I’ve had this drum machine, the Alesis HR-16b, for about 15 years. I haven’t used it much lately because although it’s a nice machine which is really quick and easy to program and a lot of fun to use, it has to be said that the sounds are a bit too clean. It’s hard to get it to sit right in a mix, it always needs a bit of distortion or lo-fi filtering.
And with the easy availability of far more sophisticated software drum programming tools these days it just isn’t worth using drum machines for recording unless you really like the sound of them. And even then it’s usually easier to record the sounds you like and sequence them in a software environment.
I would, however, like to incorporate this machine into a live performance set-up. For one thing, I like to have sounds coming from multiple sources – rather than all out of the same usb port on the laptop, into the same channel on the desk, through the same effects etc – and secondly, like I said, it’s a really fun machine to use. Soooo…..
…. I was pretty excited to read this very informative thread which explains just how easy it is to warp the HR-16 / HR-16b to get a huge variety of distorted, gated, glitchy, modulated sounds. Oh Joy!
What I learned (from the link above) is that all of the sound data is stored on two 32-pin chips, and connecting any two or more of the pins on each chip to any other pin on either chip gives you the bends (so to speak).
I also learned to “stay away from pin 1 (program supply), 22 (chip enable), 24 (output enable), and 32 (+5V)”.
These are the chips:
And here’s what happens when you connect a few pins by poking screwdivers at them:
Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.
So the trick is to make some kind of switching system to make connections between the pins, preferably something a bit slicker than poking screwdrivers at them while the machine is running.
Most people who bend this machine seem to like a patchbay system, others use quite complex switching systems, usually built on the top of the machine which has a big flat space with plenty of room for sockets and switches.
There’s plenty of room inside, too – no trouble finding space for extra wiring. It’s a bit like working on an old car.
For some reason (beyond my current understanding of such things) the data is identical on both chips so it’s only necessary to use half of the pins. This makes soldering a bit easier as you can just use every second pin, alternating between the chips, and still get all the possible connections.
From a comment in the same thread (above) I got the idea to do all the switching and bending in a separate box, so I’ve used one of those long connectors you find on the back of a floppy drive (which is exactly where I found it) and wired each usable pin to a corresponding pin on the connector.
Now I can use a flat ribbon cable and experiment in a separate box without destroying too much of the case of the machine until I know what I want to do. Also I can potentially build several different boxes with different switching systems so it keeps the whole thing nice and flexible. Also I can put the machine back together and keep using it as normal while I experiment with breakout boxes.
————————————–
As you can see, so far I’ve just been putting together other people’s tried and tested ideas but next comes the interesting part – coming up with some way of switching the connections. I haven’t given it too much thought yet but I do have the idea that some way of fading between connections and combining different bends in a smooth, fluid way might be fun.
Maybe something with LDRs (photocells)? I think I’ll start with pots and faders and see how I go. Any ideas appreciated.
3 Comments
Man, you gotta look up the easy little sequencers one can build using a counter chip, the 4016… or is it the 4017? actually it’s not a counter so much as a sequencer of sorts. The input takes a pulse and passes it along ten different outputs depending on the sequence you choose and it had a reset so it can be 8 steps. A search on LED counter should get you there. I’ve been doing stuff for ages concerning the 4017 with a 555 counter and then using a 4066 which is a set of mosfet switches where the input pulse turns on a switch, each 4066 has four switches, spst. so you can see where I’m going with this?
By having the 555 counter controlling the 4017 which steps out a sequence, then the 4066 doing your switching you can automate the patterns of switching on your drum machine.
If I were you I’d check first if the data output from the chip is a straight soundfile by inputing it to an amp, through a cap of course, and that would give you the basis of how the system works.It would make sense then as the sounds are modulating each other as opposed to straight mixing. That said, given it is the case, then putting a resistor on the end of each signal would mix them and a pot accros them would define the area between modulation, of sorts, and mixing.
Another thing, which may or may not work, is starving the chip on its voltage input. It’s most probably, if it can be done, a very small span of voltage, that send the outputs into territory where they just work. Too much voltage starving and the chip won’t have enough current to operate and would most likely shut down but there may be an area that yeilds interesting results.
I like the idea of the Floppy drive input. Then you can use the controller you make for other circuit bending tasks.
Yeah I see where you’re going, in fact I’m already there! I saw this one the other day: http://www.youtube.com/watch?v=Qe1ButmzNWY and it made me wonder about exactly what you suggest re automated switching between the pins.
Not sure what you mean about checking the data output – how would I go about this?
And the voltage starving thing, is that what’s happening already within the chip when the pins are “crossed” to create some of the glitchy sounds?
Opps, not really one data pin so much as checking whether the pins that you are crossing actually output an audio signal or whether it’s a digital signal which is then processed somewhere else. If it is an audio output, off each pin, then a switch and a pot would offer all sorts of possibilities.
I don’t think the crossings of pins would create voltage starving so much as have the singles fighting each other and being in phase and out of phase all over the place and thats what would create the glitchy sounds. A straight voltage starve on the supply pin would , well I’m not really sure, but for starters it might mis-bias the internals and have gating type effects while also having more dropout on fuller sounds.
4 Trackbacks
[...] year’s version is a noisy, glitchy, live-in-the-studio jam on my drum machine with a ring modulator (driven by the Good Will Audio Generator), a crusty old sampler, an optical [...]
[...] Alesis HR-16B drum [...]
[...] Bending the Alesis [...]
[...] Alesis HR16-b drum machine [...]