In the first post in this series, I described experiments with different types of inductor pickups. At the end of this, I had 3 types of pickup which I thought would be of further use: a mono guitar pickup; a stereo pickup using the insides of two Telephone Pickup Coils; and a stereo pickup using two 100mH inductors:
Later I added a similar pickup with two 200mH inductors I had been able to get hold of.
As can be heard in that earlier post, I had used my Macbook as the sound source, but I thought a better one would be a hard drive.
Some time ago I had bought a job lot of broken hard drives from eBay at a cost of about 50p each. My original idea for these was to scavenge parts from them – magnets, discs, etc. – and to use the arms for a purpose which I may yet get the chance to write about. (See this post for a description of how to dismantle them).
In this instance, however, I was looking for drives which would power up – many of them didn’t – and might emit interesting noises. Not noises you could hear directly, but noises that could be picked up by an inductor.
I went through most of the drives – of which the above picture shows just one boxful – and tested them with the 3 inductors I had used in my last test on my laptop: the standard guitar pickup; the two Telephone Pickup Coils; and the two 100mH inductors. The playing technique was simply a matter of powering up the drive and then passing the pickup slowly over the surface.
Surprisingly, the three different systems didn’t pick up exactly the same noises, and the range of sounds detected seemed sufficiently varied to make an instrument based on this approach worthwhile. This recording is of the drive pictured, using the three pickups described above, in the order in which they are mentioned, the guitar pickup, Telephone Pickup Coils, and 100mH inductors:
*
It would have been perfectly possible just to use the hard drives spread out on the table, but I decided it would be much neater to make a proper instrument in a box, with its own power supply and preamp. I was lucky enough to come across a supply of cheap wooden boxes which would be ideal for this project and, hopefully, some future ones.
There would be plenty of room inside for the power supply, circuitry and hard drives.
The first thing I added was an inexpensive handle to the outside (secured with nuts, bolts and washers, rather than screws, to ensure it won’t come off):
The second thing was to install a power source for the hard drives and circuitry. The obvious thing to do was to use a typical hard drive power source, which would typically provide 12v and 5v DC – the 3.5″ drives would need 12v to spin the discs; 5v would be a suitable voltage for 2.5″ drives and for the electronic circuitry. These were the main items I bought:
The 240v adaptor fitted into the corner of the box like this:
and an on/off switch was included inside the box lid:
Unfortunately, the mains power supply was too noisy to be used for the electronic circuits, so I used the mains adaptor for powering the hard drives and a PP3 battery for the electronics.
I used a box-type battery holder as it had an integral on/off switch and I was able to glue it securely to the inside of the instrument’s wooden case.
*
So far, so good; but for some reason the hard drives weren’t performing the way they were doing when I had been experimenting with them, and sometimes didn’t even appear to be powering up. Only when the power supply exploded one day with a loud crack and the lights went off did the penny finally drop! . . .
. . . I went away and researched the amount of power required by a hard drive – and it’s much, much more than I imagined. The ratings are normally found on the drive itself: the ones I had been using, when I finally read the labels, were rated up to 720mA @ 5v and up to 900mA @ 12v, although often somewhat less, but averaging out at about an amp/amp and a half. Here’s a couple of examples of where you can find this information on a typical drive:
As you can see, there are big differences between the ratings for these two drives – and even the 5v consumption can be surprisingly high. And that’s in normal use: when powering up, they can easily consume over 2A each during the first 2 or 3 seconds – no wonder my poor power supply couldn’t cope! You can see from the label that it’s only rated at 2A at a time for each voltage, so is really only suitable for a single drive.
So, my first step was to buy two new power supplies, 5v and 12v, each rated at 8A – a bit of an expense I wasn’t expecting (about £8 – £9 per device)! These were no longer to be incorporated into the box; they would be external, connected via two typical centre-positive power sockets on the rear of the box.
I was aiming to incorporate 5-6 hard drives in the box, so the two supplies would provide sufficient power for normal operaration – especially given that there would be no data input or output from the drives – but I had to arrange the power switches so that no more than 2 or 3 of the drives would come on at once.
I could have done this with just a row of power switches, of course; but a more interesting method was to purchase 4 of these ‘delay relays’ at only just over £1 each:
The large blue component is the relay. It’s not completely in focus in this image, but you can just about see that it can handle 10A (10A of mains voltage, in fact), so was well up to the job in hand.
Item 1 in the above picture is a multi-turn preset, which allowed me to set the delay so the relay wouldn’t come on for at least 2 or 3 seconds – the length of time a high-amperage spike might be caused by a hard drive powering up. This particular device could be adjusted for a delay of up to 10 seconds, so I set the 4 devices to work as follows:
When the on switch is operated, 12v and 5v power is immediately passed to the first one or two hard drives; after 3 or 4 seconds, power is passed to one or two more drives; and after 6-8 seconds to the one or two final drives. In this way, overlapping spikes are avoided, and sufficient power is available for the drives to work properly after they are fully powered up.
A series of 6 LEDs, 3 for 12v, 3 for 5v, showed when the power connections were made.
Item 2 in the above picture is where the power lines are connected. The centre connection is the ‘in’ or ‘common’ connection; either side of this are ‘normally open’ (normally disconnected) or ‘normally closed’ (normally connected) connections, which are then reversed by the operation of the relay. I needed the normally open connections, so in two devices the centre connection was a 12v line, and in two the centre connection was a 5v line; the normally open connections were connected to the LEDs and the hard drives’ power connectors.
The 4-pin Molex connectors taking power to the drives are wired like this:
Pin 1 (yellow) = 12v; Pin 2 (black) = Ground; Pin 3 (black) = Ground; Pin 4 (red) = 5v
So, the power section now looked like this (note: the on/off switch wasn’t quite in place when I took this picture):
For a short clip of the startup procedure, click here.
In this test I only used one hard drive connected to each of the three sections – the third one being unusually noisy! They are, of course, all broken in some way, but you can hear that the first two, as they start up one by one, are not at all as rattly; but these are not the sounds the instrument is designed to create: as we will hear later, each drive creates its own interesting sounds when probed by the instrument’s inductors.
*
So much for the power connections. Next, the electronics.
The first part of the circuitry was a preamp for the inductors. For this I used the same transistor-based preamp I had used before for electrets and inductors, with the inductor connected on the left where the microphone is shown:
As this was a stereo instrument, with two inductors fitted side-by-side, as shown above, I used two of these preamps.
*
I tested this, and it worked fine with the mains adaptors for the hard drives and the battery for the preamp, so I turned next to adding a tone control. I thought, as the instrument was based on inductors, that an inductor-based tone control would be the ideal thing, similar to the design that I had made before in the Bits & Pieces series, the ‘Active’ Tone Control – which, in reality, is a passive tone control with a x10 amplifier in front of it to counteract the drastic loss of signal strength.
This is my design for the two-channel version:
In this case I used a TL072 instead of the 741 in the original: it was more up-to-date, less noisy, and neater, having two op-amps in one single 8-pin package; I also altered the resistors between the inputs and outputs (pins 2 and 1, and pins 6 and 7) from 100k to 1M to further increase the amplification.
The only thing I found is that I sourced the parts for the original about 20 or 30 years ago, when it was evidently much easier to obtain a 1H inductor – this is a very large value, rarely seen nowadays, and I couldn’t find one.
However, inductors are like resistors, you can put them in series to obtain larger values, so I bought 10 @ 200mH, which only cost about £1, enabling me to create two inductors of 1H. I spaced them out on the circuit board, hoping to minimise interaction between them, and connected the outputs of the preamps directly, without a switch. This tone control varies the sound quite a bit over its full range, and I was fairly sure there would be one position which would be very similar to the unaltered sound of the inductors picking up the sounds of the hard drive in action.
(In the event, I had a problem with the circuit around the TL072, so the amplifier and the tone control parts of the circuit ended up on separate boards, as can be seen in later pictures).
The 8-way phono socket panel in the bottom left is where the pairs of inductors plug in, and allows 4 separate pairs to be connected at the same time. Multiple Molex power connectors like the ones illustrated above allow a number of different drives to be running at the same time, giving the possibility of more complex, multi-layered sounds.
*
The following picture shows the electronics in a more or less finished state in the lid of the box:
On the left, from top to bottom are Panel 1: 9v Battery Power on indicator light, 3.5mm stereo audio out socket, Tone and Volume controls;
Panel 2: 4 x Stereo Inductor inputs;
Panel 3: An LED on the left for 12v power on, and an LED on the right for 5v power on – the third hard drive or pair of drives.
On the right, from top to bottom, are the circuit boards for the tone control, the transistor-based buffer/pre-amp, and the op-amp-based pre-amp; and at the bottom, the 9v battery box with integral on/off switch.
*
To finish the instrument off, I just needed to arrange for the hard drives to be secured in the main part of the box.
I began by putting in a layer of foam rubber, mainly with a view to deadening the sound of the spinning drives. Some time ago I had purchased a roll of foam, advertised as a yoga mat or sleeping mat. It only cost about £4.50 and was quite big – about 2 metres by half a metre (perhaps rather narrow for sleeping!), and seemed the ideal thing for this purpose. I lined the box with the foam, sticking it down with hot glue. (Again, this is before I replaced the power supply).
To keep the hard drives in place, I used blocks of polystyrene, and cut more squares of foam to insulate drives which would have to sit on top of others.
Cutting polystyrene is messy and rarely successful, so I used an electric polystyrene cutting kit with a heated blade, like the one shown below. This cost under £10 and proved considerably easier and neater in this and other projects – and elsewhere in the house.
Taking care not to set light to anything, or breathe in fumes from burning polystyrene, I trimmed the pieces without causing any mess.
Turning to the case, I decided a companion box was needed to transport the leads, power supplies and spare hard drives; so I fixed another of the carrying handles to the new box, and stuck on two small engraved plates to indicate which one was the instrument and which one carried the parts.
Using the two boxes, the instrument and accessories could easily be transported together.
A small length of yellow plastic from a cable tie was fixed to the right-hand side of the lid of the instrument box to ensure that it stayed open at the best angle.
I recorded the instrument using Audacity on one of my old MacBooks. The pair of 100mH inductors were very noisy. I didn’t have time to find out why, so I unplugged them; but I found that the most productive technique was to use two sets of inductors – the 200mH and the telephone coils – one in each hand. This enabled me to search for interesting sounds in two places at once, to balance these sounds, and also on occasion to create interactions between them.
It was also possible to lay one set carefully on a drive, to continue picking up sounds, and leaving one hand free to operate the tone crontrol.
The following sound file illustrates some of the typical electrical/mechanical/drone sounds I was able to get from the drives:
Music Electronics 