Tech Thoughts: “Spielsachen” – growing sideways

Sometime in July, the Spielsachen setup had reached the Future Squared integration stage.

As for this strange name, I had at the beginning of my Eurorack journey created setups in modulargrid.net which were called “Today” (for what I had at that time) and “Future” (for what I might do in the following step). At one time, “Future” became “Today”, so there was a “Future of the Future”, or “Future Squared” for short. Consequently, the next stage will be called “Future Cubed”. I still got aleph-zero names available.

At this point, the setup could be considered pretty complete. As the last remaining big steps, the front row had been converted to three skiff cases (Neutron left and DFAM right, flanking a Moog 104HP case), a few modules had been swapped, and the stompbox choice was up to spec.

PICT_20191112_135658.JPG
Future Squared

Growing Sideways

What’s now bound to happen is what I call “sideways growth”. The setup hasn’t grown physically for a long time (pretty much since I got the Monster Base and the two wooden boards in the current setup), and with the move to very small modules already pretty much accomplished, there’s neither new or more powerful components on the horizon, at least not in the bigger view.

There’ll be changes in infrastructure (mainly PSUs), wiring, maybe the keyboard stands, and a lot of tasks in geometrical integration. For the actual functional components (read: synths, stompboxes, modules and the like), two branches happen independently on the upper tier and in the Monster Base. But first, let’s revisit where I wanted to go with this setup, because those goals pretty much hold today, and I don’t plan to change them anytime soon.

Requirements Revisited

The requirements are:

  1. A performance-oriented synth setup
  2. Analogue signal path
  3. Modular, more specifically Eurorack
  4. Portability (possible for a single person) and easy setup/teardown, a.k.a the “Live Requirement”

Design decisions from the past (which are still unchallenged today) included a setup on a two-tier keyboard stand, a top tier with small mixer, effects (stompboxes), an analogue drum machine and a polysynth keyboard, and a modular setup on the bottom tier comrising of a Monster Base (MB) and a row of skiff cases.

The MB is a flightcase design, so portability is not an issue (even though it’s heavy). For the front row (the skiff cases in front of the MB) and the top tier, sizing decision were made to allow transportation in a typical five-octave keyboard bag. Which allows MB in one hand, bag over the shoulder and stand in one hand.

There’s a few things lacking at this point. First of all, the devices on the boards (front row and top tier) just sit on top of those boards, so violating (4). A proper mounting is in order, and wiring to go along, also to allow for easy setup. Furthermore, a bunch of wallwart or brick PSUs currently reside on the floor next to a large power strip – that needs to change. And the keyboard stand (X-type stand with added second tier) is not easy to set up or carry around.

Otherwise, with those items taken care of, I’d be pretty much done. So let’s start with all these lacking things.

Geometrical Integration and Infrastructure

Let’s start with the keyboard stand: one of those centre-post things would be nice. They’re easy to set up for both tiers at once, and easy to carry. Something like this:

KB Stand.jpg

For the PSUs, there’s only two components with internal PSUs, namely the Mackie 1202 VLZ Pro mixer and the MB, and I’m not going to change that. Apart from that, there’s devices which require 9V (stompboxes, minilogue), 12V (Elektron Analog Rytm) and the +12/+5/-12 combo for the skiff cases. For all of the 9V demands, the solution is to rework an existing stompbox-oriented 9V supply with a stronger SPS so it can also feed the hungry minilogue. For everything else, a self-built thing is in order by using COTS (commercial of the shelf, i.e. readily available in a store) SPS components for probably the single most expensive component in this chapter.

There’ll be riser constructions on the top tier, to hide especially power supplies and some of the cabling, thus making more room for the actual devices. Devices will be screwed (PSUs, mixer) or velcro’d to the board, and cables cut to length and tied down for an orderly appearance.

There’ll be a single, shrink-tubed cable run from the top to the bottom tier, containing audio, MIDI and USB feeds, as well as power for the skiff cases. For this, the skiff cases need to be modified to be able to accept an external Eurorack power supply with four lines.

Power will go together on the top tier from the two custom PSUs, the mixer and the MB. I may use either a small power strip or, better still, an encapsulated screw terminal for this.

The remaining tasks for the front row are pretty simple: velcro the cases to the board. Colour-marking the two ethernet cables which provide multicore from the MB to the front row and the MIDI cables might help here (the audio cables are already colour-coded).

Ending this chapter: I’ve decided to make a sensible design choice by choosing white boards (and thicker ones as before, as they’re a little challenged physically) and light blue for the risers.


Further activities then require some choices. Fortunately, there’s an either-or choice each for the modules in the MB (“Power or Flexibility”) and for the devices on the top tier (“Cheap or HFE”), which are independent of each other. Fortunately, there’s no decisions to be made for the front row – it will remain as is.

A or B: Power or Flexibility

Recently, a lot of the design changes had been about quantitatively increasing a feature or reducing footprint. Plans for Future Cubed are no different. There might be a Doepfer A-111-6 to replace the Nanozwerg, a Pico BBD to replace the A-188-1 etc. On the other hand, I recently revisited some modules that had been removed from the setup for now (mainly because they’re too large), and there’s some among them (such as the A-143-2 QuadADSR) that I’d like to be able to easily swap in from time to time.

In the end, that means I could either go with a setup with increased functionality (“power”), or one stripped to the absolutely always required modules and a lot of space to swap modules in (“flexibility”).

Two current design drafts (again, done in modulargrid.net) show the general idea:

A: “Power”

In the “power” design approach, a focus was really on increasing power or reducing footprint, but in an ideal world both, while keeping the entire setup full. Major changes include a second LPG pair, a passive 4-channel switch, a smaller noise module with added T&H, to name a few. It’s more of a decent cleanup than a full Kaizen cycle.

Note that some of the changes visible below could also be accomplished here, and a few further ones. Examples include replacing the polarizer with a WMD Triple Bipolar VCA (same width but one more channel), the Doepfer A-111-6 in place of the Nanozwerg or the Hatz instead of the MFB CY-522.

B: “Flexibility”

For the “flexibility” design, removing components was done gently, while replacing was done more radically than above. There’s especially quantitative changes (such as removal of another A-110-4 and one of the A-111-3). As the only completely new component, you see the Clep Diaz.

The free space here amounts to a whopping 96TE – which is pretty close to a single 104TE row, and considerably more than a standard 84TE one. In other words, more than 1/4th of the MB would be vacant. I could easily return a QuadVCO and Source of Uncertainty to add some non-deterministic drone stuff, a Reflector and A-136 DIS for a processing-focussed setup etc.

As you might have guessed, even though the “flexibility” setup is smaller from the total module width, its resulting delta cost is also higher. Considerably.

1 or 2: Cheap or HFE

My first idea for reoganizing the top tier was more or less a tidying up and making use of the planned riser to make some space. In a very rough photoshopjob, the “cheap” design looks like this:

Tier Two New.png
1: “Cheap”

As you can see, there’s no additional components – but there’s less crowding, and there’s free space for a potential Volca (blue) and a further stompbox (green), which would have the fitting wiring put in place in case I wanted to add them.

After thinking about this, I reflected what I liked least about that top tier, or rather about the setup as a whole. It wasn’t unorderly cables. It wasn’t lack of specific modules, either. It was the slim kebyoard of the minilogue (and that idiotic lever didn’t help either).

So I set out to define the “HFE” (for human factors engineering) alternative:

Tier Two New Alternate.png
2: “HFE”

The key change relates to the keyboard, which is a CME Xkey37. It’s got USB connectivity and power over it, so I can easily connect it to the Hermod with a single cable. It’s also rather expensive.

And then I’d need a new polysynth, as that had become a kind of requirement. More specifically, an analogue one.

From what I already have and which would fit the size requirements, there’d be the Tetr4 only. Great synth. But not enough control. Make no mistake, I really like the minilogue’s sound, and I like to twiddle its knobs. It’s just about the keyboard.

No way to continue while ignoring the elephant in the room that is the DeepMind 12D (desktop version). It’s generally well-liked. It’s hugely powerful (twelve analogue voices!). And it’s not too large. Incidentally, it also has an internal PSU, so no need to blow up the existing 9V for this solution.

So this is the option 2: get rid of the minilogue and replace it with a keyboard and DeepMind.

Putting it together

The “Power or Flexibility” decision is an easy one, simply because I can move back and forth between both without any substantial lost investment or major efforts (it’s mainly swapping modules and maybe no longer needing the blanks). Also as stated, this decision doesn’t affect anything else.

This isn’t true for the “Cheap or HFE” choice: a move from HFE to cheap would mean considerable loss of investment, plus a lot of effort, as the top tier would have to be set up again. Unfortunately, this would also affect parts of geometrical integration, fortunately to a lower degree.

It’s noteworthy that while the driver behind the HFE approach is in fact HFE, more specifically the minilogue’s minikeys, this variant also comes with vastly improved power. Let me say it again: twelve analogue voices.


In the end, it’s not an easy decision. But one I need to make before I actually start gigging with this setup. Well, what if there was a much simpler setup that I could use for gigging…

But that has to wait for Spielsachen: growing smaller.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s