Modular Synth build

20 posts in 2 and a half years, and nobody has built anything? Somebody build something, anything.

Enter the Soundbrigade, and thanks Tubelab for hinting me about this thread.

Shortly: ALWAYS been interested in sound and generation of sounds. Started to build the Popular Electronics Psych-Tone in early ´70's. Got intrigued by the Practical Electronics Minisonic and later ETI's 4600.
Never built any of these, but recently I made a new start, buying a PCB for the (orig) Soundland from MFOS and later Soundlab Mk2. Both assembled today, but non hooked up to the panels.
I also ordered the Anushri (kit) from Mutable Instruments, ready built but not tested. Got two keyboards, 2 and 3 1/2 octave and the MFOS 32-step sequenzer in the production line ...

However I haven't come anywhere with all my projects and I am eagerly awaiting my apathy and letargy to wear off, but I have little by little started looking into a modular synth. Some modules are finished, some almopst finished an other just planned (Gerberfiles prepared).
My modules:
CGS Tube VCA (because tubes are fun)
CGS Digital noise
CGS Analog Shiftregister
CGS Infinity Melody
CGS LFO
BMS Arpeggiator
BMS Quantizer (x2)
Arduino Nano ADSR
A homemade clockmodule that clock certain modules.
I weill also add a homemade joystick (according to MFOS)
and
MFOS Keyboard controller

Much of these modules are "for fun". To get some backbone I am planning for the following:
Some Yuisynth VCOs
Yuisynth VCF
Yuisynth ADSR
A delay (PT2399)
A dual SSM2164 Based Voltage Controlled Amplifier
MFOS Ring modulator
Yuisynth Noise/SH

I will fix PCBs for most of the modules and will propably use either the Soundlab or a Thomas Henry SN Voice as a BIG module.

Btw, the modules are built on home made panels, not very fancy and will probably give the synth a rather amateurlike appearance. Pictures to follow, now it's bedtime in Sweden.

(I need a mixer and maybe some more LFOs)
 
Hi,
You can see mine in post #13 in this thread. It has several CSG modues, a few MFOS modules. I made my own PCB's for the simple ones, lfo's EG's, etc, since they had the circuits and descriptions published. Then I eventually learned enough to design my own modules. I miss the internet of those days (90's, early 2k's). People were so incredibly willing to share and teach. Still are around here, but it was a lot more common back then.
My band mate died and I put the modular on hold for a while, then life got busy. Now I've got a 4 year old and no time for anything. When he's a bit older and I have time again, the modular is first on my list :)

All the panels are self made. Cut the aluminum w/ a circular saw. Used a drill press. Labels are laser printer decals.
The section on top in the wood frame are all timing and control modules. LFO's, EG's, Pulse generators, multipliers, dividers, and simple sequencers.
The mid section are all percussion modules.
Then the big case on the bottom are mostly VCO's, VCF's, effects (reverb, delay), etc.. all the sound generation modules and audio mixers.
I can sequence entire songs on it.. No digital sequencers or anything, just all timing pulses controlling EG's to fade things in and out, timing pulses for the percussion, analog voltage generators going to quantizers for the melody, etc...

At the time, me and a couple friends were playing music at raves and clubs, and it was my goal to do 100% modular show.. I could do a few tracks.. maybe 15 minutes, but ran out of options after that. I had one more big section of modules planned that should have let me do it. I'll finish them some day, but probably won't be playing it live anywhere like I could have 15-20 years ago...
 
home made panels, not very fancy and will probably give the synth a rather amateurlike appearance.

As mondogenerator mentioned in that other thread:

Some inspiring stuff on "Look Mum No Computer" on YouTube.

Check out some of his front panels....and he has a unique YouTube channel with a serious following. I have been one of them for a couple of years.

His DIY modular:

YouTube

Mine is now sitting in pieces on the floor again. The case has shrunken to where it wont hold all the stuff. I plan to fix that with some table saw time on the next dry, reasonably warm day.
 
Back then (1970's) I couldn't afford the components to build a synth. The LM301's for the ETI4600 were expensive and my salary low, so the Minisonic was a good option and even had an etched PCB lying around for years. Today I realise that the Minisonic had some flaws and wasn't that potent.

With my many projects I have painted myself into all corners of my small flat, but I think I have a clear view what I have to do.

I am intrigued by the idea to have something plying almost at its own, why a sequenzer comes in handy. 32 steps or 2 by 16 steps doesn't sound like overkill ... ;). Adding the CGS Infinite Melodi also adds numerous variants and ways to "play" random tunes.

But of coarse I need a manual input and I will wire up my two keyboards AND a ribboncontroller for that purpose.
Then as I wrote, I will add the usual voltage controlled suspects. This will be the modular. Then I have to find ways to add the Moog Concertmate MG-1, the Anushri and the Soundlab Mk2.

As said the design isn't up to highest standard, but maybe an amateurish synth looks more inviting?! And I have seen a few of the Look-mom-no-computer videos and most of Ray Wilsons where he described his stuff. I have to know how things sound like.

IF I ever get my thumb out, I have some fun ICs to use for various addons, but that will be a later issue.

I am no musician, so I have to learn and it's my dream to learn how to play the guitar I built for myself. Guess that will a tough job :boggled:.

Here are pics
 

Attachments

  • 4546418976_IMG_6377.JPG
    4546418976_IMG_6377.JPG
    152.7 KB · Views: 222
  • IMG_7730.JPG
    IMG_7730.JPG
    127.7 KB · Views: 142
  • IMG_7716.JPG
    IMG_7716.JPG
    162.2 KB · Views: 118
  • IMG_7502.JPG
    IMG_7502.JPG
    198.8 KB · Views: 116
  • IMG_7501.jpg
    IMG_7501.jpg
    394.2 KB · Views: 121
  • IMG_7147.JPG
    IMG_7147.JPG
    219.4 KB · Views: 231
  • IMG_6259.JPG
    IMG_6259.JPG
    627.8 KB · Views: 235
  • IMG_6125.JPG
    IMG_6125.JPG
    275.2 KB · Views: 230
  • 4549058416_IMG_6376.JPG
    4549058416_IMG_6376.JPG
    163.4 KB · Views: 226
I plan to fix that with some table saw time on the next dry, reasonably warm day.

Despite weather forecasts to the contrary I had a few hours of 39 degree F ( 4 C) weather with sunshine and minimal wind yesterday So I started on modular cabinet #3. The rain came before I could finish so I tried again today, but there were strong cold winds out of the north that slowed my progress......which came to an abrupt halt when the snow started.

I am now in the process of figuring out how many mistakes I made playing in the cold, what's not done, and deciding what my next moves are. The next 2 to 4 day's weather forecasts are NOT compatible with my outdoor job site table saw.
 
Yesterday I saw a video from Behringer announcing the "System 100." It is basically a clone of most of the old Roland modular stuff from the late 70's, but dome in Eurorack format. Each module will be available separately for USD $50 to $100.

Today they fired another round in the clone wars by announcing clones of the old 1960's Moog modules, also in Eurorack format. No prices announced.

It seem that buying built versions of many modules will now be cheaper than buying the "board and panel" set form Synthcube and then buying all the parts and building it.

YouTube

YouTube
 
Today it rained all day. Tomorrow it should be below freezing all day, the cold and wind will linger for a few more days. That has put an end to the short trips outside to fix minor issues, and big ugly screw ups in "modular cabinet V3.0."

So, I decided it was time to stick with what I have built so far......If V3 is anything V1 or V2, I'll be ripping it up again, or building V4 soon enough, so I decided to stuff this thing with what I have.

Nothing is glued yet. Everything is put together with screws so that anything, and everything could be changed if needed, so it gets tested as it is for now. The pretty face can wait, I want to hear her sing first, then maybe she gets some clothes and make-up to be YouTube ready.

I would have liked to fit "blue face" into the rack somehow, but there are some serious issues with any possible path. Blue doesn't have a front panel, and Blue was conceived without thinking much beyond making a big prototype digital synth playground.

Blue could get a front panel, and a slight size reduction to be capable of fitting in a double height euro rack space. It's rather haphazard construction would likely not survive this. The best course of action, and the current plan is to build a new Blue. In the short term, the existing Blue will likely just be fastened to the side or top of the rack. Maybe Blue, and the two MFOS boards get their own box since they're all bigger than Eurorack......I don't need to decide this yet.

The next project will be to wire all of this up.....since working with power tools outside is not an option for several days.....now is the time.

After all of this stuff is playing with nicely together over one MIDI cable (yeah right) I can finish populating the boards for the 5 modules in the rack whose panels are present, but the boards are not. A couple of those boards are nearly finished, while the others are still in the mailing envelopes from Synthcube.

I already have several breadboard synths and circuits that are working, and can be made into Eurorack size modules fairly easy, so they get stuffed into the box.
 

Attachments

  • P3350133_x.jpg
    P3350133_x.jpg
    871 KB · Views: 153
  • P3350135_x.jpg
    P3350135_x.jpg
    861.6 KB · Views: 143
It's old and the CRT is weak so I have to turn the lights off to use it, but it is a handy tool for tuning stuff (like crystal filters) up...or finding out that the cellular band "Yagi" that I got off Amazon for $15 is a total piece of useless crap. I could do better with a bent coat hanger.....tuned up on the network analyzer. The $50 Wilson, however is at least close to 50 ohms in the 700 and 800 MHz cell bands.

It's actually the same unit that I had on my desk at Motorola about 15 years ago. They got me a better one, and scrapped this old one. I managed to buy it from the scrap company for like $100 because it didn't appear to work in a room full of bright lights. It still has the Motorola calibration sticker from June of 2000.

I have collected a lot of old HP test gear from hamfests and Ebay. I'm slowly putting together an RF test station similar to what I had at Motorola for the last 10 years I worked there. Some serious RF tinkering is beginning to take place.
 
I put a picture of my newly created modular cabinet in post #48. It had several blank panels from "board and panel" kits that I have purchased from Synthcube and Modular Addict over the past year or so.

Synthcube had another of their many "sales" so I ordered a few more bare bones kits (board, panel, and maybe the rare parts or pre-programmed chip)....…...This has got to stop.....Just Say NO...…there are about a dozen boxes with parts and stuff I have collected for synth building over the past 10 YEARS, yet very little building had occurred.

Time to FIX that. Read past the "Steps" to see the real reason for this insanity.

Step 1, drag out every box that could possibly have synth stuff in it. Dump them all out, and toss all of the useless bits into the trash. Good news, I found the MFOS Ultimate and Ultimate Expander boards that I haven't seen since moving into this house. They were in an old PC motherboard box. Bad news, Many of the pots have been stolen, probably for another project, but I don't remember doing it. Good news, I found my long lost Axoloti board, and a couple of other digital synth boards that I made......just what I need, more projects.

Step 2, I chose 5 Eurorack modules that I would build based on a few common traits. All were old school, using leaded parts with no unusual techniques or parts needed.

Step 3, dig through all the boxes with parts lists for the 5 boards in hand. 5 boards quickly became 4 since I realized that one of the boards I have didn't match it's documentation.

Step 4, Go looking for the docs (found), and get an email bomb from Synthcube offering 20% off, so I fall off the wagon and order 3 MORE PROJECTS (one is SMD)...NO......

Step 5, order the missing parts for the 4 boards I plan to build, and most of the stuff I will need for the 3 new projects.

Step 6, I make the stupid decision to attempt to build 9 Eurorack modules AT THE SAME TIME......what could go wrong.

Step 7, build day #1.....The Synthcube order has not even left Boston yet, so it's several days away. The Mouser order arrived a day early and the Digikey order is in the mail girl's Jeep on it's way here.

Step 8, I make the decision to divide the build into SMD, and through hole, which removes 3 modules from the build plan. I have parts and docs in hand for 4 modules, so it's build time.

Step 9, setup a mini assembly line on a large industrial table in the basement lab. I haven't taken on a project like this since leaving Motorola 6 years ago, so the process evolves as it gets used. This will become evident once a video is made....if that turns out to be possible.

Aren't there 12 steps in the recovery process......can they be replaced with solder?

My hands have grown shaky and numb in my old age, so I decide to break this up into several short sessions first thing in the morning when my abilities are best.

Why did I do this crazy project? I need a confidence builder / learning project......Tubelab must evolve or die. The techniques that I used when Tubelab (and myself) was younger don't really work today. My attempts at documenting the new TSE-II board have been failures so far. The tube amp market is shrinking. New products are needed, and my history with DIY synths goes back to 1971 with the PAIA 2720 kit. Some DIY synths / modules may be in the future. Maybe even some with TUBES in them.

I need to document the build up of two new Tubelab PC boards, plus a third when it is ready. In the past I have taken several pictures of each part as it's being soldered into the board, and written a paragraph on how it's done....Heathkit style. This is overkill and takes months, especially when you drop a camera on a concrete floor.

Now that we have YouTube, there is another way. Some serious experimentation is needed to learn how to best present a PC board, or complete amplifier build up. A full HD video of an entire amp build up would be very long, and would need to be broken up to avoid serious boredom for the viewer, especially if the viewer already knows how to stuff a board.

These board builds are an exercise in learning how to do this. For now I'm experimenting with time lapse. I basically take one still picture every second with two cameras, that make a time lapse video from the resulting pictures. I think a good presentation will need real time video, time lapse, and still frame pictures all integrated into something watchable.

There have been several short sessions over a week and a half. Some went too long which caused errors, mostly in the camera management department.

I am nearly done with 4 modules, and have taken about 15,000 pictures all in 4 K or better resolution. I have made far more mistakes in the photography aspect of this effort, than the actual build. I have learned that my lighting sucks. I have run out of memory card twice, and killed the battery in one or both cameras three times, so some of the build was not filmed. These are the little details that I'm learning to overcome before doing some tube amp builds on video.

So now there are two through hole modules and 3 SMD modules waiting to be built. Despite my best efforts two of the built modules are still missing one part each. They have been ordered. I plan to experiment with lighting and power / time management before starting the next build.
 

Attachments

  • P3600771_x.jpg
    P3600771_x.jpg
    771.3 KB · Views: 123
  • P3500154_x.jpg
    P3500154_x.jpg
    820.4 KB · Views: 197
  • P1290294_x.jpg
    P1290294_x.jpg
    699.7 KB · Views: 205
  • P1290087_x.jpg
    P1290087_x.jpg
    709.7 KB · Views: 200
  • P1240157_x.jpg
    P1240157_x.jpg
    607.3 KB · Views: 195
  • P1140019_x.jpg
    P1140019_x.jpg
    620.9 KB · Views: 202
Last edited:
A year and one month ago I made a short not in this thread and I think I just commented elesewhere mentioning the MG-1, that is about to be screwed together after a thorough update.

What started as a purchase of odd modules long time ago to complement the SoundLab (not finished) suddenly became an obsession. A year ago I listed modules I had and others I planned to have, even shipped gerberfiles to China.

I have since ordered some more PCBs and possible bought modules (kits) and I have also sort of decided what I am aiming at. The assorted modules from last year hinted no direction, but very much described Tubelabs 9 or 12-step method - "This one looks SOOOOO COOOL, gotta have it!!!"

I may have uploaded the picture below elsewhere but as it looks right now:
5 VCOs (Yusynth + Thomas Henry XR2206)
3 VCFs (MFOS 24dB, Yusynth diode ladder, Polivoks)
2 (3) VCAs (Thomas Henry and CGS tube VCA)
2 (3) ADSRs (trouble getting the one I have working OK - Yusynth)
3 LFOs (MFOS dual LFO, CGS Psych-LFO, Digisound-80 4xLFO)
1 Noise + 1 Noise & S/H (CGS digital noise, Yusynth)
1 Echo/Reverb (Scott Barnardi PT2399)
1 Ring Modulator (MFOS)
1 Kbd Controller (MFOS)
1 Quantizer (Barton)
1 Arpeggiator (Barton)
1 Analog Shift Register (CGS)
1 Infinite Melody (like the turing machine) (CGS)
1 16 step sequencer (Thomas Henry)
1 Joystick (mee own)
1 USB-MIDI to VC + gate (Mutable Instruments)
1 Clock module for stuff that runs on clocks (mee own)

My module standard is choosen to be 7" as I wanted my panels to be narrow (missed that :eek:) and as I have free access to sheet metal. I will also never have the need of swapping modules with anyone else.
This project was supposed to be a budget one, but I also realise that I will have a lot of value built into the machine when (if) it is finished.


I have started to have a look at the Polivoks. It has 8 PCBs mounted on 1 (2) motherboards making it very much pre-patched. The boards looks a bit not so professional and I got the idea of fixing new PCBs and one by one replace the old boards. Just an idea ... :idea:)
 

Attachments

  • IMG_20210128_124306_450.jpg
    IMG_20210128_124306_450.jpg
    422.1 KB · Views: 116
  • IMG_0060.JPG
    IMG_0060.JPG
    477.8 KB · Views: 110
  • IMG_8598.JPG
    IMG_8598.JPG
    516.8 KB · Views: 102
I just realized that over a year has passed since I have visited this thread. I have built, or partially built most of the modules that I had collected for my modular, but have not put all of them into the modular or even tested some of them. I successfully resisted the 3 or 4 sales at Synthcube, so my unbuilt module collection has dwindled to near zero. :)

However.......PJRC had sent me a free Teensy 4.1 board quite some time ago. I promised to make something with it, so what would it be? Another synth? Well, yeah! This time I wanted something small. The idea would be for a micro system small enough to throw into a backpack and take anywhere. Blue, the synth with too many pots seen next to my modular was too big and never meant to be transported due to it's front panel being perfboard.

After a few false starts, and lack of a plan, I backed up and decided to make a tiny version of Blue with a minimum UI....yeah, menus suck, but they are needed for small size. The hardware is just 10 rotary encoders, a small touch screen, a MIDI input, and a 16/44 stereo DAC with a headphone amp. All runs from USB power, or a phone charger bank. Software development is still a work in progress, sometimes slow progress with a few backwards steps. Code was never my strong suite, despite a computer engineering degree.

Given the processing power of a T4.1 this became an exercise in "how many poly synths can I stuff into a single board?" Currently the answer is three 6 voice poly synths and a crude sample based drum set. CV/gate was ditched, this box is MIDI in only. I have used about half of the T4.1's memory and about 30% of its CPU power with 18 notes playing with some simple drums.

I rushed assembly of this in order to take it on a road trip to Florida last September. I had planned to make some "synths in the sand" videos on the beach at sunset, but Hurricane Sally formed right on top of us during our trip, so one poor phone video with only a keyboard and the synth and no sound recording equipment was made.

I wasn't about to take my new expensive laptop, the Focusrite interface, or the NDLR out in the storm. Here the synth and the Arturia Keystep go play in the random light rain and blowing salt spray on day 1. The weather degraded quickly to no power, and flooded roads, leaving us stranded on the island for several extra days. I didn't even take my camera outside until the storm had been gone for several days, all pictures and video proved that the iPhone 10 is indeed waterproof.

The calm before the hurricane - no music, just the surf, the wind and the birds - YouTube

The usual setup includes the synth, the Keystep (mini keyboard and polyphonic sequencer), the NDLR (another sequencer, with algorithmic pattern generators), and the laptop running Audacity for recording the results, or Ableton Live for some poor attempts at music composition.

One day I was wasting time on YouTube when a shiny red box was dangled in front of me. As much as I wanted to resist, I failed. Within a few days, a T-Synth was on my bench and built. It is quite similar to the multi synth that I had just built, but it uses pots instead of encoders. Input is MIDI only. Currently both are connected into my modular synth.

I also revisited my vacuum tube synth design a few times in the last year. The VCA has been totally redesigned. It works better than expected. I'm now back to the vacuum tube ladder filter....or filters. The input and output circuitry is the same, but I have 4 different "ladders" with another in the works. All sound and behave differently. I ventured down a rabbit hole that has had me occupied.

Those of us who ever worked with really old radios or old ham gear may remember the regenerative detector, or the "Q multiplier" devices that would attach into the IF amplifier circuitry. These worked by applying just enough positive feedback to a tuned amplifier to bring it to the brink of oscillation, but not into oscillation. This creates a resonant peak in the gain / frequency response plot at the desired receive frequency while rejecting off frequency signals. The Moog ladder filter works in a similar manner. The resonance knob works like the "regen" knob on an old radio. It applies adjustable feedback such that one or more of it's 4 poles gets pushed toward the right hand plane.....the instability region.

The next step in radio's evolution was the super-regenerative detector. Here you turn the knob into the unstable zone. This would ordinarily create an oscillator, as it does on a ladder filter. Oscillation does not occur instantaneously, it takes several microseconds to built up. The super-regen allows the instability to build up right to the point of oscillation, then breaks the feedback path in order to "quench" the oscillation. The path is then closed and the oscillation begins to build up again. This boosts the circuit "Q" and creates a sharper resonant peak. The "quench rate" was usually in the tens of KHz. Can this technique be used to build a better ladder??????

One can not have too many sequencers, right? Again a Youtube video dangled a shiny silver object in front of me. It would fit right into my modular, but does not yet reside there. Yes, I got another module, and built it in two days. I then connected it to Blue where it has been ever since (three weeks). The power supply is from one of my early versions of the modular. It can't feed todays version.

The Super Sixteen is a variable length sequencer, up to 64 steps with some features for live performance. It is CV / gate only, no MIDI, hence the easy connection to Blue, which does not speak MIDI. Both of these will go into the modular when I clean up the workbench for the next project.
 

Attachments

  • P3990165_x.jpg
    P3990165_x.jpg
    951.4 KB · Views: 102
  • P3990161_x.jpg
    P3990161_x.jpg
    852.7 KB · Views: 112
  • P3990193_x.jpg
    P3990193_x.jpg
    959.4 KB · Views: 102
  • P3990224_x.jpg
    P3990224_x.jpg
    929.6 KB · Views: 116
  • P3990310_x.jpg
    P3990310_x.jpg
    823.8 KB · Views: 102
  • P3990311_x.jpg
    P3990311_x.jpg
    599.9 KB · Views: 101
  • P3990336_x.jpg
    P3990336_x.jpg
    900.3 KB · Views: 100
  • P3990338_x.jpg
    P3990338_x.jpg
    875.7 KB · Views: 105
  • P3990353_x.jpg
    P3990353_x.jpg
    951.7 KB · Views: 92
  • Like
Reactions: 1 user
Super Sixteen ... :scratch: ... gotta check that one up.

Someone said that you cannot have too many LFOs and I recently heard that you cannot have too many VCAs, but too many sequencers???!!! Yeah, why not?

I will have two, the 32-step designed by Ray Wilson and the 16-step by Thomas Henry. The latter looks like a 32-step, but it has dual potentiometers. The second set can be used to control a VC clockgenerator that clocks the sequencer, hence making it possible not only to adjust pitch but also length of each tone :cool:.
Ray Wilson constructed a sequencer with such a function but used a different (more complex) approach.

I can also use the sequencers in the Behringer Crave and Behringer MS-101 (*), but recently a different type of sequencer popped up in my head (well the idea did). A few years ago I saw on YT a sort of device where you put marbles in holes of a board. The thing was that the marbles either triggered different instruments or somehow changed the tones, but I saw how cool it had been to make a, let's say an 8-step sequencer where each step is hooked up to four switches and switches in each row are gated together to give 4 gate or trigger outputs. Have absolutely no clue if such a device is useful, but somehow I can see it as a very handy device.

Another matter I have to solve is powering all modules. Today I got two small regulators from China - +/-12V and +/-15V. I have a very suitable transformer, but there's a slight problem. Most modules run on either 12 or 15V, but some have their own regulators and would require a raw and filtered DC of 18V. But I may just desolder the regulators.
 
The Super Sixteen has two CV outputs per step. One is for pitch, and it gets calibrated accurately to 1V/oct. The other is for velocity, or whatever you want. It is close to 1V/oct since it comes from the same dual DAC chip, but there is no separate cal routine. There is a knob to set the step length on a step by step basis. There are several features including a "mutate" button with three modes, to make for some live pattern changes so that the sequence doesn't get boring. I have been playing with it on a mono synth for a while, and still haven't discovered all that it does.

The designer has a YouTube channel where he goes through the entire design process in several videos:

Extralife - YouTube

For now, it's a learning tool for when I decide to try and code my own into a Teensy board. I would like multiple concurrent sequences for playing lead, bass line, and chord patterns all at the same time with some intelligence for changing patterns, much like the NDLR does. MIDI and CV/gate outputs are both needed too.

I have rebuilt the power system in my modular synth twice, and a third rebuild in in the works. The first version is the perf board supply seen powering the Super Sixteen. It made 5 volts, +/- 12V, and +/- 15V from a small toroid and some TI linear regulators. Most Eurorack stuff wants +/- 12 V. The MFOS stuff wants +/- 15 V, most of the microprocessor based stuff that I make needs +5v, and +/- 12V. There are now 4 Behringer synths in my rack. The Deepmind 12 plugs directly into line power, but the Model D, the Neutron, and the K2 all want 12 volts at 1 amp each. They are currently powered by the wall warts that came with them. Everything else is running from a power supply robbed from an old computer.

Power supply 3.0 is being built. It will use concepts that I learned back when I built my first computer in 1975. There will be one bulk power source in the bottom, and separate linear LDO's for each row in the rack.

I have a MeanWell 15 volt 7 amp SMPS that's turned up to 16 volts. It feeds a 15 volt LDO, or a 12 volt LDO as needed throughout the rack. There is a separate 12 volt LDO for each Behringer synth, and two for the rest of the rack. 15 volt LDO's can run off the same MW SMPS as needed when I finish the MFOS stuff.

There is another similar system for the negative supplies.

There is also a MW 5 volt 10 amp SMPS, that's turned up all the way (about 5.9 volts) to feed individual 5 volt LDO's as needed.

As this thing evolves there could be some vacuum tube modules, and some MFOS synths that will be adapted to a semi modular format similar to the Behringer synths. Both of these mean that there will be some modules that are taller than the Eurorack standard.

The Deepmind is big. At 8.75 inches tall, it eats up 5U of rack space. The MFOS stuff is even bigger, but I'm laying out a new front panel to shrink it for modular use. I haven't made up my mind yet on the size of the larger stuff, but it will either be 4U at 7.0 inches high, 5U at 8.75 inches, or the "Kosmo" format proposed my the LMNC guy on Youtube at 20 cm high (7.874 inches).