Frac Rack DIY

[soup]

So who's still out there building frac? Show me some pictures!
I've had some time on my hands and nowhere to go, here's what I've finished lately...

[werock]

Your stuff always looks so good

What is the double envelope? The panel is very Blacet looking!

I haven't got any pictures as nothing is completed, but have a SSI2144 filter that is waiting for a panel, and just hooked up a VCO yesterday so need to do a panel for that. And I'm building someone on here a CB-55 drum module in Frac, again waiting for a panel to arrive from Germany. But mine are in mini-jacks

[KSS]

Ooh! I like the look and layout of that Sequential Switch! Clever to tie it together and have it look interesting, while keeping knobs, switches and LEDs accessible!
What circuit is behind the panel?

[soup]

Thanks! The dual envelope is a mfos dual ar. It is very much a riff on the blacet style. The sequential switch is from cgs.

[Cat-A-Tonic]

I have a couple old Frac modules on the wiring table.
Currently working on an Ian Fritz TGTSH,
and a combo panel with a CGS Psycho 6 LFO feeding an Ian Fritz Double-Well Chaos.

After finally acquiring a Kilpatrick Phenol and finding some better quality banana cables,
I have to say y'all were right.
3.5mm jacks suffer from the cable crackle
which can really disturb your patching flow
and disrupt large patches in a way that is frustrating to troubleshoot.
3.5mm cables are harder to make and break more over time than good bananas.
Banana connections are more reliable.

I kinda regret having invested so heavily in 3.5mm jack stuff.
I suppose some designs aren't as easily available with bananas (LZX for example),
but envy all your lovely banana systems.

Another format choice I wonder why we don't see more of is simple rack panels.
1u rack would be the smallest module unit.
This thinking seems to make a lot of sense for combo-modules.

[uniqview]

After finally acquiring a Kilpatrick Phenol and finding some better quality banana cables,
I have to say y'all were right.
3.5mm jacks suffer from the cable crackle
which can really disturb your patching flow
and disrupt large patches in a way that is frustrating to troubleshoot.
3.5mm cables are harder to make and break more over time than good bananas.
Banana connections are more reliable.

I am completely rethinking 3.5mm for my FracRak SDIY. Banana jacks have tremendous appeal, A lot as you indicate, due to mechanical integrity, But another key reason for me is the UX of having color coded jacks, so that functional distinctions are significantly clearer. Makes the patch language more understandable, easier to follow, esp, with specific goals in mind.

[Moog$FooL$]

WOW!!! nice builds.

it's great to see some "old time" frac users..... kinda reminds me of when i first started coming around here.

[soup]

Thanks!
I've never regretted going bananas.
I built a few new things in the last couple of months...

[diophantine]

@soup - lovely builds! Great layouts, particularly on the Stroh modules...

I sometimes regret not going with bananas, but at this point it would just be a huge undertaking. Sometimes it is also good having something "different" from my banana systems...

But another key reason for me is the UX of having color coded jacks, so that functional distinctions are significantly clearer. Makes the patch language more understandable, easier to follow, esp, with specific goals in mind.

Obviously YMMV, but I would suggest that it may often be better to not be overly clever or complex. Some color schemes (like the Buchla 200/200e) are often too distracting to me, but others (like the Serge and Buchla 100) are just the right amount where it is helping me "subconsciously" without me actually having to think about it.

[uniqview]

Obviously YMMV, but I would suggest that it may often be better to not be overly clever or complex. Some color schemes (like the Buchla 200/200e) are often too distracting to me, but others (like the Serge and Buchla 100) are just the right amount where it is helping me "subconsciously" without me actually having to think about it.

A fair point. And essentially the goal: enable the ability to work subconsciously, through a very clean, balanced UX design. Avoid bringing up unnecessary cognitive processes that delay or defer creative results.

Which all reminds me of a poster that was at a start up company I was involved in 5 years ago. It said words substantially to the effect: "A good user interface is like a good joke. If it has to be explained, it's not a good user interface."

[soup]

I don't think there's a right answer but most of my stuff is done with 8 colors. If you have enough labels on the jacks I think people can use the synth without knowing/remembering the color code but as they use it they'll figure it out which does makes patching faster. I also like how it looks.

[uniqview]

Starting to rethink this again, as I've discovered there are now color ⅛"/3.5mm cords. I'm probably slow: these cords have likely been available for a while, but only now that I'm starting to integrate FracRak modules have I discovered them.

As well, for certain situations, it's nicer to have the input jack that is not plugged in to anything be able to have a switch to ground, so that the particular input is not floating around in the wind, so to speak.

[uniqview]

My first SDIY module, in a really long time. Actually, about 39 years.

I call this synthesizer the Model III, because it's my "third" system, after two others I first built in 1976 and then in 1979-81. Over all this time, I have flip-flopped over implementation technologies numerous times. I struggled with synthesizer designs using DSP in TTL hardware versus CMOS hardware versus FPGA hardware. Or several looks at implementing synthesizers just with software, esp. in the last 5 years. I also spent a lot of time discovering implementations the way Bob Moog originally did them in the mid-1960s, direct construction with discrete silicon transistors. That was a very valuable experience. I learned how to do a lot of system functions at high performance, using devices that cost pennies each today. This entire quest is something I've blogged about at Model III EMS.

In analog versus digital, I have finally decided I want to complete a life-long exploration with unique analog circuit design techniques. I feel compared with digital, these are things that are now unexplored. Digital is ubiquitous, analog is more unique. And for me, this kind of synthesizer is a kind of working with the "reals", mathematicallly speaking. I certainly do remember when 12-bit DACs were really expensive. Yet now one could construct kinds of synthesizers using a USD $35 Raspberry Pi, with ease. And have that computer tied to a HAT that has 24-bit 192 KHz stereo audio output.

But what about other concepts for sound synthesis? What if analog was explored in more depth? I have been fascinated since my late-teen years with circuit design using a really unusual op amp called the LM3900. This is a current-mode op amp that which National Semiconductor first made in September 1972. So, now in 2020, what I'm building is a new kind of analog synthesizer, mostly based on the LM3900, as well as some of my discrete transistor research, and 4000-series CMOS. I have some initital blog posts about my LM3900 research at Beyond AN-72.

I selected PAiA FracRak as a mechanical packaging system around the early 2000s. I really like the this format, for a number of reasons. I originally conceived of the Model III as a 5U system, but over time I realozed the space required for that format would be prohibative. Eurorack came along later, but I find that format much too small for my eyes and haptic dexterity. FracRak is perfect, right in the middle, able to flexiby support a lot of different circuit ideas, very friendly toward SDIY. I also like the PAiA case itself, becuase the depth is sufficient for complex discrete or other circuits, in terms of supportable area for PCBs. The case is also designed for SDIY options, and there are flexible and space effcient power distribution options.

Today what I completed is just a test and measurement module: a truly simple control voltage generator with rotary switches. I built this so I can do testing with new circuit boards that I'm presently working on with the LM3900 designs. This design used a precision voltage reference, ultra-precise 0.05% resistor strings, and CMOS chopper-stabilzed opamps with 30 microvolt offset voltage. Basically, this module is intended to provide precision octave and semitone control voltages for various modules under design. It does not have fancy panel graphics, just label maker tape. But it works!

Here's some "module selfies" during the evolution:

Every journey begins with a first step. Other, much more sophisticated modules will follow