8000:

c. They’re saved as ‘Other uncompressed files’ and, by clicking ‘Options’, the header is set to ‘WAV (Microsoft)’ and they’re encoded as ‘Unsigned 8 bit PCM’

Once the file or files are done, I copy them to the folder with LopGenerator in it, run LopGenerator and I get this message:

If you try one that’s too big for UCreate, you get this:

I just did one file here as an example, but you could put any number of .wav files in the folder and they would all be changed to .lop files at the same time.

*

Step 3: The final step is to load the sound files into the UCreate.

To do this, you’ll be typing 2, then ENTER (‘Make SamplePackage’, as the Help screen tells you), which will create a file which you can load onto the UCreate with a ‘Write’ command.

But first, you need to rename the sound files you created, and put them in particular directories.  There are 12 buttons on the UCreate for sound files – not including the two at the bottom for your own recordings – so u-load will be expecting to see 12 files with the names ‘sample_0.lop’, ‘sample_1.lop’, ‘sample_2.lop’, etc., up to ‘sample_11.lop’.

So, find the 12 sounds you want to load, rename them like this, and put them in a folder called ‘Samples’.

Type 2, then ENTER, and u-load will ask you to give a name for the file it’s going to create from your samples.  Type in the name, press ENTER, and it will gather the 12 samples together and create the file.

I wasn’t sure whether the file required the extension ‘.upload’ or not, so I created one with and one without.  The first screengrab shows the name ‘Beats’ being entered, then ENTER:

At this point, I had to close u-load, switch the UCreate off and on again, and reopen u-load.

Type 4, then ENTER (‘Write SamplePackage’): u-load will ask you which file to upload.  Give it the name of the file you just asked it to create, it will load the samples onto the UCreate, and you’re ready to go!

*

To add files to the two spaces originally used on the UCreate for your own recordings, create .lop files as above, and put them in a folder called ‘Recordings’.  They can be loaded into the two special spaces by typing 5, then ENTER (‘Write RecordPackage’).  u-load will ask you for the name of the first file to load: tell it the name, and press ENTER.  It will then ask you for the name of the second file: give it that, too, press ENTER, and the two files will be loaded onto the UCreate.

As a matter of fact, marbs recommends storing all sound files in ‘Recordings’ at first, and writing them to one of the two special spaces to check if they work properly, before renaming them and putting them in ‘Samples’.

*

Well, that’s the theory.  In practice, I was able to deal with the two ‘Recordings’ files exactly as described; as for the ‘SamplePackage’ – it looked as if it was working, as in the screenshot above, but the files wouldn’t play, and I don’t know why.  I seemed to be able, using the same process, to reload the files that were previously in it – but I don’t know if I was really doing that, or they hadn’t been removed in the first place and I was just doing something that allowed them to play again.

I assume the .lop files were OK, as I actually tried with alternative sets from the UCreate website, not ones that I’d created myself, so what the problem is, I can’t say.  If anyone reading this knows why – especially ‘marbs’ – please get in touch!  If I find out any more, I’ll post again.

Guitar FXBOX – Part 1, the software

I haven’t blogged about the guitar before, although this is the conventional instrument I play.  There are enough places on the web to find descriptions of commercially produced effects – of which there are many thousands – and, with possibly one or two exceptions, I won’t be writing about these.

However, I thought it might be worth describing some which I’ve been using which were written using PureData and which together I’ve called the ‘FXBOX’.  Some of the effects I’ve bought have cost £100 – £200, but for no extra expense other than a computer – which I assume you must have, as you’re reading this – and a little know-how, all these effects and more are available.

I have to say, I didn’t write the majority of these effects myself. The know-how has come from examples in tutorials or instructions for PureData, like this one: http://en.flossmanuals.net/PureData/ or this one: http://pd-tutorial.com/ or from the PureData forum at http://puredata.hurleur.com/.

A particular inspiration was Pierre Massat’s ‘Guitar Extended’ blog. Pierre has developed a sophisticated system using an Arduino and a Raspberry Pi to make and control the sound of the guitar, which you can read about in detail here: http://guitarextended.wordpress.com/.

In my case, there were a small number of particular effects I wanted to combine in a single unit, and my good fortune was to find this Spectral Delay patch by Pierre Massat (from an original by Frank Barnecht, based, in turn on work by Johannes Kreidler, whose tutorial is referenced above): http://guitarextended.wordpress.com/2012/02/07/spectral-delay-effect-for-guitar-with-pure-data/

This became the central element of the FXBOX project. What is does is to give different delay times to different harmonics of the input, producing an effect which Pierre likens to ‘hundreds of sparks or stars that fall like raindrops or broken crystal’ – a poetic, but apt description.

More conventional echo effects are also available from this patch.

In the FXBOX I put a couple of conventional effects before the Spectral Delay, namely Chorus and Distortion, and a variable attack patch; at the end of the chain I added patches for panning and stereo imaging.

The two interesting effects which remain are a pitch controller – the normal guitar input can be raised or lowered by up to an octave – and an unusual ‘freeze’ effect, which can produce either a static drone or a very odd series of pitches in response to the input.

This screenshot of the main page shows the various effects and their controls:

I’m very grateful to Pierre and others for making their patches available for everyone to use. One or two people whose patches came from posts on the Pd forum, I know only from their online names: ‘ralf’ for the Freeze function [http://puredata.hurleur.com/sujet-6100-fft-freeze-help], and ‘dack’ for the Pitch Shifter from his ‘Guitar multi-effects rig’ [http://puredata.hurleur.com/sujet-2525-guitar-multi-effects-rig].

The various patches which are required to make the FXBOX work are here:

http://homepage.ntlworld.com/andymurkin/Electronica/FXBOX

All the variable controls are available on the screen, but I decided for practical performance purposes it would be best to have presets which could be selected quickly.

In addition, after starting to use the FXBOX, I rapidly came to the conclusion that it needed a controller which was easier to use than the laptop trackpad. For one thing, you can’t adjust a setting manually and still keep playing.

So I set about designing a controller which would also be compatible with PureData. Some controls might practically be operated by hand, but essentially this would have to be a foot controller with buttons and pedals for the various presets, and for adjusting settings which might need to be changed while playing.

The second part of this post describes the controller.

UCreate Music, Part 2

In my first post on the Radica/Mattel UCreate, I mentioned adding In/Out connectors to enable the UCreate to be operated by external controls – e.g. joysticks – or the UCreate Button to be used to control other devices.

This post describes two of my devices which I modified for this purpose: the StyloSim and the Black Widow.

*

The StyloSim is a two-joystick controller, used for simple flight simulation games.  It has two medium-sized joysticks, which are very nice to operate, but no buttons.  Examining the controls using the [hid] object in PureData suggested that the chip it uses would support the use of buttons, but this function is not implemented.

The essential task, then, with the StyloSim was to add two DB9 connectors, matching the connectors on the UCreate, so the UCreate’s effects could be controlled by the StyloSim, and whatever the StyloSim was used to control (at the moment, just one PureData program I’d written to add volume, filter and panning effects to an audio input) could also be controlled by the UCreate Button.

I checked to see that the ‘high’ and ‘low’ ends of the potentiometers in the StyloSim were both the same: they would have to be connected together to control the UCreate, but would also have to be left in a state where they correctly controlled the StyloSim chip.  They were connected, so I snipped the 6 wires between the circuit board and the potentiometers of the right-hand joystick (which I called ‘VR1′ and’VR2′).

The wires from the potentiometers were connected to the ‘Out’ socket, and the wires from the circuit board were connected to the identical pins of the ‘In socket’.  In this way, whatever else I connected, a DB9 lead connecting these two sockets would allow the StyloSim to function as normal.

In fact, I had another addition to make: the potentiometers would have no effect unless the UCreate ‘Hold’ switch was on.  As with the UCreate itself, I added a 3 way toggle switch, centre off, momentary in one direction, latching in the other, and connected this to the appropriate pins on the DB9 ‘Out’ socket.

These connections were enough to ensure that, with the use of DB9 leads, the StyloSim could control the UCreate effects, and that the UCreate Button  - which had only one joystick – could control at least some of the things the StyloSim could control.

However, the UCreate also now had a ‘Volume Pedal’ output, which just required a potentiometer connected via a 3.5mm stereo socket.  As I had two more potentiometers available in the StyloSim, I connected one of these to a switched socket.  I used ‘VR4′, the up/down potentiometer of the left-hand joystick, as this was set not to return to centre when released, so would be very suitable for setting a volume level and then leaving it.  When nothing was plugged in, the joystick would remain connected to the circuit board inside the StyloSim; when a lead was plugged in, it would control the UCreate volume.

(In practice, unlike the volume pedal, the joystick – because of its limited travel, presumably – didn’t take the volume right down to zero, so was less effective than the pedal, but useful as long as complete silence wasn’t required).

I also made two more modifications, which weren’t strictly necessary, but which were not too difficult and, I felt, enhanced the design.

First of all, I chopped off the USB lead and added a socket instead.  This is only because I find it annoying to have fixed leads hanging off devices – it makes them awkward to carry about and store away.  Sometimes USB leads are small and fiddly, but at least they’re colour-coded.

(There is some variation in exactly which colours are used, however.  There’s supposed to be a convention, but as you can imagine, manufacturers find plenty of opportunity to use colour combinations of their own.  Those who are colour-blind – like me – have to be especially careful, but more often than not you can work out which lead is which.  Looking into a socket from the outside, 1, on the left-hand side, is +5v, and should have a red or orange wire connected to it; 2 is ‘Data -’, and has a white or sometimes a yellow/gold wire; 3 is ‘Data +’ and has a green or sometimes grey wire; and 4 is Ground, with a Black or sometimes Blue or Brown wire.  I have come across other combinations, unfortunately, of which those using white or yellow for ground are the most annoying.  Often there are 5 wires, with an extra connection – frequently black – for the shield around the cables.

Looking into a plug, 4 is on the left-hand side and 1 is on the right.

I’ve mentioned before that I shouldn’t be using Type A sockets as the output of a device that’s being connected into a Type A socket – on, for example, a computer.  I should be using a Type B or mini USB socket; but this rule is to avoid connecting two devices together that both supply power, which might cause excess currents and start fires, and this isn’t going to happen with the devices I’m using – the StyloSim, for example, receives 5V from the computer, but doesn’t provide power of its own).

Finally, partly as an indicator that the connection with the UCreate had been properly made, partly because flashing lights are always good, I also added blue and red LEDs to the front of the StyloSim, and connected these to the relevant pins on the ‘Out’ socket.  These flash in time with the rhythm of the sounds from the UCreate, when the ‘Hold’ switch is on, so you can tell if the UCreate is ready to receive instructions.

These pictures show the new additions to the circuitry:

and this one shows the StyloSim in operation, controlling the UCreate, with the DB9 and 3.5mm connectors in place, the volume set fairly low, the ‘Hold’ switch on, and the LEDs flashing:

*

Essentially, I did exactly the same to the Black Widow.  The big joystick on the right-hand side was connected to the DB9 ‘Out’ socket, and the ‘throttle’ on the left-hand side to the 3.5mm volume socket.

In this case, however, there were buttons available, so I was able to use the ‘F4′ button on the top of the joystick as a momentary ‘Hold’; for a latching Hold, I added an SPST switch at the bottom of the front panel, plus the two LEDs, which are illuminated when either Hold switch is activated.

And this is the rear of the instrument – not that neat, but it all works:

*

As for the UCreate itself, I made three further changes – but these were more additions, rather than modifications:

1.  External power supply. I was pretty certain the UCreate would work with a 5v supply, and was about to use an old mobile phone charger for this purpose; but while I was looking through things I had lying about, I found a better quality one which I’d been given and which was rated at 5.5v, 350mA.  I replaced the connector with a 3.5mm mono plug to match the socket I’d installed in the UCreate, and it seemed to work perfectly, cutting out the battery supply when plugged in, and powering the device.

2.  Switch box.  It occurred to me that there might be occasions when, if I was using the Black Widow and the UCreate at the same time, it might be handy to be able to swap the joysticks quickly from controlling one thing to another, and a way of switching the DB9 leads from one device to the other would be useful.

I was looking into buying a DB9 switch box, which would have been about £5 – £6, but in the end I decided to be stingy and bought three DB9/DB25 adapters for about £4, as I found an unused two-way DB25 switch box amongst my stuff.  I had bought this for an as-yet-unrealised MIDI project: as this will probably remain unrealised for some while, I thought I might as well use it in the meantime.

I did find a diagram on the internet to show how the DB9 pins were, according to the RS232 standard, allocated to pins on the DB25 connector, but it doesn’t really matter, as all 3 connections (in/out, A and B) will be the same.

The Black Widow ‘Out’ is connected to the ‘In/Out’ socket on the switch box; ‘Out A’ is connected back to the Black Widow ‘In’ socket; ‘Out B’ is connected to the ‘In’ socket of the UCreate.

3.  Feedback circuit.  As the UCreate has input and output sockets next to one another on the back, I thought a circuit that connected part of the output signal back into the input, in conjunction with some of the effects – for example, the filter or flanger – could potentially produce some interesting sounds.  Having just finished the modifications described in my first post and put everything back together, I decided to do this externally, so added two 3.5mm splitters to the line in and out, and connected them with a lead containing a volume control:

This allows some feedback sounds to be added to the linked samples or to sounds at the line or mic in sockets, and the amount can be limited by the volume control.

UCreate Music, Part 1

This post concerns a very interesting device which was made a few years ago by Radica, a Mattel company.  It was manufactured for a very short time between 2009 and 2010, and supported only until 2011, but examples still appear on eBay, sometimes for very reasonable prices. I got mine for less than £10, which I thought was pretty good for a comparatively sophisticated machine.

The way it works is by playing loops, which you can choose from its memory – one each from 4 banks of 3 loops, in the categories ‘Back Beats’, ‘Riffs’, ‘Licks’ and ‘Runs’ – and apply effects to.  You can also record two of your own samples to add into the mix.

This is how you would normally use it (ignore the toggle switch and sockets on the left-hand side: this and other modifications I made are described later):

There are two reasons why the UCreate captured the imagination of electronic music-makers.  First of all, you could connect it to your home computer via a USB socket on the back and make use of software that would allow you to save recordings of songs, reorganise the loops and effects and download new loops from Mattel’s UCreate website.  I’ll return to this topic later.

The second thing was the range of 8 special effects, and the fact that these are available not only to the loops played back by the UCreate, but also to any audio source connected to the Mic or  Line in sockets.  The effects – referred to as ‘FX and Filters’ – comprise Tremolo, Distortion, Flanger, Phaser and Echo, a variable low-pass filter, and two unique and unusual effects called Forward/Reverse Looper and Rewind Spin Looper.  These work by recording very short samples and replaying them in various ways controlled by the user.

The way the effects are controlled is also highly unusual:  a large Button on the front panel can be pushed and tilted left/right and up/down to vary two parameters of the effect – for example left/right controls the speed of the flanger, up/down controls the depth.  If you find a setting you want to leave for a while, a ‘Hold’ button fixes it where you’ve set it until ‘Hold’ is pressed again.  The fact that the whole Button is lit up when in use  with flashing blue LEDs is just the icing on the cake.

Leaving aside the loop playback feature for the moment, this effectively makes the UCreate an inexpensive, but versatile multi-effects unit, playable in real time.  Although only one of the effects is available at a time, the Forward/Reverse Looper and Rewind Spin Looper in particular, together with the ability to control these in real time with the Big Button, makes the UCreate a useful and unconventional device to have.

I began by using the UCreate in this way, making just a couple of small modifications to it.

First, I added a socket for an external power source; then, as I had done with a number of my other instruments, I added banana sockets for connecting a larger external 8ohm speaker and a DPDT switch to cut out the internal speaker when this is in use.  There is a headphone/external speaker socket on the back of the Ucreate (which also cuts out the internal speaker when a plug is inserted), but this is a 3.5mm stereo socket, as you would find on a PC or mp3 player and is more suitable for use as a Line out.

These are the audio and USB sockets on the back of the device:

Next, imagining a situation when both hands might be occupied in operating the loops and effects and not able to control the volume, I added a socket for an external volume pedal.  This was a 3.5mm stereo socket with internal switches, like the Line out socket.  I used a small size socket purely because of the lack of space inside the case.

On the small circuit board attached to the on/off/volume control, I broke the connection to the centre of the volume potentiometer and rewired it to the socket so that when nothing was plugged into it, it was connected directly to the main Ucreate circuit board, as originally designed; when the volume pedal was plugged in, the potentiometer in the volume pedal was added into the circuit.  This would enable the maximum volume to be set by the original volume control and the pedal to move between this and zero volume.

The pedal itself was simply a cheap second-hand Bespeco volume pedal.  I removed the original sockets and the circuit board inside and connected the potentiometer to a 3.5mm socket, wired in a similar way to the socket inside the Ucreate.  The tip was connected to the input from the Ucreate and the ‘high’ end of the potentiometer in the pedal, the sleeve to Ground and the ‘low’ end of the potentiometer, and the ring to the potentiometer wiper, the centre tag. (I should have built this before, as it would have been useful with many of the instruments I had made or modified, and I’ll have to consider retro-fitting sockets to them so it can be used).

I then decided to take a closer look at the big control Button.  I tried dismantling the mechanism, but couldn’t seem to get it completely apart.  This may have been because it was pressed or glued together after the circuit board was wired in, and I wasn’t going to risk breaking it by trying to prise it apart if it wasn’t meant to do that.

However, I got it apart far enough to see that it used a joystick for the left and right and up down movement.  This was mounted on a small PCB, and on the bottom of the PCB there were three momentary switches, set out in a triangle.  These were like the ones you often get on game controllers: they’re soft and squishy, and when you press them they join two contacts on the PCB; when you take your finger off, they spring back into shape and the connection is broken.

All three switches were connected the same, and later experimentation showed that they had exactly the same function as the ‘Hold’ button, except they were momentary instead of latching.

This gave me two thoughts: first of all, with essentially a joystick and a momentary switch under it, it would be possible to use the UCreate’s Big Button to control another instrument or effect that normally used a joystick or two separate potentiometers; and secondly, there was no reason why the UCreate couldn’t be controlled by two potentiometers or an external joystick.

The way to do this would be to put the Button back and separate the connection between the UCreate’s main PCB and the Button PCB, and then route these connections elsewhere.

The link was made with a 9 way ribbon cable; the names of these 9 connections were printed on the main PCB, and even where this didn’t mean a lot, it was easy to follow the the tracks on the Button PCB and see what their functions were.  So I cut the cable.

From top to bottom, the connections were:

VCC_33 – which connected to one side of all three switches

IOA15 – connected to the other side of the three switches

GND_ADCVCC33 – connected to one end of the two joystick potentiometers (the ‘low’ end, presumably)

LINE 3 – the centre tag of one of the potentiometers (the ‘up/down’ one, which I called ‘Pot 2′)

ADCPVCC33 – the other end (‘high’ end) of both potentiometers

LINE 2 – the centre tag of the other (‘left/right’) potentimeter which I called ‘Pot 1′

R154_1 – one side of one of pair of surface-mounted blue LEDs on the Button PCB

R68_1 – one side of the other LED

V BAT – the other side of both LEDs (+6v, presumably)

Essentially what I did was to connect the end of the ribbon cable that came from the main PCB directly to a DB9 connector on the back on the case.  This was marked ‘In’.  The other end of the ribbon cable, the one from the Button PCB, was connected to another DB9 connector, marked ‘Out’.

In this way, if you wanted to control the UCreate from another device – a larger joystick, perhaps – all you would need to do was connect it to the DB9 ‘In’ socket; if you wanted to control another device with the Big Button, you would connect the other device to the DB9 ‘Out’ socket; and to use the UCreate as normal, just connect the two sockets together with a DB9 cable.

In order to make space for the DB9 sockets, which are quite big, I had to remove part of the bottom half of the case, which stuck up inside:

I think this was probably a carrying handle, but I didn’t think I needed it, so I sawed it off and created a lot more space in the back of the case.

In fact, I didn’t connect the Button directly to the ‘Out’ socket.  Although the Button works brilliantly well for the Reverse and Rewind ‘stuttering’ or ‘scratching’ effects, there was a lack of precision when it came to such things as the filter cut-off frequency, speed and depth of flanging, and so forth.  Apart from anything else, joysticks don’t usually use much of the possible travel of  an ordinary potentiometer, so there was also a restricted range over which the Button was operating.

So I decide to squeeze a couple of potentiometers into the case, which would be selectable in place of the Button.  The two connections for the centre tags of the potentiometers (‘Line 2′ and ‘Line 3′)  coming from the Button PCB went to one side of a DPDT switch, and the poles went to the DB9 ‘Out’ socket.  The wires from the other side of the DPDT switch went to the centre tags of two potentiometers, which I squeezed in the front of the case.  The two connections for the ends of the potentiometers went to the potentiometers and to the socket.

In the event, I also added a 10k preset, set at about halfway, in the circuit at the potentometers’ ‘bottom’ end: it seemed to me that some parameters – e.g. the filter cut-off frequency, and the volume pot when using the tremolo effect – were going too low, at the expense of effects that could be obtained with higher resistance.

The way the UCreate works, the potentiometers  - and the potentiometers under the Button, come to that – have no effect unless one of the ‘Hold’ buttons is pressed, so I needed to add a momentary button, preferably somewhere near the potentiometers.  There was just about room, and what I decided to use was a toggle switch with a centre off position, momentary on in one direction, latching on in the other.  This would enable me to engage the momentary switch, adjust a potentiometer, then when I had exactly the sound I wanted, latch the switch on.  So the two connections for the switch went both to the DB9 socket and to this new switch.

In fact, they went to a third place: a standard (1/4″ or 6.35mm) mono jack socket to which a ‘Hold’ footswitch could be attached.  I used a standard size jack in this instance because I had some nice ready-made footswitches: they’re apparently sold for use with tattoo machines, but come with standard jacks attached, which is very handy.

So that’s how I modified my UCreate, producing a versatile and quite easy to use multi-effects device.  This picture summarises most of the changes I made:

I don’t know if they’re all the same, but the grille on the front of mine came off very easily, so I took the opportunity to remind myself of what the 8 effects are, and what order they come in.

The front and back of the device now look like this:

and here’s what it looks like in operation:

The Electronic Canary

OK, this wasn’t the most versatile circuit I’d ever built – but it wasn’t the most difficult; and, more importantly, it illustrates an important source of inspiration which could easily be overlooked – the ‘educational’ electronics kit.

I remember when I was very young, receiving one of these for Christmas, produced by Philips – my first exposure to the world of electronics, I suppose – and being fascinated by making lamps flash on and off (this was before the invention of the LED!), and listening to Radio Luxembourg (and before Radio 1 . . .) through a single crystal earpiece.

So I was delighted, when staying in Leeds for a few days recently and popping into a small neighbourhood charity shop, to find this Radio Shack 30-in-One Electronic Project Lab.

It only cost £2, and seemed remarkably complete.  Amongst the sound producing circuits described inside were The ‘Invisible Power’ Radio, The Bee, The Highs and Lows of Oscillation, The Two-Transistor Oscillator and, of course, the Electronic Canary:

Each one was described like this, the various different components are explained, and users are encouraged to experiment with the circuits they make.  The booklet gives not only the necessary wiring instructions, but also a circuit diagram.  This would enable you, should you be so minded, to solder together a permanent version of any of the projects described.

I lost no time in getting down to cutting and stripping wires and joining them up using the little spring connectors, and in just a few minutes, my board looked like this:

It wasn’t necessary to put the earpiece in the ear – the sound was perfectly loud enough to be heard around the room, a point on which my very patient hosts were too polite to comment.

In fact, I did need to replace the earpiece, which wasn’t functioning quite correctly, but the other components I needed all worked fine.

So, my advice is to keep an eye out for these kits if you see them going cheap: there are bound to be music or sound circuits you can make – and there’ll be opportunities to modify them if you have the knowledge and expertise.  Mine is certainly now going cheep and will, I’m sure, find its place alongside, perhaps, the Cracklephone, which also has distinct bird-like qualities.

Sounds of Nature with the SoundSpa

After finishing the Black Widow, Mk II, I had some fun using natural sounds with it – rain, birds singing, that kind of thing.  The final feature I had added to it was the ability for it to play automatically, which, though not quite as good as manipulating the samples with the Flight Controller throttle and joystick, would allow the Black Widow to keep operating, leaving the hands free to play other instruments at the same time.

While I was thinking about natural sounds, I came across an interesting-looking device called the SoundSpa, which is designed to soothe you to sleep with sounds of nature, including rainforest, ocean, rain, waterfall and ‘summer night’.  This sounded like a good candidate for some circuit-bending!

These machines do appear on eBay from time to time, and I was very fortunate to get one there for three or four pounds, including a handy mains adapter.  I hadn’t really done any circuit-bending, except for my work with Stylophones – like the Alien and the Hedgehog – but I described that as modification, in the sense that I knew what kind of circuit I was dealing with, and I knew more or less what I wanted to do with it.  The art of circuit-bending is, by tradition and by its very nature, more experimental.

As it happened, just at this moment, I came across an ad for a circuit-bending workshop, which I decided to attend.  So the story of the creation of the BentSoundSpa is also the story of my day at the wonderful Music Hackspace in Cremer Street, London.

The workshop was organised by Susanna Garcia and run by Tasos Stamou.  Tasos has created some great music, partly using circuit-bent instruments; you can read about him and hear some examples of his music on his website: http://www.tasosstamou.com and some examples of the circuit-bent instruments he uses here: http://stamouinstruments.blogspot.co.uk.

There were about a dozen of us taking part, and enough musical toys to go round.  I took the SoundSpa along and worked on that.  First we were encouraged to start our devices playing, take leads with crocodile clips and find connections which produced effects that were, well, interesting, unexpected, dramatic . . . all of which we did. There were glitches, repeats, jumps and crashes  - but the devices seemed surprisingly resilient: when they crashed, it was just a matter of taking a battery out for a moment, putting it back, and everything would be working again.

The workshop was designed to be suitable for complete newcomers – which several of us were – so we were deliberately encouraged not to speculate on what the parts of the circuits were for, and proceed by exploration.

Nothing very exciting was produced by the SoundSpa until we looked specifically at making pitch changes, dabbing at resistors with a damp finger.  Having found the right one, we cut it from the board and replaced it with a potentiometer.  A range of values was provided, so we could pick the most suitable, and soldering irons were available to replace the crocodile clips when we were happy with our choices.

I found a very high value, 2.2M, which took the pitch of the sounds right down to almost standstill.  The only problem with taking the pitch up high was – just as I had experienced with the Stylophones – the resistance was so low the device crashed, so a small resistor was needed to stop this happening.  I calculated this would be around 10k.  On the day, I found a value high enough to prevent crashing, and when I got home I replaced this with a 10k preset, setting it just high enough to keep the device running when the potentiometer was at maximum.

After this picture was taken I decided to add a second, smaller potentiometer – 100k – in series with the large one as a kind of ‘fine-tune’ control.  This has a much larger effect at higher pitches when the resistance is low on the bigger one.

Then, to finish, we went down to the workshop, drilled holes for the potentiometers and added a 3.5mm output socket which would cut out the speaker when plugged in.

I considered replacing the original on-off switch/volume control since it was a bit dodgy – as was the one on Tasos’s example, one of several of his instruments which he brought in to show us – but it works OK with a little persuasion.

Obviously, the effects available from it depend on the wide pitch and speed variations available.  Like the Black Widow, the addition of a variable filter such as the StyloSim or the Active Low-pass Filter would add to its versatility.  I haven’t used it properly yet, but I’ll post a sound file as soon as I have one.

The Black Widow MkII and more on sample manipulation

I recently came to use the Black Widow sample manipulator again and decided to make some improvements to it.

Initially, I improved some of the existing features: what I had at first perceived as ‘unreliability’ in its responses to button commands was really only a difficulty in making the required number of presses to ‘fix’ volume, filter or pan settings on a sample.  This involves making one, two or three quick presses on the hatswitch button – like making a single, double or triple mouse-click reliably – not perhaps one of my best ideas, and a little difficult to execute, especially without any feedback as to whether the fix had been made or not.  So I added indicators to show if the fixes were on or off.

I then moved the sample arrays onto the same page, so there was no need – as there had been in the original – to move from page to page on screen to see what was happening with the samples being played.  This screen also now has the various indicators for volume, filter, playing position within the sample, etc., visible next to the representation of the waveform in the array:

Above each array are also indicators ‘REVERB ON/REVERB OFF’ and ‘ECHO ON/ECHO OFF’.  I thought simple versions of these two effects would be useful, and added them to each of the output stages.  Having run out of buttons on the Black Widow, these effects are manually turned on and off by using the ‘W’ and ‘E’ keys on the computer keyboard.  (Although ‘E’ makes sense for echo, ‘W’ doesn’t for reverb, but I chose these two letters so as not to conflict with another program I commonly use, where ‘R’ and ‘V’ have particular functions; and ‘W’ happens to be next to ‘E’).

Also, on the right-hand side, indicators for ‘Auto Speed’ and ‘Number of sections done’ refer to the next change I made: a function to allow the Black Widow program to run automatically.  It’s designed to do this in a similar way to the way it’s operated manually: one sample at a time is operated on, and there are built-in delays between stages to allow a particular selected combination of settings to play for a while before being changed.

The particular settings are chosen by random numbers generated by the program.  In normal operation this will allow them to range between minimum and maximum values in each case.  In the event that this wide range is not required (for example, turning the  filter to minimum usually has the effect of turning the sound off, as does turning the playback speed to zero), there is now a page on which minimum and maximum values can be set for each parameter of each sample:

It’s possible here to fix any parameter, so it cannot be changed during ‘auto’ mode; and to fix any sample to play back at normal speed, full volume, average filter setting and centre panned.

Finally, I added a third section in which samples can be recorded into the arrays, rather than loaded from existing sound files.  This could be useful for variety in a ‘live performance’ in which short samples (90 seconds is PureData’s default maximum) could be recorded and almost immediately loaded and manipulated, manually or automatically.

The screen that opens automatically when the app is run has had to be expanded, and now looks like this:

It looks a bit complicated, but I haven’t mastered the method of producing a neat GUI; however, it works fine, and is easy to use if the numbered steps are followed.

The PureData file on which the app is based is here.

It’s perhaps worth mentioning here, as I haven’t done so anywhere else, a very handy little device I used, not wanting to have a full-sized keyboard in the way when I was trying to use the Black Widow controller.  This is the so-called iPazzport, a miniature keyboard and trackpad, which connects to the computer by USB.  It looks like this:

If you imagine that the trackpad is slightly smaller than the trackpad on a laptop, this gives you an idea of just how tiny the keys are – but I was still able to press the correct key when required.  In case of difficulty it comes with a stylus and a small attachment which fits on the end of the finger, with a point underneath it.

Mine was about £10 off eBay, but I don’t know how readily available they are now: the one you see these days is a wireless version which costs a bit more, between £15 and £20.  There are now some similar items – also wireless – styled rather more like TV remote controls, which look as if they might have larger keys.

Also worth mentioning here, if you’re interested in this kind of thing, is the work of Karlheinz Essl.  You can read about him here: http://en.wikipedia.org/wiki/Karlheinz_Essl_junior and on his website at http://www.essl.at/.

Of particular relevance in relation to the Black Widow project is one of the various programs Essl has created for sample manipulation, fLOW, which is described like this: ‘fLOW . . . generates an ever-changing and never repeating soundscape in real time that fills the space with flooding sounds that resemble – metaphorically – the timbres of water, fire, earth, and air. This ambient sound scape generator adjusts itself through various parameters and controllers that are represented in real time on your screen.’  It comes with four samples, which are are loaded and manipulated automatically via a comb filter, ring modulation, frequency shift and a flanger, but you can also load your own samples into it for a more personalised experience.

You can read about fLOW and download it from http://www.essl.at/works/flow/download.html.  I don’t remember how much it costs, but it isn’t very much, and well worth the money.

It’s well worth reading more about Essl, too.  His main preoccupation is one that I’ve become interested in in recent years: ’music that is created at the moment of its sounding (“realtime composition”)’. Improvisation, looping and manipulation of sound and music samples are all part of the same field, and chance can play a significant role.  In a later post I’ll be mentioning another his programs, REplay PLAYer (found at http://www.essl.at/works/replay.html), a multi-featured program for manipulating a single sound sample.

Finally, a very nice freeware program with similar features which I’ve used is Sineqube’s Sapling, which is described here: http://www.sineqube.com/blog/?page_id=157.  Sapling provides an easy means to load 4 samples and vary – manually or automatically – the speed and volume of playback, and the length of a loop created within each sample.  It works well with short samples and allows a very useful combination of automatic and manual adjustments to be made in real time as the samples are played, and a facility for recording the output to disk.

With all of these programs, including the Black Widow in Auto Mode, I enjoy setting up the conditions and listening to the samples play, although for a more permanent record of what was created, I normally edit the recorded results.

[Edit: more modifications have been made to the Black Widow (although not for sample manipulation). See this post].

Stylophones 4 – The Stylophone 350S, Part 2: Simple mods

After opening my 350S and giving it a good clean, I decided to carry out a few simple mods before putting it back together again.

The first thing I did was to detach the external connections to the two circuit boards to make it easier to take everything apart and get at.  These connections were:

• keyboard
• speaker
• power
• styluses
• pitch

I carefully desoldered the wires, and replaced them with 2 or 3-way Molex connectors, like these:

*

At this stage I decided not to make any further modifications to the keyboard.  The keyboard PCB now plugs into the main circuit board and is much easier to remove for cleaning and for further potential modifications.

This picture shows the front and back of the keyboard PCB with a 2-way Molex socket fitted, connecting top and bottom of the chain of resistors which produce the different notes:

*

In the case of the speaker wires, the Molex connector makes it easier to move the main circuit board around while working on it, as the wires are no longer than absolutely necessary and the speaker is firmly fixed to the top half of the Stylophone body.

The modification I made to this – which I’ve done with several of my instruments recently – was to add a switch to swap between the internal speaker and a larger external speaker (as described here).  I chose a large DPDT rocker switch, which seemed to be in keeping with the 350S’s style.  I don’t know how necessary it was, but as the internal speaker is 35Ω and the external speaker is 8Ω , I added a 3W 27Ω resistor in series with the output, which is a pair of 4mm banana plug sockets.

*

As far as power was concerned, I first wanted to replace the large PP9 batteries.  Not only are these heavy and expensive, but they take up a lot of room inside the Stylophone case, which might be needed to house extra circuitry.  So what I decided to do at this stage was to replace them with something more practical: rechargeable PP3′s.

I wasn’t sure these would be powerful enough to allow the 350S to function properly, but I exchanged the PP9 wiring for PP3-sized battery clips and everything seemed to be working.  I then looked for some PP3 holders that would provide a more permanent fixture for the batteries.  This type seemed to fit the bill:

There was just enough room to fit these side by side into one of the covers formerly used for access to the PP9′s, each one attached with 4 small nuts and bolts.  Although opened from the outside, these battery holders occupy the internal space originally taken up by one of the PP9′s.

Clips for the two PP3′s are connected to the power Molex connector via a 3.5mm mono socket with an integral switch, so that anything plugged into the socket automatically disconnects the internal batteries.

Later, the socket might be used for an 18v power supply, but for the time being I attached the discarded PP9 wiring and clips to a 3.5mm plug, so that PP9′s can still be used, but don’t have to be installed inside the body of the 350S.

*

Unlike the regular Stylophone, the 350S has two styluses: one for normal playing, sounding continuously for as long as the stylus is in contact with the keyboard; and one for ‘Reiteration’ mode – with the appropriate switch selected – producing a fast or slow series of pulses, in imitation of a banjo or mandolin, on which it’s common to pluck a single note repeatedly.

However, I had found while modifying normal stylophones, that it was sometimes handy to have two styluses, one in each hand, for playing quicker or more intricate passages; so I decided to rewire the two existing styluses as standard, and add two extra ones for Reiteration mode.

With the Molex connectors in place, it was easy to wire all four styluses up, but not so easy to find a way to secure the extra two to the Stylophone in such a way that they would be easy to reach.  In the end, I used a pair of clips like this, sold on eBay as penholders and meant, I think, to clip onto a pocket:

I had some spare white styluses, so the ‘normal’ styluses are black, and the ‘reiteration’ styluses are white.  I attached a holder each side of the Stylophone in which the white styluses sit.  The long wires attached to these can be pushed inside the body of the Stylophone when not in use.

I wasn’t able to find an exact match for the wire used by Dübreq for attaching the styluses.  It’s only just over 2mm in diameter, and very flexible; there are no more than 10 or 11 strands of wire inside quite a thick outer layer, and a non-conductive cord running along the length of it, on the inside – presumably for strengthening.  If I ever find out where to get it, I’ll add it as an Edit to this post: in the meantime I had to make do with a standard multi-stranded white ‘hook-up’ wire of about the same width.

*

Dealing with the pitch of the 350S didn’t involve detaching external wires, in fact, but I added a 3-way Molex connector to the tuning control to make it easier to experiment with.

Unlike some of my previous Stylophone mods, I wasn’t looking for extreme pitch changes this time, but something more along the lines of a synth modulation wheel.  Strangely, these seem to be very rare, but I found one produced by the German company Doepfer, described here.  It comes as a kit of parts, like this:

The pot supplied with it is a 10k, which has a knurled shaft fitting tightly inside the hole in the ‘half-wheel’.

I  wired the wheel in parallel with the existing tuning control, and its effectiveness depended on three things:

1.  The setting of the tuning control: the higher it was set, the less variation produced by the wheel.  Not much I could do about this, as the tuning control is used to set the 350S to the correct pitch, compared to other instruments.  If it proves a problem, it could perhaps be solved in the future by a slight adjustment to the keyboard resistor chain.

2.  The value of the pot.  I found that a 2.5k pot was the most effective, but couldn’t find one with a shaft compatible with the Doepfer wheel.  So I added some 10k resistors in parallel with the 10k pot.  Originally I added 3, which would have made the pot 2.5k, but 2 seemed to be enough (3.3k), and took up less space, so I left it at that.

3. The part of the potentiometer track covered by movement of the wheel.  The wheel wasn’t able to move the wiper of the potentiometer round the whole track – which is normal for mod wheels, joysticks, etc.  It took a bit of experimentation to find the right place, which essentially meant turning the potentiometer to exactly the right position before attaching the wheel.  It needed to be at zero when the wheel was deflected fully down, and eventually I found the right place, wired leads and a Molex plug to it and fixed it in place with small nuts and bolts.

The whole construction took the place previously occupied by the left-hand PP9, with the wheel appearing through a slot in the top of the 350S .  The Doepfer kit cost about £10, so it was a bit of an extravagance, and something like it could probably be rigged up more cheaply.  However, it adds an interesting feature to the 350S which it never had before.

This picture shows the pitch wheel assembly in place and also, in the background, the speaker switch and banana sockets.

This picture shows the top half of the 350S body, with the new components and the Molex connectors in place, with the circuit boards removed:

*

After fitting everything, it was time to put the 350S back together.

The first item to go back into place was the main PCB.  This picture shows the board in position, with the 6 fixing screws marked:

Next, the Keyboard PCB was installed.  The 4 fixing screws are marked:

Before the bottom half of the 350S body was attached, the PP3 battery clips were fed into the battery holders:

The two halves of the body were fitted together and batteries inserted:

Finished and ready to go!  The underside of the 350S now looks like this:

and the front and back like this:

This gives a good view of the power socket on the back left, as you look at it; the speaker switch and sockets on the back right; the pitch wheel on the top on the right; and the white ‘reiteration’ styluses in their holders.

*

Finally, with the 350S back together and in operation, I looked at the suggested external addition, a volume pedal.  According to the 350S manual, this would replace the photo control, and adjust not only the volume, but also the waa filter and the vibrato depth.

The manual recommends a ‘standard Foot Pedal’: but what was a standard foot pedal in the 1970s is not what we might consider a standard foot pedal – or ‘expression’ pedal – these days.  What’s required here is a 50k-100k log pot, which plugs into the 350S via a 6.35mm (1/4″) mono jack plug.

I had an old volume pedal (probably dating almost from that era!) which I was able to adapt.  The original cable was crackly and the pot was scratchy, so I shortened the cable to remove the section that was obviously damaged inside, and replaced the pot.

I didn’t have a 47k or 50k to experiment with, so I used a 100k, but that seemed to be fine.  The only oddity is that the ‘waa’ works backwards, in that the filter is at the ‘high end’ with the heel down – as compared to, for example, a guitar wah pedal, where heel-down is the low end, and toe-down is the high end.  I tried putting a polarity change switch in the pedal, but that didn’t work, as the pedal mechanism – just the same as the pitch wheel described earlier – is set to reach its minimum when the heel is fully down, and doesn’t cover the full travel of the pot, so when the two ends of the pot were swapped, the pedal wasn’t reaching zero, which it needed to do to produce the full ‘waa’ effect.  I’ll just have to get used to it.

After playing the instrument for a while, I noticed that one of the switches was a bit crackly, so this is something I might tackle later on, together with a couple more mods I have in mind.

The Black Widow – Sample Manipulator

I was lucky enough to get a decent joystick for a good price from eBay, a Speedlink ‘Black Widow’:

Obviously, this is a bit more sophisticated than the controller with small joysticks which I used for the StyloSim, and the kind of thing used for semi-realistic flight control programs, having a throttle on the left-hand side and joystick on the right, and a nice feel to it.

As for musical applications, it connects to a computer via USB, as can be seen in the photo, so is very suitable for use with PureData’s ‘hid’ (Human Interface Device) function.

I thought at first of designing an instrument, but felt this would be less easy to operate than the Cybersynth or Theresynth, which used gamepad-type controllers.  An effects device like the StyloSim would be possible, but at the moment would seem like going over old ground: something I might come back to later.

Instead, I decided on a device for manipulating samples.  At the time I thought this was an original idea, but I have subsequently discovered someone who does this – in quite a different way, but using the same kind of device: http://www.youtube.com/watch?v=9wdb6-QLnQ0.

That was Johannes Kreidler, a very interesting contemporary German composer making music with cutting edge electronics.  Check out his website at http://www.kreidler-net.de/english/index.html or type ‘Johannes Kreidler’ into YouTube.  Kreidler is also an expert on PureData, as it happens, and wrote the very good tutorial ‘Programming Electronic Music in Pd’, translated into English at http://www.pd-tutorial.com/english/index.html.

So, I used PureData to create a simple device for manipulating samples with the Black Widow.

There are 4 buttons on the front of the Black Widow:

These are used to control up to 4 separate samples: when the Black Widow Pd app is first opened, the buttons are pressed one by one to load the 4 samples; when in use, the throttle, joystick and other controls operate on the sample whose button was last pressed.

As for the other controls, these are: on the left, the throttle, which includes two buttons on the reverse,left and right:

Pressing the left-hand button on the back of the throttle and moving the throttle forward from the central position increases the sample playback speed; moving the throttle backwards from the central position increases the playback speed, but plays the sample backwards.  When the button is released the speed and direction are fixed.  Pressing the right-hand button on the back of the throttle toggles the playback direction.

Moving the joystick to the left pans the sample to the left; moving it to the right pans it to the right.  Moving the joystick forward from the central position increases the playback volume; moving it backwards from the central position increases the centre frequency of a bandpass filter.

The other buttons on the joystick have the following functions:

The button F3, when pressed as the sample is playing, sets the start point of a section within the sample; F4 sets the end point.

The hatswitch has 4 functions: RIGHT cycles playback from the start point to the end point of a section set with the F3 and F4 buttons; DOWN pauses playback; UP Restarts playback; LEFT fixes a volume or filter setting.

F2 restarts playback at normal speed from the beginning of the sample.

This is the control window, in the process of loading a sample:

There are some problems with it: it doesn’t always seem to respond to instructions, particularly the ‘Fix’ command from the hatswitch, don’t quite know why.  Also, the interface isn’t very practical: if you want to look at what you’re doing as well as listen, there are separate windows for each sample which you have to bring to the front.

However, it mostly works as described, and is quite entertaining to play with!  I’ll post a sound file shortly.

[Edit: the Black Widow has been superseded by the Black Widow, MkII, described here].

Bigfoot – automatic/remote stylophone control, Part 2

I was  constructing a device to play a modified stylophone remotely and automatically.  Using a 16 way analogue switch, the 24-pin 4067 chip, I designed a device where any one of 15 intervals on a 2-octave tonic sol-fa scale would be triggered by changing the chip’s 4-bit binary input.

First of all, I had used a physical control, a 16 position binary or hexadecimal rotary controller; what  I needed next to find was chips that could be made to output sequences of 4-bit binary numbers.

There are several of these, and I went for the 4516, which is a pre-settable binary counter.  It can, if left alone, repeatedly count upwards from 0 – 15, outputting numbers in binary form (’0 0 0 0′ to ’1 1 1 1′) on the pins marked ‘Q1′ to ‘Q4′ in the diagram below at the speed of a pulse connected to its clock input (Pin 15); or downwards from 15 – 0.  But by pre-setting a certain number, in binary form, on 4 extra binary inputs, marked ‘P1′ – ‘P4′ in the diagram, it can also be made to count upwards from this number to 15; or downwards from this number to 0.

This is how the 4516 is usually represented in circuits:

Q1 – Q4, as mentioned above, are the outputs, and P1 – P4 are the inputs for the number the count starts from, both in the form of a binary number.  The ‘Preset Enable’, pin 1, is usually held low (0v): when it’s taken high (+v) the number on the inputs P1 – P4 is loaded in and the next count starts from that number.  ’Preset Enable’ is sometimes referred to as ‘Load’ for this reason.  The ‘Carry Out’ is normally high, but goes low when the count ends.

The ability to count downwards from a set number would be useful for an arpeggiator, which could be set to repeat a sequence with a length of 2 – 16 notes, using the rotary encoder, described in Part 1, connected to the 4 binary inputs to preset the sequence length.

The circuit for this device was extremely simple, requiring only the rotary encoder, a momentary switch to tell the 4516 to load the sequence length number, an on/off switch and two inverters from a 40106 (which has 6 in it altogether) .  One of the inverters was connected as an oscillator, which was connected to the 4516′s Clock input: this determines the speed at which notes sound; the other inverter was connected between the ‘Carry Out’ and ‘Preset Enable’ pins: the ‘Carry Out’ is normally high, so the inverter keeps the ‘Preset Enable’ low; when the count ends the ’Carry Out’ goes low and the inverter sends a ‘high’ pulse to the Preset Enable, reloading the start number.

Pin 10 is connected to 0v in this circuit, which tells the 4516 to count down, not up: this was the easiest way to make sure it counted the right number of notes in the sequence.

In fact, counting up or down would result  only in a scale or part of a scale being played, so I made the output a bit more interesting by reversing the 4 outputs.  Instead of connecting the A output of the 4516 to the A input of the 4067, the B output to the B input, etc., I connected it so that A B C D were connected D C B A.  In essence this meant that consecutive notes in the sequence would not be consecutive notes in the scale, which I thought would be more interesting.

This produced method 2 of controlling the Stylophone: automatic arpeggiation.

*

The third method of controlling the Stylophone automatically used 3 more of the inverters in the 40106 which had been used for the 4516 clock and ‘Carry Out’ inverter.  The inverters were wired as oscillators.

This was the idea that came from the ‘Slacker Melody Generator’, described at http://electro-music.com/forum/viewtopic.php?t=27239&postorder=asc&start=50.  Each of the 4 oscillators is connected to one of the 4 inputs of the 4067; each runs at a different speed, changing the value on that input from low to high, or 0 to 1.  The different successive combinations of 0s and 1s produces a random melody, which can be changed by adjusting the speed of the oscillators, increasing or decreasing the rate at which each particular input changes from ’1′ to ’0′.

The reason the four oscillators have two capacitors each is simply because the original circuit I used suggested values of 220n; I soldered these in place, but the oscillators seemed to run too fast for my liking, and it was easier to add new ones in parallel than take the old ones out and replace them.  The result of putting capacitors in series is the opposite of putting resistors in series – instead of the overall value decreasing, it increases; the capacitance is larger and the oscillators run slower.

Having put the 4067 and the five DPDT switches in place, I then had to connect the relevant input/outputs to 24 different resistors, in a chain (or ladder) like the original one inside the stylophone.  I suppose it would have been possible to calculate the exact resistances, but I had some time ago obtained a hundred 10k presets for about 7p each, for exactly this kind of situation, so decided to use those and tune it by ear.

This took some time, but at the end of it I had a substitute resistor chain for the SoftPot Stylophone and some methods of controlling it automatically.

*

It then occurred to me that with this arrangement, all this extravagance could only control one stylophone at a time, so I had a think about how to connect more instruments (and possibly instruments other than stylophones!).

The way to do it, it seemed to me, was to use the binary inputs to the 4067 as an output: any device could then be controlled, just by installing the 4067 and the five ‘major/minor’ switches in it – or perhaps some other suitable arrangement.

So I added two 5-pin DIN sockets as outputs, the five terminals being A, B, C, D and 0v.  Each of the four A, B, C, D outputs was buffered, using four of the six buffers in a 4050.  The 4050 is similar to its sister chip, the more well-known 4049; but whereas the 4049 inverts its outputs, the 4050 doesn’t.  This chip has even cleverer properties, which I will be using in a later project, but here I used it to ensure the binary outputs were of sufficient strength to make their way through a connecting cable and satisfactorily operate external circuitry.

I also added at this stage Clock In and Clock Out sockets, which would enable Bigfoot to set the tempo of a piece involving different instruments, or follow the tempo set elsewhere.  These two input/outputs passed through the remaining two buffers on the 4050.

The final thing was to add two more 5-pin DIN sockets, this time as inputs.  This would enable external circuitry to control the 4067s.  I had several more ideas of suitable external devices which could be used to do this, and I hope to be able to get around to making these quite soon.

The only other unusual component needed to get all this to work was a suitable master switch, to select the various external and internal inputs to the 4067s.  This had to have 4 poles – the A, B, C, D binary inputs – and 5 positions.  4 pole, 3 way rotary switches are easy to come by, but 4 pole, 4 or 5 way are not.  Fortunately, I was able to source a 4 pole, 5 way switch on eBay from a supplier in Hong Kong for just a couple of quid, so everything was in place.

With a circuit like this – just a handful of chips and a few external components – you either get a neatly laid out PCB or a rats’ nest of wiring.  I ended up with a rats’ nest of wiring . . . however, it worked, even when crammed into the case, with the addition of an extra section underneath the ‘big foot’ I had selected.

This picture shows the two binary input sockets on the left.  The 5 way switch is the knob on the front of the Bigfoot, just the right of centre in this picture.

Due to a certain amount of experimentation along the way, some changes of mind about the functions, and some difficulties in getting all the switches and sockets to fit, there were some extraneous holes which I had drilled in the case.  The plastic frogs are there to hide the holes.  I also added a square of velcro on the back where I could attach a battery holder, as I had done with a number of previous projects.

Bigfoot – automatic/remote stylophone control, Part 1

I’d made enough instruments for the time being, and it was time to construct some automatic controllers – sequencers, arpeggiators and the like – as an alternative to playing them by hand.

When I made the SoftPot Stylophone, I had added a socket which allowed external circuitry to replace the chain of resistors which govern the pitch of the instrument.  This project was to make a device which would be able to use this feature to operate the SoftPot Stylophone remotely, and this rather blurry photograph shows the result – Bigfoot:

I got the inspiration from several places: the arpeggiator and sequencer from Fun with Sea-Moss: http://milkcrate.com.au/_other/sea-moss/; the melodygenerator by Slacker described on the electro-music.com forum: http://electro-music.com/forum/viewtopic.php?t=27239&postorder=asc&start=50; and the Intro to Lunetta Synths at https://docs.google.com/document/edit?id=1V9qerry_PsXTZqt_UDx7C-wcuMe_6_gyy6M_MyAgQoA&pli=1,  All these sites are full of great ideas and practical examples.

The main chip used in the circuits described above is a 4051, which is basically a single-pole 8-way switch.  It’s usually depicted in circuits like this:

The way it works is like this: it’s an analogue switch, not a digital switch, meaning you can connect anything you like to the pole (pin 3, marked Z in the diagram) and the 8 switch input/outputs (on the right-hand side, marked Y0 – Y7).  It doesn’t have to be logic high or logic low (i.e. +v or 0v) , it can be any voltage, an audio signal, anything – just like a physical switch.  Any one of the 8 input/outputs can be connected – one at a time – to the pole, not by turning a physical switch, but by the logic high or logic low status of the 3 ‘Select’ inputs (pins 9, 10 and 11, marked S1 – S3).

You can have every combination of logic high and logic low on the three Select inputs, ranging from 0v on all of them, 0v on one of the three and +v on two of them, +v on two of them and 0v on one, or +v on all of them.   There are eight possible variations, starting with 0v on all of them, which you could represent as ’0 0 0′ or the binary equivalent of the number zero, to +v on all them, which could be represented as ’1 1 1′ or the binary equivalent of the number 7.

If you feed 0v to all three of the Select inputs, or ’0 0 0′, this is lowest possible binary number, so the lowest or first input/output is connected to the pole (Y0, pin 13); if you connect, say the one on pin 9 (S3) to +v and the other two to 0v, this would be the binary number ’1 0 0′, the equivalent of the number 4.  Because the sequence starts with ’0 0 0′ , or zero, feeding in ’1 0 0′  connects the 5th rather than 4th input/output to the pole (Y4, pin 1).  By connecting all the Select inputs to +v, or ’1 1 1′ (the number 7), the 8th input/output is connected to the pole (Y7, pin 4).

In the circuits I looked at, a common type of connection would be to have the pole connected to the part of an oscillator circuit that determines the pitch, and 8 input/outputs connected to different value resistors.  This would mean that a different resistance would be connected to the oscillator and a different pitch would be sounded when each of the 8 input/outputs was connected to the pole.

You could determine whether each of the Select inputs was a ’1′ or a ’0′  with  three 2 way switches, +v one way, 0v the other way, and change the notes by moving different switches up and down.  But this would be rather tedious.  By adding a circuit that automatically changed the ’1′s and ’0′s, you have a melody generator, arpeggiator or sequencer.

This was the kind of circuit I was after.

However, 8 notes was bit restricted.  Not restricted because there are 12 notes in one octave, though: I reasoned that you could make life easier for yourself by only allowing notes in a single scale – the ‘do’, ‘re’, ‘mi’ approach so succinctly captured in the Rodgers and Hammerstein song from The Sound of Music (‘Do a deer, a female deer/Re, a drop of golden sun’, etc.).  There are only 8 notes in a  ‘do’, ‘re’, ‘mi’ scale, including the next ‘doh’ up from the one you started from.  If you just use those, you’ll never get an ‘out of tune’ note in your arpeggio or sequence.

The proper name for the ‘do’, ‘re’, ‘mi’ system, by the way, is ‘tonic sol-fa’, and was invented here in East Anglia by Sarah Ann Glover of Norwich, who lived from 1785 to 1867.  This 1868 woodcut shows Sarah Ann teaching ‘do’, ‘re’, ‘mi’ to the musical children of Norfolk:

(Why this public domain picture  is held by Music Department of the Bibliothèque National de France is not adequately explained by the Wikipedia, where I found it.  I suppose the fame of ‘do’, ‘re’, ‘mi’ is international).

No, it was restricted instead because the SoftPot Stylophone has 12 ‘do’, ‘re’, ‘mi’ steps from the bottom of the keyboard to the top – and in any case could be made to produce notes outside the range of the built-in keyboard.

So I decided I needed 16 steps (2 octaves, including ‘do’ two octaves up from the start), and found a chip, the 4067, to do the job.  The 4067 is a single-pole switch like the 4051, but with 16 switches instead of 8.  The only way it differs in operation from the 4051 is that it requires 4 Select inputs in order to go all the way from ’0 0 0 0′ (zero, meaning the first input/output is connected) to ’1 1 1 1′ (15, meaning the 16th input/output is connected).

The 4067 usually appears in circuits like this:

It’s very similar to the 4051: there’s a Pole (pin 1, marked Z); 16, instead of 8, input/outputs (right-hand side, marked Y0 – Y15); and 4, instead of 3, Select inputs (pins 10. 11, 13 and 14, marked S0 – S3).

I also decided to make things slightly more complicated by considering alternative scales.  If you follow the ‘do’, ‘re’, ‘mi’ scale of the Rogers and Hammerstein song, this is a major scale.  If, on the other hand, you wanted to play, for example, a minor scale, you would find that ‘mi’, sometimes ‘la’ and sometimes ‘ti’ have to be changed to be a semitone lower.  And occasionally you might feel like making ‘re’ and ‘so’ lower as well.  (‘Do’ and ‘fa’ can be left alone!).

I’ll explain in a minute exactly what scales I had in mind when doing this, and where I got the idea from, but adding the ability to sharpen or flatten certain notes of  the scale meant that I needed 25 notes instead of 15, so the 4067 was wired up like this:

The notes depicted are the notes that would be used in the key of A.  Since the SoftPot Stylophone has a tuning control (in fact two tuning controls!) on it, it can be made to play in any key, not just A; the circuit here doesn’t need to be changed, only the tuning on the SoftPot Stylophone itself.

Each of the 16 outputs of the 4067 is connected to a resistor in a chain.  The top of the chain is connected to the tip of a 3 way (‘stereo’) 3.5mm socket; the bottom of the chain is connected the ring, and the sleeve is connected to pin 1 of the 4067 – the pole of the 16-way switch.  When plugged in, it takes the place of the Stylophone’s own resistor chain.

Note that switches allow you to choose between 1) major and minor 2nd (‘re’); 2) major and minor 3rd (‘mi’); 3) major and minor 5th (‘so’); 4) major and minor 6th (‘la’); and 5) major and minor 7th (‘ti’), as you see fit.  C1/C#1 and C2/C#2, D#1/E1 and D#2/E2 etc. use the same switch, so there are 5 of these switches, not 10.

The reason I chose to do it this way is because of an extremely interesting article – series of articles, actually – which I read on The Tonal Centre website, written by Andrew Milne.  I’m not in the slightest bit concerned that the theory described there is ‘unconventional and some of the concepts . . . quite novel’, as it seems to me to make perfect sense, and presents a coherent view of scales and chords which I’ve found quite easy to understand, and useful to use.  Furthermore, Milne’s motives for writing the articles are ones with which I would hope none of my readers could disagree: ‘not for theory to be an intellectual straight-jacket which smothers spontaneity, but as a springboard for creativity and, even more importantly, as a foundation for exploration’.

Essentially, the articles do precisely as the author says in his introduction: ‘convince you that there is a lot more for the tonal composer to experiment with . . . than just the major and the minor scale.’

I can’t explain everything in the articles because a) there is too much, and b) I don’t understand it all; but essentially, the points I want to draw attention to are these:

1. What constitutes a useful and versatile scale?

A scale should constitute ‘a unified collection of notes – a selection which is in some sense complete and to which any addition is heard to be extraneous’.

2.  What makes a scale useful as a melodic resource?

A scale should be ‘reasonably smooth and even, without sudden gaps which sound as if a note has been omitted, or sudden concentrations of notes which sound as if an extraneous note has been added’.

3.  What makes a scale useful as a harmonic resource?

Because three-note major and minor chords are the basis of our kind of western music (like C-E-G and C-Eb-G), a scale shouldn’t have any notes which aren’t part of a three-note major or minor chord.

Of all possible scales there are only five prime scales which satisfy Milne’s criteria, as above. (These are the main criteria, but see the full article for a couple of others).

All of these scales contain, as it happens, seven notes, and these are clearly the most useful and versatile scales to use.  This was good news for me, as the Bigfoot would inevitably use 7-note scales.

There are 8 different scales altogether in Milne’s system, not just 5, because of  differences between major and minor, and so on, and these 8 variations of the 5 ‘prime scales’ (in the key of C) are:

1.  The diatonic scale, major and ‘aeolian’:

C-D-E-F-G-A-B

C-D-Eb-F-G-Ab-Bb

2.  The harmonic minor scale:

C-D-Eb-F-G-Ab-B

3. The harmonic major scale:

C-D-E-F-G-Ab-B

4.  The melodic scale, major and minor:

C-D-E-F-G-Ab-Bb

C-D-Eb-F-G-A-B

5.  The double harmonic scale, major and minor:

C-Db-E-F-G-Ab-B

C-D-Eb-F#-G-Ab-B

So, there are 8 different scales you can use, which all allow you to make interesting melodies and chords.  Each one has its own ‘character’, and some are much more commonly used than others.

This series of articles seemed to me when I came across it to be an extremely good guide to useful scales, and could be a help to anyone: you could use the description above to work out what scale or scales you commonly use, and then try writing a composition or improvising a solo using a completely different one.  There’s bound to be at least one you’ve never thought of using before!

Bigfoot allows the 2nd, 3rd, 5th, 6th and 7th (D, E, G, A, B in the above examples) to be individually adjusted, so arpeggios and sequences in all – well, almost all! – of these scales are possible.  The double harmonic minor isn’t possible because Bigfoot can’t produce F# and G at the same time; but 7 out of 8 isn’t bad!

So, 16 individual intervals are available  from the Bigfoot, spread over two octaves; the tonic is repeated 3 times, at 3 octaves; the 4th is repeated twice, at two different octaves; the other 10 notes are switchable between a ‘normal’ or ‘flattened’ version, which is semitone lower.

Hang on, that’s only 15 intervals . . . Well, since all 16 Select input combinations from ’0 0 0 0′ to ’1 1 1 1′ could be used to produce notes, there might in some circumstances be no way of stopping the Stylophone from sounding; so what I did was to start with ’0 0 0 1′ (the second output) and make that the lowest note, reserving ’0 0 0 0′ (the first output) for a rest where no note would sound.  I added a switch so that the first and second inputs could be connected together for those situations when this would be better.

I also added a START/STOP switch, which is what pin 15 of the 4067 does: if connected to +v it stops, and all the switches are disconnected, regardless of the state of the Select inputs; if pin 15 is connected to 0v the switches start to work.  (The 4051 also has this feature).

In practice, I actually installed a second 4067, with the two 4067′s being connected only at the 4 Select inputs (pins 10, 11, 13 and 14).  I wanted to have an LED indication of which switch was connected, and had to separate this function from the resistor chain that produced the notes.

So the pole pin of the second 4067 was connected to +9v via two 1k resistors [not one, as shown in the diagram], and each of the 16 outputs was connected to a green LED (matching the green case the circuit was built into).

In order to test the LEDs – and later to test the notes which were being produced – I needed some way of connecting exactly the right input/output to the pole of the switch, so I would know I was adjusting the right preset.  This meant feeding exactly the right combination of  +v (’1′s) and 0v (’0′s) to the Select inputs, to get exactly the right output.

I considered four 2-way switches, +v one way, 0v the other way, and changing the notes by moving different switches up and down, as I described before – but it turns out there is a device which does this job very simply, just like turning a rotary switch: a 4 bit binary (sometimes called hex) rotary encoder.  I wouldn’t say these are extremely easy to come by, but this is the one I got:  http://uk.mouser.com/ProductDetail/Alpha-Taiwan/RE2001F-40E2-20F-4B/?qs=yA6kp8fx8Y4fjZ7sDt2l6A%3d%3d.

(The above picture shows a typical rotary encoder made by Alpha Electronics.  RS online sell a couple, but looking at the product details, I don’t think the connections of the ‘Code 033′ version they sell is right.  There are lots of 2 bit encoders, and lots of encoders which are not binary or hex.  They won’t work – it has to be 4 bit binary with 16 positions, starting with ’0 0 0 0′ at position 1 and stepping through the binary numbers 1 – 15, ending up at ’1 1 1 1′ at position 16. These are referred to as ‘hex’ because the hexadecimal system has 16 numbers in it [usually written as '0 1 2 3 4 5 6 7 8 9 A B C D E F' - a more user-friendly way of depicting '0 0 0 0' to '1 1 1 1']).

I needed to use the encoder for another part of the circuit, which I’ll come to later, but for the time being its 4 outputs were connected directly to the 4 Select inputs, ‘A B C D’, of the 4067s.  Its other connection, ‘Common’, was connected to + volts.  To test it, I used  4 LEDs, and could see that turning it from position 1 to position 16, it automatically output the binary numbers in order from ’0 0 0 0′ to ’1 1 1 1′.

It’s worth mentioning an important point, to avoid later confusion, which is that ‘D’ is actually the bit on the left in a binary number such as ’1 1 0 0′, and ‘A’ is the bit on the right.  You might sometimes see ‘D’ referred to as the ‘Most Significant Bit’ (or ‘MSB’) and ‘A’ as the ‘Least Significant Bit’ (‘LSB’).  That means the number sequence goes like this:

D  C  B  A

0  0  0  0

0  0  0  1

0  0  1  0

0  0  1  1

0  1  0  0

etc.

The other thing about rotary encoders is that they don’t usually have a stop, they just go round and round.  This is fairly useless if you need to know where ’1′ is, or where ’16′ is, and this is the main reason why I decided to incorporate the LEDs as a visual indication.  The other reason is that sequencers and so forth really ought to have flashing lights on them.

The rotary encoder is the knob on the right-hand side of the Bigfoot, just to the right of centre in this picture:

I glued the LEDs in place on the top and connected up the rotary switch.  Sure enough, with each turn the LEDs lit up one by one, one at a time, and now it was possible to tell which was position 1, which was position 2, etc.

Not only that, with the lack of a stop at 1 and 16 – which you would expect with a normal rotary switch – if nothing else I had Method 1 of controlling the Stylophone remotely: a manual method of arpeggiation by spinning the encoder backwards and forwards! . . .

. . . Entertaining, but not the automatic method I was looking for, however, so I moved on to Part 2 of the construction.

BigBoy BeatBox

You’ve got a Stylophone, you’ve got a Stylophone Beatbox – but don’t you sometimes wish the two could be combined into one instrument? . . .

Well, now that wish has become reality, with the ‘BigBoy BeatBox’: two great Stylophone products in one!

As the picture suggests, the BigBoy BeatBox is, in fact, two great Stylophone products literally glued and bolted together, with some of their internal circuitry combined.  The way it was created was like this:

1.  The Stylophone

The Stylophone half of the instrument is, in fact, a recreation of the original ‘Big Boy’ – a regular Stylophone S1 inside a Beatbox case.  As mentioned in an edit to the original post here, I managed to inflict terminal damage on the ‘Big Boy’ by reckless experimentation.  I normally do this before finishing an instrument, this time I contrived to do it afterwards . . . so the first thing I had to do was remove and replace the electronics with a new donor Stylophone I had lying around.

The actual process closely followed the construction of the original, but was made easier because of the sockets and wiring still remaining in the Beatbox case.  First of all, the end had to be sawn off the Stylophone circuit board, which is too long to fit in a Beatbox case; then the lowest 12 notes of the keyboard were connected to the 12 outside pads of the round Beatbox keyboard.  Fortunately, the wires attached to the Beatbox keyboard remained in the case, and just needed connecting to the appropriate Stylophone keys.  The Beatbox’s amp circuit board was taken out, but the Stylophone’s was kept and connected to the Beatbox’s speaker.  A power socket was connected to the Beatbox’s on/off switch, and the Stylophone’s on/off and vibrato switch circuit board disconnected.

I decided to replace the ‘Big Boy’s troublesome original 3-way octave switch with a simple  pitch potentiometer.  I used a 100k for  coarse tuning, in series with a 10k for fine tuning and a 100k variable preset to fix the highest pitch available.  Previous experimentation with Stylophones had taught me they have no objection to going down to very low pitches, but they cease to function – usually temporarily – if the pitch is taken up too high: on resetting, when this happens – by switching the power off and on – sometimes they will begin to work again, sometimes they won’t.

That’s what I did to the original ‘Big Boy’, and there’s no cure apart from throwing the circuit board away and starting again.  The likelihood of this happening is increased because I don’t just replace the tuning potentiometer pin-for-pin – the range of voltages available between the two pins the Stylophone uses isn’t wide enough for very large pitch variations, so I use only one of the pins that the original tuning potentiometer was connected to – the left-hand one – but connect the other one to +v.

The two new pitch controls were fixed to the front (the rounded end) of the Beatbox case, as was a replacement 10k log volume control.  The problem with the Stylophone’s original volume control was not that it wouldn’t work perfectly well, but that it would have had to be on the side of the case which I was intending to fix permanently to  the other Beatbox.

The original ‘Big Boy’ had no vibrato, but I decided the recreation should have a variable control, as fitted to the ‘Alien’, my first Stylophone modification project.  All this involved was connecting a 1M potentiometer instead of an on/of switch between the two vibrato connections next to the power connection on the main Stylophone circuit board.

Apart from an output socket and a switch to cut out the internal speaker, that half of the BigBoy BeatBox was done.

2.  The Beatbox

The other half of the instrument was a plain beatbox, with very little in the way of modifications.

The first thing I did to it was to replace the tuning potentiometer with a larger one of 100k (a direct pin-for-pin replacement this time), allowing for considerable slowing down and lowering of the pitch of the drums and other sounds.

I also followed an excellent example in this YouTube video: http://www.youtube.com/watch?v=xXdelnxXF7A to add buttons in parallel with the ‘Record’ and ‘Play’ pads normally operated by the stylus.  The trouble with the stylus-operated method is the delay in time between activating ‘Record’ with the stylus, and then using the same stylus to stop recording and begin playing the pattern you want to  be looped, as the loop begins the moment ‘Rec’ is selected.  With a small normally-open tactile switch as an alternative method of beginning and ending the recording period, you can be much more accurate as regards timing.

It’s worth mentioning at this point that, like the original Stylophone itself, the Beatbox comes in more than one variety, as far as circuitry is concerned.  I noticed two significant variations between the Beatbox I used in my ‘test to destruction’ phase, and the one that eventually found its way into the finished instrument: in one case, there was a tap from the battery compartment at 3v, which fed into the circuit (via the 3-way tone control) as well as the full 4.5v; and the layout of the circuit board was different.  As it happens, spots on the board marked ‘Rec’ and ‘Play’ were easily accessed in one case – the test unit – but not in the other – the one I was eventually to use.

In my experience, the Beatbox is a very delicate circuit, and it doesn’t take much to do something to it that will cause Record or Play to malfunction, or the output quality of the sound samples to degrade; so, proceed with caution, I’d say.

A third new button I added to this unit was ‘Reset’.  The Beatbox’s method of erasing an old loop and re-recording a new one is to switch the power off and on again.  The original power switch had to be removed as it was on the side of the case which was going to be fixed to the other Beatbox case – and there would, anyway, be a single power switch for both units: so the third button is a direct replacement for the Beatbox power switch, but now a normally-on, push-to-break, supplying power to the Beatbox side only.  To reset and re-record just takes a quick press and release of this button.

Finally, a new 10k log output volume pot was fixed to the front of the unit.

3  Joining the two halves

Superglue and two bolts was all that was required to physically join the two Beatbox cases, plus a couple of holes through which wires could pass from one side to the other.

The easiest way to connect the power seemed to be to detach the +v and 0v wires from the battery compartment on the Beatbox side and attach these to the original ‘Big Boy’ Stylophone side, which had a power input socket.

In the end I decided that whereas the battery compartment of the Big Boy Stylophone had to be removed – there was no room for batteries as well as the Stylophone circuitry – the one in the Beatbox could be used.  So I wired in a power cable which ran out of the back and was just long enough to reach the power socket in the other half.  In this way the instrument could be powered from an external source, or from internal batteries, and there was no need for a switch to change from one to the other.

The two 10k volume controls were taken to two individual tone controls.  I just wanted something fairly rudimentary, so I used a circuit from http://www.muzique.com/lab/swtc.htm called the ‘Stupidly Wonderful Tone Control’.  The component values I used were quite different – and I have no idea why – but the format of the circuit was more or less the same, and gave a little bit of variation to the tone.

After the tone controls, the two outputs were joined with 10k resistors to the original Beatbox volume control, and then the ‘Big Boy’ Stylophone amp circuit board.  This meant that the sound from both units was going to the ‘Big Boy’ half , and the volume and tone of each unit could be independently varied.  The input to the Beatbox amplifier was disconnected, and the speaker removed.

Now I had an instrument in a single conjoined case, with a single power supply and output through a single speaker or output socket.  There were two styluses and two keyboards, and – as I had hoped, but not expected with any confidence – both styluses work on both keyboards!  This means that both units can be played with a stylus in each hand, and quicker and more rhythmic patterns can be played.  I extended the wires to the styluses slightly to make sure they could reach right across both keyboards.

When using two styluses on the ‘Stylophone’ side of the unit, the ‘Beatbox’ side needs to be set to ‘Play’, otherwise that stylus will only work for a very short period and then not sound any more.  I haven’t timed the ‘very short period’, which might give a clue, but this is probably to do with the circuitry which regulates the maximum of 8 seconds (at ‘normal’ tempo) for which the Beatbox can record.

The complete circuit looks something like this:

The following pictures show the inside of the instrument shortly before it was finished:

This is the ‘Big Boy’ stylophone half.

1 = speaker cutout switch

2 = socket for external 4.5v power source

3 = 3.5mm sound output socket

4 = Stylophone S1 circuit board with permanently soldered connections to first 12 keys

5 = socket for extra stylus, remaining from original ‘Big Boy’ design – not really needed now

6 = fine tune pitch control

7 = variable preset to prevent the Stylophone’s highest note from being too high and causing the circuit to malfunction

8 = coarse pitch control

9 = Stylophone volume control

This the Beatbox half.

1 = the original Beatbox output and ‘mp3′ input sockets, no longer used

2 = Stylophone tone control

3 = Beatbox tone control

4 = Original Beatbox volume control, now master volume

5 = ‘Reset’, push to break switch

6 = wires going to ‘Play’ and ‘Record’ push to make switches mounted on top surface of Beatbox

7 = original Beatbox tempo switch, still in-circuit, but no longer used

8 = Beatbox pitch control

9 = Beatbox volume control

The features visible on the outside were these:

Before finishing I gave the speaker grilles a coat of blackboard paint.  The reason I used blackboard paint was a) it was the only black paint I found in my garage not in a spray can, and 2) it gives a pleasing matt finish, but is more durable than water-based matt emulsion.

The rear of the instrument was sprayed black and the holes masked with painted material.

Stylophones 3 – The Stylophone 350S, Part 1

A series on the Stylophone can only reach a climax with the mighty 350S!

The question of why the original Stylophone sold in its millions and became a world-wide success story, and the 350S didn’t, has long been debated.

According to http://stylophonica.com (‘The official home of the Stylophone’), it was ‘too costly, and lost the key uniqueness of the Stylophone itself, which was its small size and mass-market appeal’ – but it certainly wasn’t through of a lack of features.

You may be familiar with the Stylophone, but not the 350S: if so, then to start with, a run-down of its capabilities is required:

First of all, it’s certainly true to say that it’s much larger than the regular Stylophone – which is, after all, about the size of an inch-and-a-half thick postcard.  Here’s my 350S together with the regular-sized ‘New Sound’ Stylophone with which it shares many of its design cues:

The 350S is a souped-up Stylophone in every way: instead of the Stylophone’s 20 notes – an octave and a half – the 350S has 44.  That’s three and a half octaves, and you can see in the picture the difference in length between the two keyboards.

Not only that, the 350S has eight voices, as opposed to one (or even the S1′s three), and some of these are themselves in different octaves.

The voices are designated ‘woodwind’, ‘brass’ and ‘strings’.  In these days of sample-based synths, none of these sound terribly much like what they say they are, but they have the general qualities of these instruments – and, despite what you may read, one or two of them are quite like the distinctive tone of the regular Stylophone that we all know and love!

These voices are:

four ‘Woodwind’ voices pitched at four different octaves, and described (like organ stops) as 16′, 8′, 4′ and 2′;

two ‘Brass’ voices at 16′ and 8′; and

two ‘Strings’ voices at 4′ and 2′.

Because these voices are pitched at different octaves, from 2′ to 16′, in all no less than six and half octaves are available from the bottom of the keyboard to the top.  This is almost as large as a ‘professional’ 88-note synthesizer keyboard.  Up to two of the voices can be combined at any time, one each of the four octaves.

As well as this wide range of voices, the 350S has a variety of built-in effects.  Like the regular Stylophone, one of these is Vibrato – and two speeds are available, rather than one.

There is also a two-speed ‘Decay’ facility: as well as the usual Stylophone ability to hold a note as long as the stylus is in contact with the keyboard, when the Long or Short (actually, ‘short’ or ‘very short’) Decay button is pressed, the note will fade out while the stylus is still in contact.  According to the nicely-produced, LP-sized User’s Guide that comes with it, this enables the player to obtain ‘a percussive effect rather like  piano.’

However, as can be seen from the above picture, this is only the beginning of the 350S’s abilities.

The fast or slow ‘Reiteration’ button (second from the left) can be used to imitate the sound of a banjo or mandolin, and the 350S even has a second stylus which is used to produce these effects.

Normally, whichever stylus is being used, the ‘regular’ or ‘reiteration’, it’s held in the right hand; but it’s possible to play two notes at once in reiteration mode by using the reiteration stylus with the right hand, and playing lower notes with the regular stylus in the left hand.  It doesn’t work the other way round, and it doesn’t work in ‘normal’ mode, i.e. without either the fast or slow reiteration switch pressed.

The white tuning control can also be seen in the above picture – handily placed on the front of the instrument, unlike its counterpart in the regular Stylophone, which is always hidden underneath.

The most unusual effect, though, has got to be this:

Above the volume control is the 350S’s secret weapon – the ‘Photo Control’.  This device, operated with the player’s left hand while the stylus is wielded in the right, can be set to control the volume, amount of vibrato or low-pass filter cut-off point – acting as a ‘waa waa’.

On the side of the 350S, next to the Photo Control, are three 1/4″ mono jack sockets.:

While one of these is ‘sound in’ and another ‘sound out’, the middle one is a socket for a foot pedal that replaces the Photo Control – either because the player would prefer to control volume, vibrato or waa with their foot, or because the ambient light level is too low for the Photo Control to be effective.  A 50k – 100k potentiometer does the job, according to the User Guide.

My experience of light-dependent controls like this – and I’ve made a number of them – is that they are really only fully effective when quite a bright light is shining on them, which is not always the situation when you sit down to play.

Unsurprisingly, this magnificent machine requires a fair amount of juice, so it’s powered by not one, but two weighty PP9 batteries, connected in series to provide 18v of power to the 350S.  These are housed underneath the rear of the instrument:

The battery covers look as if they’re held in place by screws, but these aren’t really screws: they click into place when pushed, and just require a slight turn with a screwdriver or a thin object to loosen them.  (The User Guide suggests a coin, but in my experience modern coins are too thick to perform this function.  Maybe a 5p would do it).

This is the User Guide that came with the 350S:

http://homepage.ntlworld.com/andymurkin/Electronica/350sBooklet.pdf

Reliable information on when the 350S first came on the market, how many were sold, etc. seems hard to come by.  http://stylophonica.com says: ‘No more than a few thousand 350S’s were ever sold’; http://stylophone350s.com/ says ‘Dubreq, the manufacturer of the original Stylophone created and produced the Stylophone 350S beginning in 1971 . . . fewer than 3000 were ever produced’ and quotes a Ben Jarvis (son of Stylophone inventor Brian Jarvis and re-founder of Dübreq in 2003) estimate that only 200-300 working units are probably still available worldwide.

I’d be surprised if the numbers were quite this low, but they’re certainly not common, and those that appear on eBay in the UK frequently command in excess of £100, rarely less than £70. Stylophone350S.com in the States have access to a recently discovered cache of mint condition boxed examples, which are now on sale.  Their website tells the story of this amazing find.

The back of the 350S is removed by undoing 4 large screws in the corners and two very small screws under the front, and reveals two printed circuit boards: a thin, narrow one at the front containing the keyboard and the resistor chain – not discrete resistors, but what I’ve previously called ‘resistor modules’, since I can’t remember what the proper name for them is – and a large, rectangular one with everything else on it, including potentiometers, sockets and switches:

The circuit boards themselves come away quite easily: there are 4 screws, clearly visible in the above photograph, which hold the keyboard in place, and 6 similar ones for the larger board.  The volume control knob doesn’t need to be taken off – it fits through the hole surrounding it – but the plastic nuts on the three sockets need to be removed.

This is what the other sides of the boards look like:

Here we see the larger items across the middle of the board, from left to right: the three sockets, the volume control, the eight voice and effect switches, the pitch control and the on/off switch.  If I was an electronics expert, I could tell you what the rest of the components do; but I can’t.  I can only surmise that the round inductor next to the left-hand switch is to do with the waa circuit; the LDR (light-dependent resistor) to the left of that is the ‘Photo Control’.

The black ‘hood’ that partly surrounds the LDR was slightly damaged when I came to look at it, and it’s quite possible that I did this myself when I opened the case.  It was easily repairable with a spot of superglue, but watch out for this if you’re looking inside yours.

The ferrite core inductor is a Mullard FX2236.  In this close-up you can see that mine looks a bit broken.  I don’t know enough about these things to know if this means it isn’t working properly, but, while by no means common, they can be found – perhaps more easily in the UK than elsewhere – so I shall certainly consider replacing it.

According to the experts at www.stylophone.com, under the heading ‘VITAL INFORMATION WHEN BUYING A 350S… PLEASE READ CAREFULLY!’, one of the components you can see here – which they describe as the ‘Amp-ic’ – is highly prone to failure.

The related website, the Stylophone Information Centre at www.stylophone.fsnet.co.uk says: ‘The circuit board carries an IC which controls sound output, and this component (long since obsolete) is the single- most likely cause of the 350S to break down. If this happens . . . the unit will only be heard if played through a separate amplifier, if at all.’

The symptoms to look out for are: ‘when the stylus is applied to the keyboard, only a very faint sound is heard (if even audible at all), which fades away rapidly . . . Even with the volume control turned up to max, the sound will still be very low – then quickly fall away. The user will then be left with a ‘dead’ 350S.’

The chip in question is this one – the black one with six legs in the top left-hand corner of the picture:

It’s a Motorola MFC 6070, 1-watt power amplifier  - ‘designed primarily for low-cost audio amplifiers in phonograph, TV and radio applications’, according to the datasheet.

If you don’t know what a phonograph is, ask your grandad, he’ll remember them!  The use of this antiquated vocabulary confirms what is said above.  If you find the datasheet for this chip, it says ‘Device discontinued – consult factory’; if you try to buy one on the Net, you’ll mostly find specialist sites, dedicated to sourcing obsolete parts.

As a matter of fact, you can, at the time of writing, get one on eBay for about £20, but you aren’t going to want to do that: the problem doesn’t arise, apparently, just because 350S’s are now all old – it even used to happen to quite new ones.

Stylophone.com told me that ‘the original chips as fitted . . . were working very close to their breakdown point voltage-wise. Although theoretically all the chips supplied to them should have worked, Dübreq actually had to batch-test the chips to find those with an acceptable working voltage range, especially the maximum voltage’ (which is meant to be 20v). ‘We’ve seen some of these chips.’ they said, ‘ running extremely hot (basically too hot to touch) by simply switching the instrument on, before even playing a note.’

That’s not to say the MFC6070 was a particularly unusual part at the time – they were used all over the place, and even the venerable VCS3 synthesiser used one as a driver for its spring reverb circuit.  However, as the site offering VCS3 spares, http://www.synthi.com, says: ‘The Achilles heel of the VCS3/Synthi AKS are the now obsolete and ultra rare semiconductors that it uses’ . . .

This made me think twice about powering the 350S with a mains-powered adapter: the increased risk of overdoing the voltage and blowing the chip might not be worth it.  Dübreq themselves did apparently produce some 350S’s with an ‘adaptor socket factory-fitted’, but ‘this led to many of them blowing the chip.’

I’m not quite as worried as I was, however, as stylophone.com are now marketing a new module, the ‘Stylophone ACM’, which can be retro-fitted to an ailing 350S – or even to a working one, as a precautionary measure – to get round this problem altogether.

The circuitry inside this unit is not operating close to its limits, and makes it much safer to run the 350S from an 18v adapter.  (And if you buy a reconditioned 350S from stylophone.com, it will already have one of these in it).

As a matter of fact, this is not the only ‘obsolete’ component in the 350S.  Although the resistors, capacitors and transistors that fill the circuit board are not commonly used in new designs nowadays, they’re still readily obtainable; the round silver integrated circuit over on the right-hand side, just above the tuning control, isn’t.

It’s a General Instruments AY-1-5051, and what it does is frequency division (presumably for the 350S’s different octaves) – the kind of thing modern CMOS 4000-series chips do with the greatest of ease.  There’s a description on this website: http://www.divdev.fsnet.co.uk/repair2a.htm of how one might make such a replacement (using the example of a 1960s Elka electronic organ).  All I can say it, it looks feasible in theory, but not something I’d want  to be faced with in practice – let’s hope this isn’t a part which is going to fail!

Returning to my 350s, it looked badly in need of a clean up.  There was a lot of dust inside it, and over the years the keyboard had got very dirty:

The switches sounded OK – no crackling or intermittent operation, so I left those, and just cleaned the circuit board and keyboard.  The keyboard in particular needed attention from, in order, a soft brush, switch cleaner, WD40 and Brasso.   This seemed to do the trick, and it began to look shiny again.

I cleaned everything, including the switch rockers, the case and the tips of the styluses, and put it back together again.  It now looked much better, and sounded clearly and reliably on every note.

In my next post, I’ll take a longer look under the bonnet of the 350S and see what there is to see.

The UFO and the Shuttlecraft

The UFO is a simple device for controlling instruments with light-dependent resistor (LDR) controls, for example the Opto-Theremin described in an earlier post.

It started life as one of those battery-operated lights where you push on the top to switch it on and off:

I painted it silver, and added bits to make it more flying saucer-like, some LEDs that change colour slowly, and a 5-LED goose-neck lamp that I found in a local Poundshop.

The colour-change LEDs have no important function, but the brightness of the goose-neck lamp can be controlled with a potentiometer, and can thus be pointed at an LDR and used to vary – in the case of the Opto-Theremin – volume, pitch or filter cut-off frequency.  Here you can see the lamp and the potentiometer: I didn’t attach a knob as I couldn’t find one that looked more UFO-like than the knurled shaft:

The goose-neck lamp is meant to operate from a computer USB port, so plugs into a USB socket, with only pins 1 and 4 (5v and 0v) connected.  The battery holder in the lamp is designed for 6v, and therefore had space for 4 AA batteries, but I mostly use 9v, so I adapted it to take a PP3.  I restricted the potentiometer from putting the full 9v through the LEDs, in case it was too much for them.

The maximum voltage allowed for the color-change LEDs was 4.5v; there are 4 of them, so I connected two in series on one side of the dome, in parallel with two in  series on the other side.

I also added some extra 3.5mm mono sockets, as can be seen in the picture, as this is a system I use for distributing power.  When the Opto-Theremin is used in conjunction with the UFO, it can receive its power from there, rather than from a separate battery.

This picture shows the two being used together:

These pictures illustrates the soothing effect of the constantly changing colors:

The ‘Shuttlecraft’ isn’t really an invention of my own: in fact, it’s just a multi-LED lamp on a headband, as worn by cyclists.  It appears here only because it’s an aid to playing the Opto-Theremin.  Because light levels are often too low to get the maximum variation in parameters controlled by LDRs, it can be useful to have extra light to hand: but when your hands are occupied playing the instrument, the next best place is on your forehead.

Although the UFO and the Shuttlecraft were created with the Opto-Theremin in mind, they could be used with any instrument (or effect) that uses an LDR – for example, my first Stylophone mod, the ‘Alien’, or the Stylophone 350S.

Stylophones 2 – Variations

As I said in my first post on the Stylophone, there have been a number of variations in Stylophone design over the years, so I thought I would illustrate some of these from the examples in my collection.

The earliest Stylophone – in the days before Rolf Harris adorned the box – looked like this:

This early variation is distinguished by the non-playable black sections of the keyboard.  There were three types, distinguishable only by the body colour – the black one illustrated was the ‘standard’, but there was also a white one, the ‘treble’, and also a ‘bass’.  I don’t know what colour it was: I’ve only ever seen it in pictures, and it looks like a reddish-brown to me, but I’m colour-blind, so an unreliable witness . . . I hope in time to get hold of one, and somebody will tell me if it is indeed brown!

Here is the booklet that you see pictured above:

Read it here.

The black sections of the original keyboard had been a feature of Brian Jarvis’s prototype – which you can read about here: http://www.stylophone.com/Prototype.html - but the next generation of Stylophones dispensed with them.

There were still three types – the black ‘standard’, the white ‘treble’ and the (presumed) brown ‘bass’, and they looked like this:

Note the identical case design to the original, but the keyboard is now completely silver.  (Ignore the switches on the sides of these instruments – they’re a speaker cut-out modification I made to them many years ago).

The circuits in all these early Stylophones were quite similar, although not identical.  The instrument was subject to constant development, and there are versions with all discrete components, including the resistors which determine the pitch of the notes, and versions with different types of resistor modules – rows of resistors in a single unit.

It’s easy to peer into the inside of the original and ’2nd generation’ Stylophones: the back is designed to be easily removable, in order to change the battery, and the component side of the circuit board is visible.  This is the inside of an original version (the one with the black sections on the keyboard):

As you can see, in the middle, just above the piece of foam rubber which keeps the battery in place, there is a row of resistors connected to the keyboard, which determine the pitch.  This arrangement continued with the 2nd generation Stylophones.  This is a view inside the black ‘standard’ version pictured above:

(Ignore the large resistor at the back right-hand side – this is attached to the speaker cut-out mod).

At some time during the production of the 2nd generation Stylophone, resistor ‘modules’ came into use.  This picture of the white version pictured above, shows two orange-coloured blocks in place of the row of separate resistors:

Other slight changes were made to the component layout, and the style of the switches in the bottom right-hand corner is different.  (Once again, ignore the non-original large resistor next to the speaker).

This would be a typical version of the circuit from this period:

A slightly different one is illustrated here:

http://www.waitingforfriday.com/index.php/Reverse_Engineering_the_Stylophone.

Also at some point during production of the 2nd Generation Stylophone, there was a major change on the outside.  The shape and colours didn’t alter, but the guide to the notes, printed on the white background piece stuck around the keyboard and switches changed from showing notes (‘A’, ‘A#/Bb’, ‘B’, etc.) to numbers (’1′, ’1 1/2′, ’2′, etc.):

(These are the same two black and white Stylophones shown above).

This was the beginning of the famous Stylophone song-teaching method, which continues until this day.  Whereas the songs you learned from the original booklet were shown with notes, like this:

songs were now shown with numbers, like this:

Edit: However, there is a photograph of an object from the collection of the Museum of Design in Plastic at http://www.modip.ac.uk/artefact/aibdc–002025 which shows an original issue Stylophone (the one with the black sections on the keyboard) in its packaging: and one of the items included is an overlay for the keyboard surround.  First of all, this is white lettering on a black background, rather than black lettering on a white background; secondly, it uses numbers, not letters for the notes.

The third distinctly different type of early Stylophone was the ‘New Sound’, which came out in around 1975.  The sound was ‘new’ because instead of the transistor in the original, the oscillator used a 555 integrated circuit.  Mine came in a box featuring Rolf Harris:

This Stylophone featured, for the first time,a volume control, which can be seen on the left of the front panel, just above the on/off and vibrato switches.

The circuit for the ‘New Sound’ version looked like this:

This view of the inside of the ‘New Sound’ shows the black, rectangular 555 chip just above the centre of the circuit board:

The Booklet that came with the ‘New Sound’ Stylophone was more extravagant than the original – although it was only printed in black and white, it was 16 pages long and the pages were about twice the size:

Read it here.

All of these early series of Stylophones offer opportunities for modification and circuit-bending: the electronics aren’t complex, circuit diagrams are often available, and the components themselves are large and readily accessible.

I haven’t worked on these much, but my SoftPot Stylophone is a modified ’2nd Generation’ treble version.  The ‘Hedgehog’ uses a Stylophone ‘New Sound’.

Production of the original Stylophones ceased in 1980 and the manufacturer, Dübreq, moved on to other things (‘Top Trumps’ playing cards!), but in 2006 the design was revived.

The new ‘Stylophone S1′ had different electronics inside, but a more or less identical case design.  Only by looking carefully can you see the tell-tale signs: the extra socket on the side – an ‘mp3′ input – a volume control on the right-hand side, and a three-way tone switch on the front, none of which is present on any of the 1960s and 1970s Stylophones:

Several colour variations – sometimes referred to as ‘Special editions’ were produced.  These were all-black (‘ebony’), silver and white:

Unlike the earlier Stylophones, the colour isn’t an indication of different sounds – as all S1′s have 3 tones, there was no longer any need to make three different types.  They’re all identical on the inside – although I did have the impression that a little more care was given to the assembly of the Special editions, compared to the standard version.

In Asia an even more completely black version, the ‘Stylophone Studio’ was marketed:

They’re very uncommon in Europe and I’ve never seen one.

This is the booklet that came with the Stylophone S1:

Read it here.

Another rare variation of the S1 is the so-called ‘Raconteurs Tour edition’ – a special version made to be sold as part of the merchandising connected with The Raconteurs,  a band formed by Jack White after the dissolution of the White Stripes.

Both the colouring of the instrument and the package design were unique, with a distinctive black and gold colouring:

In addition, the contents of the booklet were customised for the band:

Read the Raconteurs booklet here.

According to a concert-goer (at http://cousinsvinyl.com/2008/dude-check-out-the-merch-raconteurs-and-black-lips-play-the-fillmore-detroit/): ‘I was very curious to see what a Stylophone was.  The box read “The Original Pocket Electronic Organ”.  My friend said, “Dude, you’re going to want to get one of these,” as he opened the box.  It was $40, but worth the money: it’s a working instrument with the Raconteurs logo on it.  The next day, it was worth $200 online.’  Mine was a lucky find on eBay, but these can be very expensive when you come across them, usually more than the original $40 price tag.

As an added attraction for those who bought the Stylophone at Raconteurs’ gigs, the band held a competition, inviting fans to submit Stylophone versions of their songs.  The competition was announced on the band’s website:

 The video made by the winner, Zach Herrmann, can be seen on YouTube at http://www.youtube.com/watch?v=2eK17MqWIo0.

The circuitry of the S1 is very different from the earlier Stylophones, being based on a tiny digital chip which you can’t even see as it’s covered in a blob of protective wax.  It has a separate amplifier circuit board.  It also runs on 4.5v, not 9v, so instead of a PP3 battery it takes three AA batteries.  These are not inserted by removing the back of the instrument like the earlier ones, but are held in a battery compartment accessed from outside.  For this reason, the S1 is glued shut and getting to the inside of it for the purpose of modification (or troubleshooting) is not a simple matter.

The only picture of the inside of an S1 I seem to have to hand is this one, which has points marked on it for a ‘feedback’ bend: but it clearly shows the components which are visible on the main circuit board – i.e. not very many! – and the amp board in the background:

The chip which does all the work is under the black blob; the resistors are tiny surface-mounted (SMD) type.

For this reason, modifications and circuit-bending opportunities are a little more limited than with the early series of Stylophones.  Elsewhere in the blog are one or two examples of my efforts: The ‘Alien’ was my first modification project; the ‘Gemini’ uses two S1 boards in a single case.

And finally, a word about the smallest ever Stylophone, the Stylophone mini:

This one really is miniature!  Measuring a mere 8cm x 4.5cm, this is an official Dübreq/re:creation product, and is a perfect reproduction of the regular Stylophone.  Powered by 3 AAA batteries, it has a working stylus and the full complement of 20 notes.  The only thing it lacks is the Stylophone’s traditional Vibrato.

Here is a Stylophone mini with a regular Stylophone S1:

Inside, there seems to be very little indeed!:

It looks as though the keyboard is connected to a small piezo element acting a sounder, with very little in between!  I didn’t take the circuit board out on this occasion to look, but I suspect, like the S1, the chip which operates the Stylophone mini is very small and surface-mounted on the other side.  It certainly looks as though modification and bending possibilities are limited.

That’s an overview of the mini and regular Stylophones; my next post on the topic will deal with the amazing machine often described as the Stylophone’s ‘big brother’, the 44-note, 8-voice Stylophone 350S:

Theremin 2 – The Opto-Theremin

In my previous post Theremin 1, I described an optical theremin circuit which was ultra-simple – and worked!  However, I described how once I had finished it and played with it for a while, I realised there were some slight problems in being able to play it effectively; and also began to wonder how I could make it sound more interesting and varied.

This post describes what I did to finish it off.

The solution I came up with to make it sound more interesting was based on something I’d seen on a couple of Lunetta-related websites: a 4040 divider outputting several octaves at once, which could be individually selected and mixed together to create the final waveform.

This was the original optical theremin circuit:

I reduced the value of the timing capacitor in the original circuit from 22n to 15n to make the basic pitch a little higher and connected the 4040 between the output of the oscillator – pin 2 of the 4049 – and LDR2.  Together with the direct output from the oscillator, I used the first 7 outputs from the 4040, covering 8 octaves overall – a rather excessive range, but the pitch variation obtained using LDR1 was very wide, and I figured the highest and lowest notes might be needed at some point – contributing to the tone, even if not the main pitch.

The 8 different octave outputs were connected together via SPST switches, and various combinations of octaves did produce a surprising variety of tones.

Just one more thing was needed, I thought, to maximise the availability of square waves at 8 octaves, and that was a filter.  I’ve written elsewhere in the blog about Ray Wilson’s simple-but-effective 741-based low-pass filter from the Music From Outer Space website, which I liked so much I built two of them: one of them as a stand-alone unit (Active low-pass filter); and the other one here in the Optical Theremin.  The only difference was that in this case I replaced the cut-off frequency control with another LDR (LDR3).

I discovered afterwards that I hadn’t used the most up-to-date version of the circuit, but adding the extra cut-off frequency fine adjustment and resonance control (the extra features of the revised version) would have been too much.  As it was, I already had 3 LDRs, and only two hands to operate them with.  I added switches to select between the two volume and filter LDRs and two potentiometers, which would enable one of them to remain constant while the other one was manually controlled.

The filter circuit went between the outputs from the 4040 and the input to the Volume LDR (LDR2).  And that was it for the design and construction of the circuit, which now looked like this:

The reason for passing the output of the 4049 oscillator (at pin 2) through two more stages (spare ones in the same 4049) was to give it the same power as the outputs from the 4040.  Using the signal from pin 2 as an output as well as feeding the input of the 4040 seemed to be too much for it, and it wouldn’t work.

Also – as I had done with a couple of recent instruments (for example, the Cracklephone) – I added a pair of banana sockets, so a larger external speaker could be used.  Not shown in the diagram is a switch to cut out the internal speaker when these sockets are in use; and a similar arrangement to the Cracklephone, two 3.5mm sockets where a small goose-neck microphone can be attached.  This is not part of the Theremin circuit, and is the nearest it gets to having a line-out.

This what the inside of the case looked like just before I put it back together.  I don’t recommend you try and stuff so much inside a small case, as I always seem to be doing.

As for the physical construction of the instrument this presented one or two problems.

The one of most interest concerns the LDRs.  Electrically, these seemed to work perfectly; but the problem with them is they’re so small, and it’s very difficult to make subtle changes to the amount of light falling on them with a large human hand.

This is a perennial problem with these small-sized LDRs, and what I decided I needed was something like a torch or spotlight has to widen the spread of light – a dish or reflector, which would effectively increase the area the hand would have to cover in order to restrict the light falling on the LDR.

Having thought of torches and spotlights, I reasoned that I could use MR16-type spotlights, remove the original bulbs, and fit the LDRs inside.  These MR16 (or, with a different fitting, GU10) bulbs can be expensive, but my local Tesco’s was selling a pack of 8 for £1.25, so I bought those, and got to work on them with a hammer.

I’m not saying I did a neat job – and I damaged a few of the 8 in the process! – but I ended up with three reflector housings for my three LDRs.  Knocking off the fittings was easy, but getting the bulb out was not: these things are evidently put together before the glass front is attached, as the bulb is considerably larger than the hole through which the electrical connections pass.  At first I tried knocking the bulb through into the body of the reflector and breaking it up – a procedure not dissimilar from that of the Egyptians, removing internal organs through tiny holes before the process of mummification – but that proved impossible to achieve without breaking the glass front; so in the end I just broke open a hole large enough for the bulb to come out of.

My original intention was to drill holes in the Theremin case large enough to insert the end of the reflectors, and superglue them in place, but in the end the hole would have to have been an inch (25mm) wide, and there wasn’t enough space inside the case to insert something this big.

So instead I mounted the reflectors on the surface.  I built up a sleeve for each one using 25mm inside diameter O-rings.  These were about 3mm deep, so 4 of them superglued one on top of another were enough to support a reflector.  More superglue ensured that the reflectors stayed in position, and that the LDRs, passed through a hole under each reflector,  were in the right place.

This picture of the back of the instrument shows an arrangement I’ve used in quite a few cases where there wasn’t enough room for the battery inside.  Close to the power socket I’ve stuck a square of velcro, and the battery holder sticks to this.

Having a battery stuck to the bottom wasn’t  a problem, as the instrument wouldn’t be resting on it: the  1/4″ Whitworth nut, glued over the hole through which the Beatbox tuning pot was accessed, is the attachment for a mini silver tripod which I managed to get hold of for 99p on eBay.  This would allow the theremin to be raised from the table top and set at the best playing angle.

I think the tripod gives it a futuristic look – or at least what was considered futuristic in about 1940 . . .

I adjusted the original tripod between these last two pictures to allow the feet to spread out a bit wider, as it was a little unstable, especially with the battery velcroed to the outside.  This just involved a little sawing and cutting and it now stands much better without being in danger of toppling over.

Theremin 1

Apart from the Stylophone, another instrument I’ve always been fascinated by is the Theremin, and I’ve always planned to make one.

I know the project I’m about to describe isn’t a proper theremin, but it has an oscillator, and pitch and volume are controlled by hand without touching it – which are amongst the essential features of  a theremin – so it seemed like a good place to start.

The Theremin is an electronic instrument named after the man who invented it in Russia in about 1920, Lev Sergeyevich Termen,  It’s called the Theremin, not the Termen, as Leon Theremin is the name by which he became known when he came to America in the late 20′s – probably a better representation of the family name, which is not Russian in origin, but French.

LeonTheremin, c.1924, public domain, from Wikipedia

Theremin disappeared back to Russia after just a few years, and did not reappear in the West for over 50 years (1989, by which time he was 92 years of age!), but left designs for his instrument which were manufactured under licence by RCA.

As I see it, there are two essential features of the theremin, pertaining to how it is played, and – more technically – how exactly the sounds it makes are produced.

First of all, as mentioned above, it is normally played without touching it: instead, pitch and volume are controlled by moving the hands nearer to or further from antennae – like radio aerials – on the instrument.  This is a pretty unique feature, and came about because Theremin invented it not while trying to make a musical instrument, but in the course of obtaining an audible response to scientific experiments he was conducting at the Physico-Technical Institute in Petrograd (St Petersburg).

The classic theremin features a vertical, radio-like antenna for pitch control (played by the right hand), and a horizontal loop (played by the left hand) to control the volume.

Secondly, the notes produced by the theremin do not come from the output of a single oscillator: instead, it has two oscillators, which run at radio frequencies (RF), and are too high to hear.  However, if there is a difference between the two frequencies, this produces a third tone, which is much lower and which you can hear.  When the player moves nearer to or further from the antenna, this alters the difference between the two high frequencies, and raises or lowers the audible tone.

(Another electronic instrument invented independently at the same time as the theremin, The ondes martenot, also uses the same principle of two high frequency oscillators producing a third, audible tone, but this is essentially a keyboard instrument  – although it also includes a ribbon controller as an alternative to the conventional keys.)

It is typical to hear melodies played on the theremin which feature a great deal of portamento – sliding from one note to the next.  This is almost inevitable, as there is nothing to guide the player to find the correct note, other than their ears.  Skilful performers can avoid doing this all the time, but it would be a terrible waste not to make a feature of it, since the theremin makes it easy.  It is typical, although not a necessity, for melodies to be quite high-pitched, and players will usually employ techniques that produce vibrato.

These typical theremin sounds – often described as ‘ethereal’ or ‘spooky’, and frequently found in horror or science fiction contexts – are a product of the two important design features described above.  Other instruments have been designed which mimic the sound of the theremin (including my recent SoftPot Stylophone), but unless the sounds are produced by the player’s proximity to the instrument, using the body’s natural capacitance to affect the pitch and volume of RF oscillators, it isn’t a proper theremin.

The Electro-Theremin or Tannerin, developed by Paul Tanner and featured on the Beach Boys’ Good Vibrations, was such a good imitation that people used to think it was a real theremin; but it was controlled by a slider.

According to Google, today would have been Robert Moog’s 78th birthday.  As you can see, they’ve illustrated this with a (functional!) version of his famous synthesizer.

Before inventing this, however, young Robert was heavily into theremins, studying the Leon Theremin-designed RCA model and inventing his own – proper – instrument.

It’s not entirely a coincidence that my SoftPot Stylophone – and the earlier Cybersynth – sound a bit like a theremin, but this was the first time I’d constructed an instrument meant to be played without touching it.

In fact, the particular instrument I built is an ‘optical’ theremin, with pitch and volume controlled by two light-dependent resistors, and the basic circuit was so simple, I had to try it:

This circuit came from Graf’s Encycopedia of Electronic Circuits (Vol. 5, I think), but seemed to me to have something of the look of a ‘Lunetta’ device about it. I’ve already digressed into the history of the theremin, so I’ll save discussing Lunettas for another time.  (If you don’t know what a Lunetta is, start here: https://docs.google.com/document/edit?id=1V9qerry_PsXTZqt_UDx7C-wcuMe_6_gyy6M_MyAgQoA&pli=1#heading=h.6a4696420d74 and here: http://electro-music.com/forum/index.php?f=160).

Anyway, I was sure I’d seen the 4049 chip on which the optical theremin is based being used in Lunetta circuits. I had all the parts to hand – including the 4049 which I had recently salvaged from a project board I’d put together so many years ago I’d forgotten what it was originally for; and even the transformer, which I’d recently bought for another project – and it fitted on a 1” square piece of veroboard tucked inside the case of one of the broken Stylophone Beatboxes I had acquired.

It was ultra-simple – and it worked!  However, once I had finished it and played with it for a while, I realised there were some slight problems in being able to play it effectively; and I also began to wonder how I could make it sound more interesting and varied.

I’ll describe my variations on the basic circuit in a follow-up post.  Information on Leon Theremin comes from Wikipedia and Theremin: Ether Music and Espionage by Albert Glinsky (Foreword by Robert Moog).  The Wikipedia article has a number of very useful links to websites on Leon Theremin and his wonderful instrument.

 

Bits & Pieces 4

I wasn’t expecting to add this post just yet, but I had a stroke of luck which has enabled me to complete the scheme for my mono mixing section which I started writing about when I described the Red Dragon the other day.

I bought a job lot of small SoundLab mixers off eBay which were said to be faulty returns.  I thought I might be able to salvage some parts from them, use bits of them in some way, or even repair them – but it turned out that several of them appeared to be in working order.

Two of them were straightforward 4 channel mono mixers – an updated version of the one I had used before, I presume – so these were immediately used for the left and right channel inputs to the mono mixer, as I described in the previous post.  Generally speaking, I wanted to have the lower tones to the left and higher tones to the right, so my ‘double bass’ stylophone was the first thing to be plugged into the left mixer; the treble stylophone and the SoftPot Stylophone in the right.

More interestingly, the two other working units were the G105C version with ‘microphone effects’ – a delay circuit which I guessed was probably based on a PT2399.  I opened up one of the dead ones, and found that this was the case.

The circuitry was very different from the original SoundLab mixer I’d acquired – all surface-mount components; everything, pots and sockets included, firmly fixed to a single circuit board – and I’m not sufficiently skilled or equipped to be able to repair something like that.  Not only was it not functioning, it seemed to short out the power when the on switch was pressed.

I sawed out the part of the circuit with the PT2399 on it, which didn’t short the power when used by itself, but didn’t do anything to the input sound either.  This section is permanently in circuit when the mixer is operating, so maybe that was why the original unit didn’t work.  In any event, I decided to put the broken ones away for another day, and concentrate on the ones that worked.  The case would find a use later on.

First of all an echo unit is a really useful thing to have – and 2 echo units with 4 inputs is a bonus!

My initial arrangement with these is to have the outputs connected to the new Left and Right Mixers.  The left echo unit is used for instrument input and the output is divided: one half of the output going directly to the Left Mixer, the other half going to the right echo unit, and from there to the Right Mixer.  As the delay time and feedback (number or length of repeats) are separately adjustable on the two units, some interesting stereo effects are possible.

Bits & Pieces 3

This Bits & Pieces post covers a few effects modules I’ve recently made.  These are:

1.  Active 3-way tone control – the ‘Tardis’

2.  Active tone control of unknown origin

3.  Active Low Pass Filter

1.  The active 3-way tone control is standard of its type, I imagine.  I don’t know where I found it – years ago I used to scour back numbers of electronics magazines in the local library and copy out interesting circuits.  No doubt it was one of those.  The circuit diagram looks like this:

and the finished article looks like this:

which is why it’s called ‘The Tardis’ – not because it’s a time-manipulation circuit, which would have been cleverer.

2.  The tone control of unknown origin is unlike anything I’ve seen before or since – or, rather, since the heart of it is quite a high-value inductor, it most resembles a variable bandpass filter – a wah circuit – but is evidently not intended to be swept up and down like a wah wah pedal.

Another of my finds in a very old electronics magazine, it was originally called a ‘Passive Tone Control’, but the reduction in the volume of the input signal was so drastic that I added an amplification stage before it to boost the level to something like the original, and it became an ‘Active Tone Control’.

Moving the single control from one extreme to the other varies the tone considerably, and it’s very useful with sounds rich in harmonics, like the various Stylophones in my collection.

This was another project after the Touch-Radio which I housed in one of the transparent jewellery cases I had recently acquired.

3. I’d heard nothing but praise for Ray Wilson’s simple 741-based low-pass filter – and indeed the whole ‘Wacky Electronic Noise-maker Thingy’ which it forms part of – so I decided to make one and try it out.  In fact, I made two, and I’m glad I did, because they’re great!  One of them went inside an Optical Theremin project, which I’m in the middle of, and which I’ll be describing as soon as I’m finished; the second one went into another jewellery box project.

I hope I’ve interpreted correctly what it says on the Music From Outer Space website, where it comes from, and it’s OK to reproduce the circuit diagram here:

You can read about Your First Wacky Electronic Noise-maker Thingy here: http://musicfromouterspace.com/ - just look for links to ‘WSG’ and you’ll find it. In fact, I just looked at it again and discovered that the circuit there is a slightly more advanced version of the one I built, incorporating fine adjustment of the filter cut-off frequency and a resonance control: looks like I’ll have to go back and make some modifications! . . . Later on I may have to build the whole thing . . .

Note in the diagram above the correct way to wire the cut-off frequency potentiometer.  I used a logarithmic pot, because that’s what I happened to have, which exaggerated the effect of my error the first time I put it together of wiring the pot the wrong way round – no effect throughout most of the travel, then a huge effect in the last quarter-turn: wire it the right way round and you get the full effect through the whole travel of the pot.  Adding a 100k pot, wired the same way round, in series with the 1M pot, at the end marked ’1′ – which is what the slightly more advanced version includes – would help to make more precise adjustment of the tone.

I should add that the whole Music From Outer Space site is an absolute mine of information and worth reading in its entirety: you can learn about synth modules, study circuit diagrams/schematics and buy circuit boards and so forth associated with the projects described.

Once I get round to doing the modifications, I’ll add a comment or edit the post and show a picture of the finished article.