Adding Eurorack Modules to a Microbrute

The electrical aspects

The control voltages the mirobrute produces and accepts are compatible with the eurorack standard, so you can just plug microbrute outputs into eurorack inputs and vice versa and expect your system to do something more or less predictible. But it may not give you the most musical result, so it is worthwhile to take a closer look at the generated control voltages: the two modulation sources (envelope and LFO), the 1V/oct pitch CV and the gate CV; as well as at the audio levels.

LFO and envelope

On my exemplar, the LFO has a voltage swing of slightly over ±5V (10V_ss), except in the very high frequency range between the last two tickmarks, where is acts a bit strange anyway. This is quite a typical range for a eurorack LFO. The original Doepfer standard defines LFO voltage to be ±2.5V (5V_ss), and many classic Doepfer LFOs have outputs in that range, but many LFOs currently available can go up to ±8V (16V_ss).

The peak voltage of the envelope is a bit over +4.5V, which is slighly on the low side for eurorack envelopes, the standard value is +8V. So if you want to control eurorack modules with the microbrute envelope, you may think about getting an amplifier module with a gain of two.

Both modulation sources loose about 0.1V if loaded with a 100k resistor (which is a typical CV input impedance for many modules), so the output impedance of the microbrute is sufficiently low to control a small number of modules.

There is one peculiarity one should keep in mind: the two CV sources of the microbrute both have an output attenuator (the two "amount" knobs in the lower row of knobs), while none of the six CV inputs in the mod matrix have an associated input attenuator. This is the other way round than the convention for eurorack modules: modules that accept a CV input usually have an input attenuator to control the effective amount of CV, while CV generators usually produce a signal with the full possible voltage swing. So you should plan to have at least a passive attenuator for each control voltage you want to route back from the rack to the brute. You can also use an amplifier with a gain knob for this, or, if you want maximal flexibility, a polarizer and offset generator.

The pitch CV

The pitch CV is 1V/octave, as it should be, and it is even less perturbed by loading with an external resistor, again as it should be, since pitch CV errors are far more audible than any other CV errors. For the same reason it may still be advisable to run it through an active buffer if you want to control several oscillators with it. There is one caveat, though: The lowest note of the microbrute is the C at MIDI note 12 (16.35Hz), and it generates 0V pitch CV for that note. Many oscillators and other MIDI to CV gateways seem to be built around the assumption that 0V corresponds to the C at Midi note 36 (65.4Hz, which different people call either C1, C2 or C3) or 24 (32,7Hz), and not all VCOs can be easily tuned two octaves down (at least I had to fiddle with the offset trimmer pot of my A-110-2 Basic VCO). So if you want an external oscillator to sound in unison with the microbrute one, you should get one with an octave select switch with a "-2" setting, a wide enough tuning range, or use some other means to subtract two volts from the CV. This is also a thing you should be aware of if you ever add another MIDI interface to your system.

Gate CV

The gate CV is the odd one out among the microbrute outputs. With a nominal value for 11.9V on my unit, it should be high enough to trigger any eurorack module. But when I connected a 470k resistor to it, it lost more than 2V. So it has an output impedance of 100k, which is on the same order as the input impedance of many modules. If you connect the gate CV to one of those modules, it will only see a gate voltage of 6V, and with two, only 4V will be left. This may still work in many cases, but in general there is no way you will be able to use the gate CV to reliably trigger two or more inputs without an active buffer.

Audio levels

Eurorack audio signals usually have a level of ±5V (10V_ss), while the mircobrute audio in- and outputs are line level, which usually means at most ±1.7V (3.4V_ss) for pro audio. The pure, unfiltered triangle wave at maximum output volume on my unit is ±1.2V (2.4V_ss), and I have managed to get it it to actually produce ±1,57V (3,14V_ss, as measured by the AC voltage function of my multimeter, I don't know where the actual peaks would be seen on a scope). So you need an amplifier with a gain up to three or four if you want to get the signal from the brute into a modular. On the other side, you must attenuate the signal from the modular by that factor if you want to feed it into the brute. You could use the input level knob at the back of the brute for that, ideally compined with an attenuator in teh rack. Another alternative would be to add a line out module to the rack and patch that into the brute.

So, which modules to get?

If you start your journey into modular with a microbrute, you are probably not in the demographics that will happily shell out the equivalent of a small car for your dream modular (at least I know I am not). Your thoughts will probably be more along the line of "I know this is an expensive hobby, I should probably budget something around two times the cost of the microbrute for a small initial outfit." Well, there are empty cases that cost considerably more than that, so targeting that range requires careful selection of the modules to add to the required buffers and attenuators. In the following I try to write down some of my thoughts about that.

Do you want another simple synth voice?

An obvious first idea is to get a handful of modules to build a standard subtractive synth voice: a VCO, an ADSR, a LFO, a VCF and a VCA, something like this:

This is a nice voice that uses about two thirds of our budget for the modules. If we add a case and enough cables we may even still have room for multipes and attenuators. It offers us a few more modulation options than the brute alone (for example, we can use one envelope each for the filter and the VCA if we use the envelope out from the brute). This approach would be far more useful if you had patch points between the oscillator and the filter in the microbrute. Then you could flexibly route either oscillator to either filter or chain things freely. But without that you are more or less forced to either use both parts as independent voices, to chain one voice as a whole through the other one, or to just not use some of the expensive new modules. (There are test points on the circuit board you can solder wires to and then drill holes for jacks in the case to make these signals accessible, but that is a topic for another day.)

With these limitations in place this approach just duplicates functionality that is already there in the microbrute. We now have basically two standard monosynths, and the more expensive one has the more boring oscillator section. I think that unles you really don't like the sound of the microbrute filter, the money would be better spent to buy modules that offer functions that are not present in the microbrute when starting to explore modular synthesis. Consequently, the goal for now is to build a more complex single voice around the brute. Building additional interesting voices out of modules can wait for after winning the lottery.

What's in a microbrute?

To find out what is missing, it is useful to check what is already there. To do so, I drew up a quick sketch of a modular system that should behave somewhat similar to the core ot the microbrute (without glide, overtone and the sequencer). The top row corresponds to the oscillator section, the bottom row to the rest.

The red cables correspond to fixed internal connections, the orange ones are the switches, and blue knobs and jacks are parts that are exposed to the user. All other knobs can be assumed to be set to some fixed, producer chosen values that mostly disable the functionality they would control. The audio in on the back side of the brute would be a fifth input to the mixer in the top right.

The oscillator section is quite powerful with the two kinds of waveshapers wired to the waveforms for which they make the most sense and the overtone section I didn't even model in the above sketch. Reproducing its functionality with standard modules takes some effort, so it would be quite a pity not to use it. (Of course the setup in the sketch would be quite a bit more powerful still with the two fully fledged waveshaper modules with many control parameters that don't exist in the microbrute. And I know that the way I coloured the Wave Mult. II does not faithfully reproduce the behaviour of the Ultrasaw knob. Just imagine someone has replaced the top three knobs on the right with some Lego technic gear contraption if that bothers you.)

One thing to note in the bottom row is that there is no direct control access to the VCA. The only way to control its dynamics is through the ADSR. This does greatly reduce the usefulness of the ADSR as an independent general purpose modulation source, as well as a modulation source for the filter.

What to use?

These observations suggest the following approach to take when adding modules to the microbrute: I want to keep using those parts that can more or less be patched as if they were independent modules, or that have fixed connections I would patch most of the time anyway. This is the main audio pipeline VCO-VCF-VCA, the LFO as an independent modulation source, and the ADSR as a dedicated VCA modulator. So everything except the red part in this modified image:

This leaves me with the audio input before the filter, one usable CV source (the LFO), and six CV inputs. It looks I really need more CV sources and at least some multiples or stack cables to use all these inputs. (This is indeed my main gripe with the microbrute: I cannot have filter sweeps independent from the VCA envelope). Maybe some more complex CV processors to have more lively modulations, and maybe a simple offset generator to go with the LFO, as I've noticed that the metalizer doesn't handle negative control voltages very gracefully. The oscillator is more flexible than most affordable VCO modules (tell me again, why is there no patch point?), so I am not quite sure what to do with the audio input. I might say that the brute has no true sine, so add one of these, but what good is a sine patched directly to a filter? Or maybe two through-zero sines to do FM and Ringmod, or maybe just a basic oscillator to fatten the sound?

Actual Modules...

What functions your first modules should perform is a highly personal choice and depends on what kind of sound you want to produce and what you miss the most. I am missing things like complex envelopes, so I will look at CV sources and processors first, but you may be after a fat multi-oscillator sound and need VCOs. But in any case we will need some utility modules that may not be as glamorous, but are at least as essential.

Utility Modules: CV in

The first task we have to accomplish is to get the control voltages from the microbrute into the rack.

The easiest part are the LFO and envelope (if you want to use them). They are strong enough to drive several inputs and don't necessarily need any active processing if you don't want to boost their levels. You can distribute them using a bunch of stack cables or one or two passive multiples (which are available in any colour and material from about every vendor ever).

As mentioned, the gate CV must be buffered. Its exact level is relatively uncritical, so any buffered multiple (there are some in the list linked to in the preceeding paragraph) will do for most purposes. But even here the high output impedance of the microbrute may be problematic. If the buffer has a similar input impedance, you may end up with a voltage under 5V, which will not suffice to trigger some modules with extravagant trigger levels (like the reset inputs of the A-147-2 VCDLFO). But you may not even need a dedicated buffer, as many of the pitch CV buffer modules contain a second buffer you can use for this purpose.

This bring us to the last control voltage to get into our rack, the pitch CV. Pitch is about the only control voltage in a synth where the exact value is extremely important, and this can make normal buffers problematic, as they may loose a bit of the input voltage (if you want to know the details: the output impedance of the CV source and the input impedance of the buffer form a voltage divider which lowers the input voltage slightly, and the voltage on the output reproduces this lower voltage exactly) or have a gain that is only almost one. This makes it advisable to use a dedicated pitch CV buffer which contains circuitry to minimize or compensate for these losses. (Although there are stories of people who have retuned all their oscillators only to realize that is was their pitch CV buffer that was miscalibrated all along...)

A typical, simple example is the Doepfer A-185-1 Bus Access, which contains a gate and a pitch CV buffer. It has the nice property that it can send both voltages to the bus. This can save a lot of cables if you intend to use many modules that can can receive these inputs through a 16 pin bus connector. As described above, this will give you a pitch CV that most people will consider as two octaves to high, but if you can live with that, a module like the bus access and a few passive multiples or stackable cables give you everything you need to get the control voltages from the microbrute into the rack.

Otherwise you need something to add an offset of -2V to your pitch CV. You cannot use any offset module for this, as they may not be precise enough. You need something that is designed to work with pitch CV, for example the Doepfer A-185-2 Precision Adder, which can be used to add pitch CVs or add precise octave offsets.

Another option is a dedicated octave switcher like the Beast's Chalkboard. This even has a second channel you could use for the gate CV, although it is a bit wasteful to use it for that.

Utility Modules: CV out

You can plug the output of any CV generator in your rack directly in to one of the CV inputs in the microbrute's mod matrix and it will do something more or less predictable, but it will also be quite drastic and likely not very musical. You need a way to control the amount of control voltage that goes into the brute.

The simplest module to achieve this is a passive attenuator that consists of just one potentiometer and two jacks per channel. They are availabel with two or three channels in a 4HP module from many manufactures. Since they are so simple, they are also an easy first DIY project if you want to get into that. This allows you to turn the voltage from the full amount down to zero.

A similar kind of module is the attenuverter. Like the attenuator, it has a pot and two jacks, but it adds an active amplifier behind the panel that allows you to change the output value from the input value down to zero and further down to the negative of the input value. Many attenuverters also add the possibility to add an offset to the voltage. Offsets can be really useful if you e.g. want to modulate a parameter that has no useful behaviour for negative control voltages with a LFO. Modules with several attenuverters often contain additional circuitry to use unused input channels as offset sources if they don't have a dedicated offset function, or to combine outputs as a mixer.

All these modules (except the last example) work on single CVs. It is not possible to e.g. patch CVs outputs from several modules into a passive multiple and then into some input to combine voltages (directly connecting outputs to outputs is one of the few things that may damage modules during normal operation). To combine CVs, you need a mixer that adds the input voltages to form an output voltage. A standard mixer like the Doepfer A-138a has a set of inputs with an input attenuator each, and maybe a further knob for the output amount. But there are also mixers that just add their inputs together without any control, which can save quite a lot of space in the rack if you know their inputs do not need to be attenuated.

The minimum you need to interface your rack to the microbrute is one attenuator channel for each CV you want to feed back into the brute. A simple mixer and an offset can be quite useful, too.

Utility Modules: Audio out

Everything said above about attenuators and mixers is also valid for audio signals. But unfortunately you will usually want to use different modules for audio processing than you do for CV processing. Operations on CVs usually have a linear effect: If a given modulation results in a pitch change of two semitones, attenuating the CV to half the voltage will result in one semitone, and amplifying it to twice the value will result in four semitones. But amplifying an audio signal to twice the voltage will not result in a signal that that is perceived as twice as loud as the original one. There is a rather complex, frequency dependent relationship between signal amplitude and perceived loudness that is more or less logarithmic. Therefore modules meant to process audio signals are usually described as "exponential" (the inverse of logarithmic) or sometimes as "logarithmic", in contrast to modules for CVs, that are "linear". This is just the difference between the linear mixer A-138a mentioned above and its exponential variant A-138b.

Of course, this is no hard and fast rule: you can always send audio through linear modules or CVs through exponential modules, but it may be more difficult to get useful results that way. The most obvious problem is that the usable parameter range of attenuator knobs will be condensed to a small region along the upper or lower end. Also, some modules meant for audio processing are "AC coupled", which means that they have a capacitor in the signal path that filters out slow moving DC signals, which makes them unusable to process most CV signals.

If you want to send the output of an oscillator to your microbrute, you must attenuate it from the modular level to the line level the brute expects, and you will also want to be able to control the strength of the new oscillator in the final result. You may be tempted to use the input level knob at the back of the brute for these two functions, but I dont think that is a good idea, as it is not really easy to use. I think it is best to use it only for the overall level reduction and use an attenuator in the rack for everything else. To set the input level, you'd first turn it all the way down, connect the output of your VCO directly to the microbrute using an adapter cable, and then slowly turn the input level up until a saw from the external VCO is about as loud as the brute's saw. For your first experiments with an external VCO it is then sufficient to patch the output of the VCO into any attenuator and the attenuator output into the brute. If you have more than one VCO, or your VCO generates several waveforms which you want to use simultaneously, you will of course need a mixer that comes with all the required input attenuators.

Using a dedicated line output module wouldn't change much. It would do the level conversion, so setting the microbrute's input level would be less critical, but unless it comes with a built in mixer it wouldn't provide a noticeable benefit. If you later build a more complete voice in the rack and do no longer route all the audio through the microbrute VCF and VCA, having an audio output module that also provides a headphone out can be useful, as can be modules that provide a symmetric output if you intend to uses your modular with long cables in noisy environments. But there are also people who send the modular output directly to an external audio mixer (better ones have gain knobs for each input that allow to adjust the modular level to what the rest of the mixer expects) and never bother with dedicated output modules at all.

So you can get away with just an attenuator or a simple audio mixer module when starting out and patch its output with an adapter cable into the microbrute line in.

Utility Modules: Audio in

For the goal of building a more complex synth voice around the microbrute by reusing its filter and VCA section it is not necessary to send audio output from the microbrute to the eurorack.

Once you want to do that, you need something to amplify the mictrobrute's output to modular level, using a dedicated line input module or a general purpose amplifier module like the gozinta. If you want to process the sound from other instruments, you could also look at modules that reconstruct a gate and envelope signal from audio input like the A-119 External Input/Envelope Follower.

But nothing of this is required when starting.

Utility Modules: Voltage Control for Control Voltages

PS: Cables

Count the number of input jacks on all your modules (plus the microbrute mod matrix). Count each multiple as one input. You will need at most that many cables, and from what I've read you will likely run into limitations if you have less than half of that. If you have fewer outputs than cables, buy more stack cables, multiples or modules.