[sdiy] Buchla 266 fluctuating voltages

[harrybissell]

You would have to do a LOT to get rid of the 'heartbeat' in a 5837.  I'd replace
it
with a simple shift register / x-or noise source... it would have exactly the
same output
levels but be ... heartless ???

H^) harry

Scott Stites wrote:

> Hi Peter,
>
> > My good synth chum Cynthia was kind enough to donate a 5837 she had to a
> > worthy cause (me), and I went ahead and built an 'inspired by page three
> of
> > the 266' module (uh-hum).  It works like a charm, but...
>
> I didn't use the 5837 - it was a tad too unobtainium for me, and I was
> a-feared it would come out sounding cyclical.  I've been dying to find out -
> can you discern a 'heartbeat' using the chip?
>
> I opted instead for a negative feedback transistor noise circuit driving a
> comparator 0 to +15V, which I figure was about the output of the 5837 -
> whether it is or not, the results have been excellent.
>
> > Any of you thinking of making one might want to jot this down:
> >
> > You'll be much more pleased with the results if you replace the documented
> > 100K pulldown at the output (after the diode) with a 220k to -15 instead.
> > It will increase your voltage range by about 4+ volts, which tends to vary
> > randomly only between 5 to about 7 without this.   And this range gets
> even
> > crummier (narrower) when you speed it up. This resistor mod will give you
> > true random between about .5 to 7+ regardless of the speed and it rocks.
>
> Other than the 5837 and the choice of counter, I've stuck to the original
> Buchla component values, and I have had a totally random swing from 0 to
> around +8V on the smooth output.  By coincidink, Jeff Pontius is coming over
> tomorrow and we're going to dive into it - I'll double-check that figure,
> but it's definitely enough to drive a VCO subsonic to ear-piercingly high
> whether the 'Probability' control is full CW or CCW.
>
> > I opted to not only bring the -3db, flat and -3 dB digital noise out to
> the
> > faceplate, but the noisey triangle wave as well, which is the heart of the
> > fluctuating voltages and really really fun as a source of stepped random
> via
> > a sample and hold.
>
> I've thrown the kitchen sink at mine - the noisy triangle was a thought - I
> discounted it, but after your post, I'll re-think that.  I've actually
> buffered the stepped output right after the 4016, and it is an output well
> worth considering - I find myself using it all the time, actually.  I also
> buffered the trigger firing the 4016 and again....use it all the time.
> Another thing I've done is buffered the ramp voltage off the timing cap.
> It's fairly handy to have around.  All of these outputs lend a certain
> synchronicity to some Buchla madness.  There are other bell and whistle
> modifications I'm considering.....
>
> > I have one small snaafu yet to fix - tit seems the vactrol controlling the
> > 556 r/c is such that the thing will crap out (stop oscillating) if you
> bring
> > the pot lower than about 2K in its range.   Simple matter of pulling out
> my
> > dog eared 555 cookbook to see what I've exceeded and I'll get to this
> during
> > the week sometime.
>
> I deviated slightly from Buchla on this - I used a 7556 (easier on the
> rails), but I've kept the same component values.  It runs from verrrryyyyyy
> slow to pretty dang quick (50 Hz or so?).  I've never made it just stop.
>
> > Cost of parts (sans the 5837) - 22 bucks and change at Mouser, including
> > pots, knobs and jacks.  Quite a bargain.  I'll see if I can post a
> soundclip
> > in a couple of days.
>
> On that, I'm always afraid to look.  If I want to know, I can always ask my
> wife (she keeps a running count on what I spend on all this folly).  I've
> pleaded with her not to tell me.........
>
> Cheers,
> Scott