Temporarily Unavailable

Modular Update (January 2016)

2016-01-29T20:11:00.001+00:00

What a difference a few weeks makes ...

The 'cases' are TipTop Audio 'Happy Ending Kits' - 84HP rails complete with a rather tasty power supply which takes up 4HP, leaving you with 80HP to fill with goodies. Unless you're prepared to go down the DIY route, Eurorack cases aren't known for being cheap - the HEKs are a good, relatively low-cost gateway since they can be used skiff-fashion on a tabletop or rackmounted as I've done - the stock 84HP rails will fit nicely, although longer rails (up to 168HP) are available.

I've decided to build up rather than along so I'm using a 12U desktop rack unit - since each HEK is 3U tall I have space for up to four of them. Currently, I just have the two but as anyone who's into modular synths will tell you, you should leave yourself some space for expansion.

My mix of modules is fairly utilitarian at the moment, but there's enough there that I can get some interesting, and relatively melodic, noises out of it. I'm using an Arturia MicroBrute as a 'controller' for the modular, since it has CV/Gate outputs - combine this with MIDI in and it's possible to either control it from the computer using Ableton Live or use the 'Brute as a CV/Gate keyboard. I've discovered some interesting things when using the MicroBrute, some of which I will go into in more detail at another time.

For now, the modules ...

In the bottom rack:

2 maker.ie LFOs
2 Doepfer A-110-1 Standard VCO
Doepfer A-118 Noise/Random
Doepfer A-143-3 Quad ADSR
Doepfer A-102 Diode Low Pass Filter
CFM Bipolar Half-Wave Rectifier
Pittsburgh Modular dual VCA

In the upper rack:

RYO 'Penta' 5-step sequencer
NoBots 'Double Buff' buffered multiple
Doepfer A-106-5 SEM Filter
Intellijel TriAtt
Intellijel uVCA II dual VCA
Befaco 4-channel Mixer
Plenty of space for future modules :-)

Like I said, fairly workmanlike but surprisingly powerful - I've got my eye on a few more 'exotic' modules (top of the list are an Audio Damage Aeverb and a Mutable Instruments Clouds) but these are on hold until I've got my head around working with the simple stuff. For future expansion, I'm working on having two VCAs per 3U of space (hence the Pittsburgh and Intellijel units) and after mucking around with the TriAtt and the Double Buff I'll be factoring in plenty of utility modules too.

Not to mention patch cables.

Lots and lots of patch cables (the Pittsburgh/Nazca Audio ones are very good)

Gate to Trigger Converter

2016-01-03T19:48:00.001+00:00

The problem - I have a couple of Korg Volca synthesisers (a Bass and a Beats) which have the ability to be synced from an external source (or another Volca). Unfortunately, whilst I found that the Volcas will sync to the signal from my clock source, the sequences progress on both the rising and falling edges of the clock pulse.

It turns out that the Volca (and many other synths, as it turns out) require a short trigger pulse rather than a longer clock or gate pulse. Fortunately, these are relatively straightforward to generate using a handful of components.

This circuit is based on one by Ken Stone (see here) - this is basically the passive part of Ken's circuit (three resistors, a diode and a capacitor) connected to a pair of transistor inverters:

Schematic of gate-to-trigger circuit.

The above circuit generates a 10ms pulse although the duration of the pulse could be modified by altering the value of the capacitor (and probably the resistors, too). If you're interested in the inverted trigger signal then you could either leave out the stage 2 inverter altogether or tab the signal at point B - I haven't bothered to do this, since I can't currently think why I'd actually want to do such a thing O_o

Here's the initial version assembled on stripboard

Gate-to-trigger converter on stripboard.

Since writing this I've added this circuit to my 'clock box' (hacked art supplies container containing the master clock, clock divider and attenuator) so that it's easily to hand when I set things up - generally speaking the trigger will be sent to the Sync In on the Volca Beats, with the sync-out being connected to the Volca bass if I happen to be using it. All good :-)

Given that this circuit only uses a small number of simple components I'm toying with the idea of using it as a basis for learning CADSoft Eagle and getting a 'proper' PCB made, since this is the sort of thing that would be very handy in the main rack.

Vibrati Punk Console - boxed

2016-01-01T18:00:00.001+00:00

A while ago, I bought a Vibrati Punk Console from the lushprojects.com website (worth a look) as a way of getting my soldering and general electronics mojo back in a relatively cheap and safe way. The VPC is sort-of like a turbo-charged version of the Forrest Mims Atari Punk Console and is capable of delivering some deliciously noisy square-wave shenanigans - amazing what you can do with a bunch of 555 timers.

The kit is provided as a bare board plus components, including a 9V battery clip and a small speaker - these do the job pretty well, but since work started in anger on the modular synth I've been meaning to put the VPC into a proper(ish) enclosure, replace the speaker with a suitable line-out jack and allow it to run from a wall-wart rather than a battery.

I finally managed to do this today, and here are the result - the enclosure is an old business card box I blagged from a work colleague. Unfortunately, the lid was extremely brittle and didn't take nicely to being drilled, so the case is lidless for now.

Overall view - the four pots pretty much hold everything in place, with the on-off switch to the right. The line-out and power jacks are also visible, as well as the voltage regulator.

Power in - nothing shocking here, just a standard 5.5x2.1 DC jack. I bought a bunch of these off eBay with wires etc. already attached. I shall be buying more - they're very handy.

Output jack - switched 3.5mm mono. Again, no dramas although the output signal is pretty hot; this can be fixed by either adding a couple of resistors to the output or running it straight through an attenuator. I do the latter.

Voltage regulator - just a 7805 and a couple of capacitors. The heatsink may or may not be necessary, but since I'm planning to run this (and other devices) from a 12V/2A wall-wart having the heatsink won't hurt. The only thing that's missing is rubber feet to stop the enclosure sliding around.

4-step sequencer (initial version)

2015-12-31T20:01:00.004+00:00

Fugly, but functional.

A 4022 counter, a bunch of pots and a piece of stripboard that was, on reflection, a bit on the small side. Still, it works pretty well and will definitely be used in future - lessons have been learned that will be applied to the next iteration, though.

And here's the schematic:

Fancy enclosures? Naaah ....

2015-12-31T12:41:00.005+00:00

You don't need to house stuff in a fancy enclosure if you're able to butcher a £4 art supplies box ... all that's missing are some M3 nuts and bolts to mount things properly.

6-channel attenuator

2015-12-30T13:03:00.001+00:00

6-channel attenuator

The most basic of devices - 12 jacks, 6 pots (2.2k in this case) and wire.
That's it.

Simple, but very handy for keeping hot signals in check.

CLK:2 and SPLTR: Master clock module and clock divider

2015-12-27T13:05:00.001+00:00

Master clock source and divider

Clock source (left) and clock divider (right) - since taking these I've added reverse current protection diodes after a little, ahem, accident.

The clock source is my second (or third?) attempt at making a clock module - initial versions used part of a 40106 hex Schmitt trigger and ignored things like buffering. Add to that a surfeit of diodes (with the attendant 0.6v drop) and the net result was something that did indeed provide a steady pulse but the output voltage wasn't enough to trigger things reliably. That particular unit has been repurposed as a square wave LFO.
The updated version uses a 4093 quad NAND Schmitt trigger - all four gates are used: one is set up as an oscillator with a variable resistor and a capacitor and its output is buffered through a second gate. Hence, two clock sources from a single chip. Use of 220μF capacitor and a 100kΩ variable resistor allows a range of tempos ranging from 'glacially slow' to 'insanely quick'.

In order to provide a steady 5V (or V_cc) pulse I figured that the best way was to generate the pulse from the supply rails, thus minimising any losses due to the use of transistors. The switching was taken care of by half of a 4066 quad bilateral switch with the output from the Schmitt triggers activating one of the 4066 switches. The switches are connected between V_cc and GND with the output being tapped by two 3.5mm jacks on each clock.

The obligatory blinkylights are handled by a bog-standard transistor switch - whilst the end result is probably more complex than it needs to be, the fact is that it works and triggers my Korg Volcas beautifully.

The clock divider is a bit simpler in terms of functionality and the circuit - it uses a 4024 7-stage ripple counter, triggered by an incoming clock pulse, to generate pulses at a lower frequency - in this particular case the pulses will only go down to f/32 although by using the additional pins on the 4024 it's possible to go down to f/128. In addition there are simple transistor switches connected to each output to provide the obligatory blinky.

It works beautifully from a 5V supply, but a better design might involve using the outputs of the 4024 to trigger a fixed-width one-shot pulse (from a 555, perhaps?) - this might be investigated in a later iteration, especially as my Korg Volcas seem to trigger on the rising and falling edges of the clock pulse.

Schematics

UPDATE (1/1/16)

It would appear that the Volcas are behaving as they should - the problem is that the clock pulses are, generally, too long. The Korgs expect a 15ms pulse to handle the synch/clock functionality. Whilst it's perfectly possible to use one of the Volcas as the master clock, I'd rather not go down this route if I can avoid it - for this reason I shall be putting together a gate-to-trigger converter which I can put between the master clock and the Korgs so that they behave themselves.

Wub II LFO schematic

2015-12-26T20:48:00.002+00:00

If you can't find a schematic package that you agree with, sometimes you've got to go back to analogue methods ... this is the schematic for the Wub II LFO.

Salient points are:

Waveform generation done by an ICL8038 - seems to work best with a dual supply (±9V in this case) and tends to run quite warm as it is an essentially resistive device. If you plan to run it at ±12V then make sure you've got good airflow.
Timing cap is 100μF which can yield a nice, slow sweep - alter to taste.
Ditto for the 100kΩ pot on pins 7 and 8 - this will alter the frequency of the waveform.
The 1kΩ pot doesn't really do much in this incarnation and can be left out as long as the resistors on pins 4 and 5 are tied to V_cc - these control the symmetry of the waveform; if they're equal then the rising and falling parts of the output waves will be symmetrical. 6.8kΩ was a rather arbitrary choice - according to the datasheet, any value between 1kΩ and 1MΩ is OK.
I used linear pots on the prototype board - log pots may be a better bet as the frequency control is extremely finicky.
Sine output is 0.22×V_cc p-p and the triangle is 0.33×V_cc p-p - additional amplification may be required, but I tend to run both outputs through an attenuator to modify the depth (putting a level pot in is left as an exercise for the reader :-p)

The schematic is after the break - it's a big image.

Synth DIY photo dump

2015-12-26T14:09:00.002+00:00

Random photo dump - taken from my Facebook page but presented here for those who don't want to step into Zuckerborg's content-free ad network

Breadboarded version of the guts of the 8-step sequencer - based on a pair of CMOS 4022 counters; one to do the actual control voltages and another to drive the blinky lights. Whilst it worked reasonably well, I came up with a better design which will get built in the near future.

First experiments with a 555-based clock source, and a demonstration of the DSO201 oscilloscope in action. It has variable M/S ratio but has since been superceded by a dual clock based on the 4093 NAND Schmidt Trigger chip.

I wish I could remember what the hell this actually is ... apart from a mess of wires.

The original master clock source - since 'downgraded' to square-wave LFO status because the output signal wasn't strong enough to trigger things reliably. Quite a few lessons learnt, there.

TipTop Audio 'Happy Ending' 3U rack kit with uZeus power supply - a big enough to allow a decent initial set of modules, cheap enough not to blow a hole in your bank account. Far less hassle than DIY.

Original clock source and 3 channel attenuator - the attenuator is as easy as it gets; two jacks (one in, one out) and a pot.

Work in progress, as was - the "Wub" LFO - used a buffered 40106 oscillator for the square wave and tapped the output of the capacitor for a decent, if somewhat weak, triangle wave. Superceded by the "Wub II" which uses the triangle and sine outputs of the Intersil ICL8038

127mm stripboard fits rather nicely in the rack. It'll do until I can make up some PCBs and tidy faceplates.

A big ol' storm of chaos ...

Testing the "Wub II" with my Arturia MicroBrute - hopefully I'll be doing a video of this soon.

Master clock source (left) and clock divider (right) - the source has two individually controlled clocks with two outputs for each. The divider will go down to /32. There are, of course, the mandatory blinkenlights.

The obligatory video - testing the clock module and divider by driving a Korg Volca Beats and Volca Bass from two different clock outputs - quality is a bit crap, but it shows that things actually work.

Modular Mayhem: The initial shopping list

2015-12-26T13:53:00.000+00:00

The current shopping list for the modular, not including stuff that I've built myself - all of this will just about fit in an 84HP rack, assuming the power supply panel takes up 4U itself:

VCO - Doepfer A-110 Standard VCO (2 off - 10HP each)
VCF - Doepfer A-102 diode low pass (8HP)
LFO #1 - Maker.ie LFO (4HP)
LFO #2 - Doepfer A-145 LFO (8HP)
Envelopes - Doepfer A-143-2 Quad ADSR (26HP)
Mixer - Befaco Mixer Kit (5HP)
VCA - Maker.ie VCA (4HP)
Other - RYO Penta Squencer (4HP)

None of this has been bought at the time of writing, so I'm open to suggestions for alternatives that aren't stupendously expensive. The Quad ADSR will likely be the first module to be bought, followed by the VCA.

Stuff that I have built myself, and will be described elsewhere in due course:

2x dual CV sources (dead simple - a couple of pots and some jacks)
CLK.2 CMOS-based master clock source
CLK.1 dual square-wave LFO - started out as a clock source but the output signal wasn't enough for reliable triggering, so I repurposed it as an LFO. The first 'module' I built.
CLK.SPLT - clock divider (CMOS-based, goes down to /32)
Wub II LFO - triangle/sine LFO based on the ICL8038 function generator IC (the Wub I was a 40106-based job which tapped the capacitor voltage for a triangle wave. It works, but the ICL-based board is more versatile)
Triple passive attenuator - two jacks and a pot per-channel. Totally passive, and extremely handy for taming 'hot' signals. A 6-channel version is in the works, but I need more potentiometers.
Passive mult - signal multiplier. Particularly useful for clock signals.

In the works:

White/pink noise source - still casting around for a decent circuit for this, ideally one that runs reliable on 12V. Most of the ones I've seen thus far require 15 or 18V. Using avalanche noise from a transistor is a logical thing to do, with additional filtering for the pink noise.
Buffered mult - to go with the passive version; I need to use up the pile of TL072 op-amps I've accumulated :-)
Sequencer - 8-step sequencer based on the CMOS 4017 decade counter.
Low-pass filter - something to complement the Doepfer
DAC - A fairly simple 5 or 6-bit R-2R DAC similar to the one described here. I have a few ideas which will allow me to create wild 'n wacky waveforms.
Buffered mixer - you can never have too many signal mixers.

It all happens here ...

2015-12-26T12:36:00.000+00:00

Adventures in modular synthesis begin here

Essential tools include:

Antex TCS temperature-controlled soldering iron - not a substitute for a 'proper' soldering station but a good enough choice if you're on a budget, this is a throwback to my railway modelling days. I've had good experience with Antex irons in the past. At some point in the future I'll splash out on a Hakko soldering station but for now the TCS does the job.
60/40 Pb/Sn solder - RoHS be damned, I hate working with lead-free solder, so I don't. Remember to wash your hands when you're finished!
DS201 Digital Storage Oscilloscope - £30 off Amazon. A surprisingly capable little unit which is no bigger than a mobile phone. Even though it's only single channel there's no excuse not to have one at that price - the stock firmware is pretty horrible, but it's relatively easy to reflash with something decent. It may get an article of its own at some point, and it's worth a purchase if you can't afford a bench 'scope. Dual and quad channel versions are available but cost quite a bit more. I honestly don't know how I got by without this.

DSO201 pocket 'scope with a Nokia 106 for comparison. Progress!

A digital multimeter - does what it says on the tin; measure current, voltage and resistance. Also has a transistor tester which is useful for checking LEDs as well. Has a 'hold' function, which is occasionally handy.
Helping hands - the closest you'll get to another pair of hands; this one has a magnifier as well but it needs higher magnification for electronic work.
Stripboard - Veroboard, call it what you will. I source most of mine from various sellers on eBay and it works out cheaper per unit if you buy in packs of 10. I generally use three different sizes - 127x95mm is handy for stuff that'll be mounted in the synth rack, since it's only about 1.5mm shorter than the Eurorack 'standard' (128.5mm) - that's close enough. 64x25mm stripboard is handy for doing 'sub-circuits' or things like voltage regulator circuits.

127mm high stripboard in a TipTop 3U rack - close enough for government work ...

A junior hacksaw and a selection of files - use the hacksaw to cut stripboard and the files to clean off any rough edges.