Buchla 208 Random PCB

[TheAncientOne]

Just had another thought about this unit. Perhaps someone more expert in things Buchla could chip in here.

I seem to remember that Buchla used 1.2V/Octave in his synths, making some setups easier at 100mV/Semitone, rather than the 0.08333 in the Moog format.

If that's the case, maybe a change in the output scaling and cross coupling might be a possible modification for most current synths. I know one can always adjust the modulation depth control, but in certain areas I feel that may not quite be enough, because the cross mod from the 680K output to input cascade is acted upon by the gain of each output current amplifier.

Dammit! I'm just going to have to build one now.

(edited for missing '0' c- for maths today!)

[Peake]

Hey Prof!

I believe that the 200e use 1.2V/8va; the vintage oscillators (258) were 2V/8va.

[Tasmanian Alkaloid]

[303monkey]

I'm in for two boards if a run is made.

[tlu]

RERUN???? Count me in please.

[inlifeindeath]

On the issue of the CD4015 locking up...
I built a circuit based on one of my favorite Lunetta patches using 40106 oscillator and a 4015 SR. I had issues getting it to work like I wanted it to because it kept locking into the same sequence. Once I added a voltage starve pot, it would create insane freak out sequences but then once it got back to full 9v supply it would lock up again. I don't quite remember how I hook up the data and clock inputs but I thought it was interesting to see it happening in a Buchla (clone) module. Going to have to throw one of these together.
Here's a video of the device in action
http://youtu.be/gM4IZ28ccEI

[LFLab]

Etched a board for the 208 random a while back. I added a doepfer power connector, onboard cliff jacks and perf area for an on-board clock. Didn't feel like finishing one of the other 500 or so projects in progress, but felt like starting something new. Plonked all components in and started it up, had a monster of a short in the psu connector so it didn't work on first startup, but when the short was removed it worked like a charm! Good thing regulators survive shorts, or at least mine seem to have.

Quite a bit of output, so these will need a divider, not sure if I want to put in fixed ones or panel mounted pots, seems a bit overkill for such a simple thing.

Haven't tried clocking it really quickly.

[ShapeSorter]

Hey, I know that this is necromancy, and a newb question:

I have noticed that people have mentioned making eurorack modules out of this pcb. Are they changing resistor values when providing +/-12v or just leaving the circuit as is?

[bubzy]

Doesn't look like theres any reason to change resistors for 12v to me, give it a go, its not gonna explode

Nothing wrong with a bit of topic necro especially for this nifty unit!

[ShapeSorter]

Awesome! Does anyone know the output level, so I may incorporate fixed dividers into my PCB?

[Snaper]

Wow, a bit to big for euro...
Anybody did a smaller layout? (around 80x80 or something like this?)

[LFLab]

My rack is deep enough, so I just have the PCB perpendicular to the panel. Even added a eurorack IDC power header.
Didn't add a divider, but did add a simple lfo as a means of a clock source.
Not sure if it's on 12v or 15v (I have both in my rack).

[Snaper]

Then I have to redraw it in eagle:(
My rack is 115 mm deep, the panel is around 125 :S

[animalman]

Been getting interested in having more randomness in my system.
Looking through the schematic and wonder if it would be possible to get random pulses/gates from this circuit as well??

Also how do you guys go about attenuating the "hot" output voltages to
suit +12/-12 systems?

[elmegil]

Yes. Generally you'd use a comparator to get gates, and then gate - to - trigger to get triggers on the leading edges of the gates (and you could invert the gate and do gate to trigger on that to get falling edge triggers if that was your thing).

[animalman]

Thanx.
Guess i should try to play around with the 4015. Here the use the 4015 together wit a 4070 XOR for a pseudo random sequence.
http://www.doctronics.co.uk/4015.htm

[JovianPyx]

This may or may not interest you:

Using 2 4015 and an XOR will give you pseudo random binary numbers. However, all pseudo random generators repeat, like a counter (this circuit is called an LFSR or Linear Feedback Shift Register) and will repeat every 255 clocks because it is an 8 bit LFSR. This is notable in some old equipment and even some old ICs (74677). When this is used as an audio noise source, you can hear an obvious repeat as a kind of chugging sound. If the random number stream is used for control, it's highly possible that you can hear a repeat in the pattern. This repeat may or may not be good depending on your intention.

My point is that if you want pseudo randomness that can't be humanly detected as pseudo randomness, you need more than 8 bits in the LFSR. For example, you can build an LFSR with 16 bits and use the lower 8 bits as the random number, but this sequence won't repeat until it's been clocked 65535 times. Increasing the number of bits gives an even longer sequence. We can calculate the time before repeats by knowing the clock rate and the number of LFSR bits.

Here is a chart which shows the LFSR configurations up to 168 bits showing which bits to tap for the XOR inputs. (note that the bits are labeled from bit 1, not bit 0). For each, the sequence length will be 2^N - 1 where N is the number of bits.

The basic circuit is the the same, a shift register and XOR or XNOR gate with inputs from selected taps on the shift register. Note that not all combinations of bit taps will yield a "maximal" sequence. A maximal sequence will have a repeat after 2^N - 1 clocks. Non-maximal LFSRs will have a repeat after less than 2^N -1 clocks. The taps shown in the chart are all maximal. There are other tap combinations that are also maximal, but finding them is not trivial.

HTH

[animalman]

wow that looks interesting. Honestly couldn't wrap my head around that bit-chart...but definitely something i'll need to research and digest.
Right on!

[Snaper]

I've finished mine, but its no working...
I used banana jacks.
I've read that Peake mentioned something about bananas, but I think its not the problem or yes?

[fluxmonkey]

[Snaper]

[Snaper]

Ok, so checked.
I have only 0.52V on all outputs, no change.

Measured the ICs and every power rail is OK.
No etching problems, short...actually, this is one of my best home etched boards...
Checked every single cap and resistor, everything is OK.

Now the IC brands :

- (logo)/812 LM3900N
- (logo)/812 LM3900N
- MCY 74016 TOMI

The 4015s (tried two type)

- (logo)39CXN7KE4 CD4015BE
- (logo)P9812AB CD4015BCN

All of the ICs are new, unused.
I think the problem is somewhere around one of the LMs, because, I can't measure anything at the DATA input legs of the 4015

Ideas?

[Snaper]

Another info :

I can measure around 13V on every out without any trigger at the input.
Right after I apply a trigger, every output goes down to 0.52 and stays there until power down and up.

Don't really know what could be the problem :S

[Snaper]

Switched the lm3900s and now two of the random outs are working...hm

[Snaper]

Replaced both LM3900s, and now everything works like a charm