4007 (was: Re: DIY parts)

[Magnus Danielson]

>>>>> "t" == tomg  <tomg at vivid.net> writes:

Hi!

 t> On 23-Aug-98, jhaible wrote:
 >> It's no good idea to have the discharge circuitry of several
 >> VCOs too close together. Your always fighting "interlock",
 >> i.e. a limited minimum beat rate between VCOs, before they
 >> are syncronized (soft sync'ed) thru parasitic coupling.

This is touching a subject that we have been discussing before, both
for the Minimoog and Oberheim SEM if I recall things right.
I also think that some of what I am about to say I have said before,
but none the less:

When squeeking my ASM-1 VCO upto 116 kHz I started to notice various
wavedistorsions around the reset-period of the wave. I used a
4-channel 500 MHz (2 Gs/s and channel) Techtronix oscilloscope with
350 MHz probes for my analysis and framedgrabbed the measurements.

Well, as the reset-cursuit turned off the reset FET I saw a distinct
ringing which had a lowest frequency of about 2.2 MHz. It was stable
as a rock and didn't change abit due to frequency adjustments. When
looking at one of my waveforms I suddenly discovered that the output
of the LM311 was swinging 15V in something like 25 ns and thus was
suddenly pulling a lot of current. This will together with the
inductance of the PCB strips and capacitor longer away from a slightly
resonant curcuit. This is well known trouble in the EMC field and any
experienced and thinking analog and digital designer working above 10
MHz should know about these things.

I could also clearly see how this impulse would eat itself into the
signalpath in various ways. Carefull examination of an unsoldered
laminate also showed that these power-rails where passing the LM311
and feed further to the CA3140 (used as the integrator of the
oscillator). Power decoupling caps where miles/mils away and thus not
very effective.

So, when people talk about intermodulation like this I can certainly
beleive it. Altought being an "analog" curcuit it suffers problems
which is "digital" (and these same problem is being viewed as analog
for those people, talk about irony). The frequency of these things is
way above what the curcuit is supposed to operate at, but it may still
make a noticable effect.

So, keeping such cursuits abit of the normal powerlines with
highfrequency caps closeby would certainly help to remove such
problems. It is standard procedure for digital cursuits, just apply it
to analog.

For anyone that is really into viewing things, peek in at this
unfinished page to see my picture:
http://www.it.kth.se/~e93_mda/synths/friends/stopp/asm1vco.html

I will make some further measurements and try out propper decoupling
to help out ASM-1 owners as well as others to get a clue on this thing.

Cheers,
Magnus