Fair warning: I severely injured my left wrist and Therefore I'm dictating this blog post instead of typing it. If the prose sounds weird + weirder than usual that might be why.
Today's project originated from a few professional 8HP Eurorack modules brought to one of my electronic music geek group's public meetings.
The idea was simple: take a Eurorack power supply's +/-12VDC rails, reduce the voltages via front panel potentiometers, and power circuits downstream, "starving" them of power.
Voltage starvation was demonstrated on a variety of synthesizer modules including oscillators, filters, and logic processors, but I was most impressed with how changing the positive and negative rails drastically--and, at times, beautifully--affected the sound of voltage controlled filters.
I figured the underlying circuit design was probably pretty simple and I could quickly and easily create a starvation circuit using variable voltage regulators instead of buying pre-made (read: expensive) Eurorack modules.
What I came up with:
I designed the schematic and PCB using
Kicad; I used through-hole adjustable voltage regulators I had on hand:
LM337 and
LM317, along with other junk box parts.
The design was lifted straight out of adjustable voltage regulator datasheets:
Postive adjustable rail....
 |
| ...And negative. |
I sent the layout for the circuit board and front panel to this blog's sponsor
PCBWay, who, as always, got everything back to the US fast, fast,
fast--Many thanks to PCBWay for generously sponsoring this blog.
 |
As always PCBWay delivered the circuit boards quickly and the quality of their work was top notch. Help the blog out and check out PCBWay here.
|
....the starvation module took less than an hour to build and "worked first time" (WFT).
 |
| Front panel artwork used a PNG image of legendary (and thin) 60's model Twiggy converted to SVG using the python uv app here. Read more about Kicad silks used for front panel design here. |
 |
| junk box parts.... |
 |
| The design uses a Eurorack standard 16 to 10 pin 100 mil flat ribbon cable from the rack's power supply to the starvation module, then the same from the module to the filter downstream. |
So how did it come out?
The starvation circuit's outputs varied from about 2.5V to 10V on the positive side and -1.7V to -10V on the negative side. I silked "3V" on the front panel--incorrect. The outputs could be cleaned up with more carefulluy chosen R1-R4 values, but in the age of AI: I am that lazy; I am letting it go.
Overall, though, this one kind of blew me away.
I put very little thought and time into the project's design but when I set up the finished module on the bench, in between my bench power supply and some voltage controlled filters, I realized filters I had trouble dialing in sonically suddenly became much more appealing.
For example, the dual MS20 filter featured in
this post never sounded quite right, I tried different combinations of LED's and resistors in the resonance path, and other tweaks to reign in the filter's overly nasty distortion.
We all know MS20 filters distort, one of the things that make them cool, but my build was too much of a thing.
However with the voltage starvation circuit in series I had much greater control of the dual MS20 filter's frequency and resonance sweeps, its distortion, and overall sound...the module behaves better to when its supply rails are dialed back....to the point where I might even be able to drop this filter into one of my racks and use it for something useful.
As always there are many ways for this circuit to be further improved.
For instance, lower dropout variable voltage regulators would allow the rails to get closer to 12 volts allowing more of "stock" filter sound with this unit upstream.
I also want to revise things so the negative rail pot goes from -3V to -10V--opposite how it is configured now. Dunno, I just think I would like that better.
A slightly more ambitious version would use
boost-buck converters to move the rails up and down to +/-15 or +/-18 volts and then use a starvation circuit--or something like it--to bring the rails back down to whatever the end user wanted.
This would allow you to overdrive the filter's supply voltage rails as well as reduce them.
But what I have on the bench now is what I have time for this month. I will do a bit more work on this one then post the whole thing to
PCBWay's community site.
OUTTRO
Big news is that I retire from my day job in about two months. This after 37 years of slugging it out as a tech/tech manager.
37 years? Yes I am that old.
"old fellow out to pasture". There are other F words you can use instead?
All in all, good news; I hope to spend my newly found free time doing more fun projects like this.
No comments:
Post a Comment