4x4pole: I dream of wobbly resonance.

2013 started brilliantly, let there be no doubt. By April, my newly found obsession with DIY synthesisers reached a peak, and a small pile of worthy project PCBs was collected. Then work took over. For a year.

There were times during that winter of DIY-lessness that I looked forlornly, hopelessly, at the cardboard boxes of fresh components, and the half-built units soaking-up dust on the shelf.

Whilst unable to actually build, dreams of what was possible prevented me from going mad, and I spent my spare moments planning, plotting, scheming for the day I would wield the soldering iron again. 

Now that that day has come I can reveal the outcomes of my lengthy pause for cogitation:

The original plan  for the 4x4Pole Mission was to house the 4 voices together with a scavenged Fatar keybed. Many panel designs were drawn-up, with this being the final iteration in that direction (click to view).

The problems with this setup soon became obvious:

  • Originally I was to cut the 79cm panel in 5mm thick Acrylic, however the panel would likely still flex during use. Not ideal.
  • The positioning of the screens and controls was too dispersed. If possible, I wanted to be able to see all screens at once, and I tried grouping them together, but still without satisfactory results.
  • Since I was going to all this trouble to mount a keybed, why should I restrict it’s use to just one synth?

In the end, I went for a compromise solution – but one I am very pleased with:

  • The keyboard/MKE and UC16 now become a master controller, with in-built MIDIpal, a 3mm anodized aluminium faceplate, and joysticks!
  • The 4×4 Pole is racked on a 4U 19″ Aluminium Panel, and with (wait for it…) more joysticks!

The redesign went much faster now that I had a clear idea of the endpoints. The alumium faceplates were quickly designed in FPD and sent to be cut by SchaefferAG.

I was very pleased with the results – with one minor niggle: I only used 2mm Aluminium for the Rack faceplate, meaning that there is some flex. This is fixable, but I should have used 3mm aluminium, similar to the keyboard panel, which has no flex at all. Oh well, you live and you learn.

Must say though, the anodized panels turned-out way sexy-looking. Victory candy included!

No excuses now….and no time to waste….

In my last post – a full 18 months ago! – I ravaged my poor old Evolution UC16 MIDI controller to prepare it for the installation of panel-mounted pots. In the intervening period the assaulted PCB eyelets gave-up hope and fell off, leaving me with a considerable amount of microsurgery to get the board working again.


Not pretty, but perfectly functional. Liberal amounts of hot glue and some PCB varnish should give it a few years of extended life.

Once that was done it was a matter of mounting the boards and wiring-up. Time-consuming but largely trouble-free.


Colourful though….


So, the keyboard components have been tested and all work together. MIDI is internally wired from UC16 > MKE > MIDpal > out. Initial tests show this to be a stunning combo. The MIDIpal rules!

All I need now are the side and back panels. I know what I want, I just can’t afford the additional cost right now (about 70 euro). However, this is no problem because the 4x4pole is quickly coming together. As of last night, all boards were working as expected, the filters have been tuned, and the sound of awesomeness is in the air.

Just the wiring to go…..


in 3 or 4 days I should be there.

Shruthi-1 SMR4 MkII – NoisyLittleBugger Edition


It’s finally here: NLB’s special edition modded Shruthi-1 SMR4 mkII.


  • 2/4 pole filter toggle switch
  • Bandpass 1/3 3-way switch
  • Filter Feedback FM toggle switch
  • Audio input routing switch – normal vs FM modulator
  • Filter FM Feedback amount pot
  • VCA overdrive pot
  • UFO Balcklight mod
  • Audio input and output volume pots.
  • OLED
  • Styroflex capacitors
  • Suitable clothes!

Check here for details of the joystick mods… here for the filter mods……and here for the drive mod.

Here is the SVG file for the case (right-click to download).

Many thanks to Mutable Instruments and the community.

Removing the spring from Gamepad Joysticks

[youtube https://www.youtube.com/watch?v=I3Nm4A_3N6Y&w=400&h=255]

To satisfy a couple of requests, I made this little vid to demonstrate non-destructive removal of the spring from standard gamepad joysticks (Xbox 360, PS3, Logitech, etc.).

If you don’t have a nephew’s games closet to raid, you can always buy the joysticks. This is my source. Tell Uwe I sent you.

Here’s where I show how to wire-them to the CV inputs of a Shruthi-1.

Back and building with vengeance…

New Drums:Yamaha RM50

This is the 1U rack version of the RY30. Release year: 1992 (manufactured from 1992 to 1995).

I’ve been eyeing them for quite a while and always thought I should pounce if one became locally available. Then two came along at once, and at 100euro per unit I had to get both.


Even without any expansion cards, for a 20-year old module it’s sounding great, with a small set of high-quality drum samples on-board. There’s a limited but interesting drum-focused synth engine. Pitched synth sounds are also facilitated, with basic but functional parameters to edit. No crazy pads here, but definitely usable for bass work – there’s a satisfying low-end.

The 64 factory kits are OK, if a little too generic. Although there are not many basic samples, the preset sounds demonstrate what is possible. My interest is creating custom kits. There are 128 slots for user-created sounds using the on-board and imported RAM waveforms. More than enough.
Each drum sound can be made-up of two ‘Voices’. Each Voice in turn contains 2 ‘Elements’, which consist of one drum sample and and it’s synth engine.

The synth engine is simplicity itself. There’s a very basic pitch envelope generator, an amp envelope, resonant low-pass and non-resonant high-pass filters ( (12 and 24dB), and LFO for pitch, cutoff and amp modulation.

The MIDI delay effect allows the first note to be switched-off, leaving only the repeats. Looks interesting, especially for glitchy/granular effects. Finally, there’s good control over velocity sensitivity, where volume, pitch, cuttoff and decay can be set to different values.

CC assignments are basic – 6 fixed parameters can be assigned for real-time control at kit level. The CC assignment can be switched off per sound, so there is some flexibility.

No internal FX, however some samples have reverb built-in, and this can be gated. Nothing fancy, but the sounds do seem to fit together, making it sound nicely cohesive out-of-the-box. The compactness of the 1u RM50 is also strongly appealing, though it is bloody heavy.

Deep editing with only the front-panel buttons may seem a challenge, but I’ve found it surprisingly intuitive and straightforward.
There’s also a cross-platform patch editor. Simple, but does the job – except no MIDI input under Windows.

Thankfully there are Headphone outputs. Until I get a proper soundcard I can’t record any clips, but soon….

Other nice things:
+ Each drum sound has two separately-editable voices or ‘Elements’ that can be mixed using the balance control which, mapped to CCs provide for morphing kits!
+ Handles heavy MIDI streams.
+ Deep velocity sensing options for dynamic kits.
+ Reversible samples.
+ Audio Trigger inputs = modular fun!

Peremptory niggles:
– CC control at kit level only, and only 5 fixed parameters. But selectable per voice.
– No real-time pitch change except via EG, which is rather basic.
– No sample start adjustments, however there’s a unique delay control per Element which looks good for glitchy sounds.
– No LFO to pan. Voice outputs are mono anyway.
– No noise samples, and a shortage of Hihats, however…

Lucky for me, one unit unit came with additional wave sample ram already installed, which allows me to upload 512k of my own samples. Not too shabby.

There is also a third-party expansion module available, but I think I’ll have enough in one.

I’ve already uploaded some Goldbaby samples via the MIDI port using Elektron’s C6 sysex manager. Easy.

A scanned manual is available, in three parts:

Overall, I am very pleased with these acquisitions. I was going to build both the Sonic Potions LXR and the TR8060, but I simply don’t have enough time recently, and it’s beyond time I stopped farting around and got serious about recording for a change.

Firmware check
– Switch off RM50
– Press and hold PLAY + MACRO + SOUND while pressing the Power On button

Factory Reset:
– Hold play + utility buttons when powering-on

3V lithium backup battery (CR2450)

Full specs:
– 16-bit AWM2 (48 kHz sampling frequency) with digital filter 22-bit linear D/A converter
– Layering 2 elements/voices, 2 voices/notes
– Polyphony 16
– Voices 500 preset + 500 variation + 100 user; optional Wave Card: 32 user (x3); optional Data Card: 500 variation + 100 user
-Rhythm Kits Internal: 64 preset + 64 user; optional Data Card: 64 user
– Operation Modes Multi Play mode, Multi Edit mode, Voice Edit mode, Utility mode
– Trigger inputs!
– MIDI Parameters Program change mode selection; Program change table; Control change settings; Control change assignment; Remote mode selection
– Displays Multi-function 48-character LCD; Edit LED (red); MIDI received LED (red)
– Expansion Slots External wave card slot (x3); External data card slot; Internal expansion memory board slot
– Connectors Headphones, Line out L/MONO & R, Individual line out (x6), MIDI IN, OUT, THRU
– Power 120/220 V, 14
– (W x H x D) 480mm x 44mm x 347mm (D); 18 7/8” x 1 3/4” x 13 5/8”
Release year: 1992 (manufactured from 1992 to 1995)

Cheap Bastard: Busboards are easy

Finally, I get some time to myself. But not much, so let’s crack-on.

First I connected-up the PSU to the mains. Now, if you are going to try this yourself I shouldn’t have to tell you to be careful, but I will anyway. BE CAREFUL. I’m not responsible if you fry yourself.

Anyway, I took no chances for first power-up. I stood well back and used a stick to switch it on, and I’m not ashamed of it 😛

But it works. So, out comes the multimeter to measure the veracity of the advertised voltages.


All looking good. Nice and stable output, no vibration and barely audible noise. The trimpot allows quite accurate tuning, however, it affects all three busses at once, and there is slight variation between them. Not a big deal?

Next was to grab some stripboard, a couple of large capacitors and suitable headers. It turned out that I ordered the wrong IDC connectors. So I have female connectors for the busboard, and male cable clamps. But this is no problem, as they can easily undergo a sex change with the help of some spare pin headers.



The connectors are wired as follows:


Initially I won’t be using the +5 bus, so I’ll only need the 10-pin headers. Stripboard makes it easy to follow the busses, and the whole thing took just a few minutes. This is just a first run. I’ll probably reinforce the tracks later. But as it is there seem to be no problems.



The capacitors are probably not even needed, but it’s no harm to ensure that the supply is filtered. Some details here.

After connecting the PSU, I measured the voltages at the pins and it seems that we’re in business!


Let’s have a dance to celebrate:

[youtube http://www.youtube.com/watch?v=PeeOPR8bxac&w=400&h=255]

Cost of each busboard comes in well under 5euro. Now I just need a finished module to power 😛

Cheap Bastard: No need for Baugruppenträger

While I’m waiting for some parts and test cuttings for the 4×4, it’s progress has been overtaken by this modular monkey business. I’m in no rush.

So, after assembling the Baugruppenträger and RAST, as expected, I required mounting rails to complete the job as first envisioned. Since I had a voucher for MusicStore, it was used buy the only rails they had – 12HE Adam Hall thingies at a fiver per pop. I got 4 to keep me stocked for future requirements.

Silly me hadn’t reckoned on them being made from a thick steel. As I don’t have the means to cut them, another much simpler and less expensive option popped-out: I could just glue or screw some 2cm wooden strips and mount the Baugruppenträgers directly onto these. Or, as I quickly realised, I could even just mount the vector rails directly this way and not require those ugly rack ears at all.



OK then – to hell with the Baugruppenträger!

However, it’s 4 vector rails can still be used to complete two 84HP racks, thus filling the RAST. So, for future reference, I will only need the Vector rails and the Thread strips. I could even do away with these altogether, but since I have them I’ll use them.

In depleting my voucher I couldn’t resist trying a Monotron delay, especially because it has exposed CV points on the board and would thus could be re-housed into a 4HE module. It’s nice but surprisingly noisy. That’s not necessarily a bad thing 🙂

Cheap Bastard: Power supply conundrum

As I’m quickly learning, starting a DIY modular is not for the faint-hearted. Of course it would be easy to cave-in, buy a basic setup and go from there. But that’s not in keeping with the skinflint ethos.

It’s when looking at how to power my modular that I fully realised how archaic and uncoordinated the Doepfer Eurorack format really is. Options for powering the thing are surprisingly limited, and connector conventions are known to vary between manufacturers. Just to get up-and running with the Doepfer PSU + Busboards + 5v converter comes to a total of 145euro. These are crazy prices. I went looking for reasonable alternatives.

The first question was whether I would build my own PSU based on the many specs found at Muffwiggler’s. After some consideration I decided not to go this route. Being still the newbie in this area, I decided against, if only because there is a real risk of frying myself and my surroundings if something goes wrong. So I’m better-off with a pre-fab PSU option.

Following-up a suggestion made on the Mutable forums (thanks yet again Frank!) I went for this one. For 24 euro it provides three power buses – 5v and +/-12v – and with a whopping 5A on the 5v bus.



The unit is heavy and seems quite robust, with clear connector options – if somewhat exposed. Handy enclosure to protect little fingers. There’s a prominent trimpot on the fron for adjusting the +5v bus. Not sure yet how noisy it is when plugged-in.

There is also a 5v =/- 15v version which will probably find it’s way here at some stage. But let’s keep focus for now.

I’m thinking to build the busboards myself using perfboard and some nice big capacitors, and based on schematics I found at Muffs.


Another option might be to just use a ribbon cable and connect the first modules in series until I can come to a more permanent solution. Much reading to do. Oh for some free time!

I have to face facts at the beginning that this modular won’t be a static device – it will likely grow and evolve as the addiction takes over. So I’ve got to be prepared to make big changes along the way, and try to anticipate my future requirements.

If I can get this PSU working as intended, I’ll be well-covered for power options.

Goodbye Tempest, Hello Freedom.

Well the Tempest turned-out to be a damp squib for me. Or rather DSI’s inability to implement features that were hotly discussed over a year ago. Maybe they will eventually pull the rabbit out.

Even so, there seem to be a lot of Tempesters who are perfectly happy with the current OS, and that has to be respected. As with everything else in life, people will have a different approach to, and expectations from, such an instrument. Everyone should come to their own conclusions.

[youtube http://www.youtube.com/watch?v=QereR0CViMY&w=400&h=255]

For me, the hardware interface, the sound and the overall concept were incredibly appealing. The forum was abuzz with excited discussions about the untapped potential, and we looked to be in line for a classic that would finally put the Roland box voodoo to bed. A powerhouse on paper, a legend in the making.

Now don’t get me wrong – in some respects it’s nearly there. I had fun designing sounds on this machine and I learned a lot along the way. The glass is definitely more than half-full, and I certainly do not regret the purchase.

But the sequencer is still too rudimentary for my needs. I didn’t see anything in the recent betas that give me confidence in user-requested improvements beyond the necessary bug fixes. DSI’s prolonged silences do not help. To me, most of the requests seem reasonable, and there appears to be broad agreement on the most important. The persistent bugs, though not serious, reveal a lack of attention.

Nobody is asking for a Cirklon-killer – only that the machine is given a sequencer worthy of it’s potential. And it seems so tantalisingly close.

At current pace, by the time the OS is ready, the second-hand market will already be mature, new competing products will have been released, and DSI will have lost sales. Brand damage has already occurred. I am amazed that the can’t see this! I would have said ‘bemused’, but it’s actually quite sad. I really want DSI and their products to succeed – the innovation is wonderful – but there seems to be no ambition to finish the job.

I can only speculate that the feasibility of new features is dependent on internal hardware limitations such as the the 4mb limit (honestly, what were they thinking!!), rather than the coder’s inability. He seems talented, if disorganized and overworked and, it would appear, fighting some serious battles to squeeze better performance out of this box.

Nevertheless, my biggest disappointed has been their habit of raising expectations and not following through. I honestly didn’t expect that, and I fell for it wholeheartedly.

They are also not giving the game away regarding what’s possible or likely to reach implementation. Engaging in further feature discussions seems pointless, as none have materialised. Beta testing has been the simple repetitive process of pointing-out the same few bugs in every release.

Admittedly my passion boiled-over after seeing the latest delay (of several months) was due to focus again shifted to the Prophet 12. OK guys, but at least tell us! It would be ridiculous to expect new product development to stop, but it is also reasonable to expect that existing products should receive the necessary attention.

In short, I am extremely disappointed with DSI’s behaviour. I would not have accepted this from another company, but I was rooting for DSI to succeed, and I trusted them. More fool me.

As a sweetener it was revealed that Tempest and P12 share a common codebase, and DSI have said that features will be rolled-back into Tempest. This sounds great, but makes me wonder, since the Oscillators are completely different, and P12 has no sequencer – the element that needs most attention. Time will tell.

I do earnestly wish DSI success with P12. It looks like it will be a winner and they have clearly worked very hard on it. They must be congratulated for the risks they are taking to innovate exciting new products. In the grand scheme of things, complaining about a few missing features seems churlish.

But when it comes to Tempest, I’ve simply lost faith. It had to go.

With the proceeds I have fostered a beautifully-built MIDIbox Seq4. Now THIS is a sequencer. It completely obliterates DSI’s effort. Open-Source with frequent updates still after 5 years of development. All that user input has been eagerly gathered and distilled into a great interface and a smooth, almost effortless workflow. And zero fuss from the unpaid developer.


There is more than enough cash remaining to start a nice modular and have a big party. I’ve already sourced the PCBs to make my own drum modules, which will be triggered by Seq4’s 8 analog outs. I can simultaneously sequence FM percussion provided by FS1r. What I lose in compactness I more than compensate for in flexibility and power.

Not only that, but there is still change to cover the entire expense of my 4×4 build AND to a fund a new build using a new third-party ladder filter in a 2-octave keyboard enclosure.

My first Eurorack module is ready to be assembled. Little Dieter is due to be born any day now…..


I’m sticking with open-source hardware from now on. No more ‘black boxes’. It’s AD 2013.

There will be no pining for Tempest.

Here are some Kits and sounds – the ones I managed to backup before the new owner ripped the Tempest from my arms. FWIW, he’s absolutely delighted with it.

Have fun!

I dedicate this one to DSI 😛
(One of my favourite basslines and easily possible on Tempest)

[youtube http://www.youtube.com/watch?v=e4OtIb6wDCQ&w=400&h=255]

The Euclidean Algorithm Generates Traditional Musical Rhythms

What do African rhythms, spallation neutron source (SNS) accelerators in nuclear physics, string theory (stringology) in computer science, and an ancient algorithm described by Euclid have in common? The short answer is: patterns distributed as evenly as possible. For the long answer please read on…