[sdiy] PSU ramblings

René Schmitz uzs159 at uni-bonn.de
Sat Jun 29 12:58:39 CEST 2002

Hello Batz,

At 11:25 29.06.02 +0930, Batz Goodfortune wrote:

>I'll leave aside the fact that they're only 300mA for the time being. 

Thought about that too, well, I got four of these units, should I run out 
of current...

>(Can't cause serious welding with that kind of current.) And I don't know 
>about the grounding thing. That could be more trouble than it's worth. 
>Though on the face of it, it would seem to be something of a solution for 
>de-synching VCOs. But! Often in mixers, I've used a pair of 78Lxx/79Lxx's 
>on each channel. An "L" series will give you about 100mA local and if you 
>had some LEDs or something that switched in with a thump, you can always 
>run them from a separate "L" series reg.

Hmm, a 7815 or 78L15 would actually also do for the post regulator, giving 
the benefit of short circuit proofness. 

>As for jacking the ground I'm not entirely sure you'd achieve anything. If 
>you did that you'd have to derive everything from that source and that 
>would be the same as not jacking it. It might provide some degree of 
>isolation depending on topology but I doubt it would be worth the effort. 
>Although it's harder to achieve in a modular or a modular mixer, a good 
>STAR point is probably more beneficial.

Actually the wireing will be star point anyway. Just that I would insulate 
some (if not all) modules from GND, makeing the connection to the actual 
star ground a hi-Z one. When there is no current flow there are no voltage 
drops. Infact if the "dustbin" would be sitting on the module, one could 
easily bus the reference ground arround.

Jacks would be connected to the PSUs ground directly, whereas the circuits 
grounds would be buffered. That is no problem as I already have a reference 
and a current ground on the power connectors of the modules.

>Star-Ground distribution is very difficult to achieve in a great many 
>designs that are modular and/or expandable. What I've found to be a better 
>approach is to run a dirty great thick ground trace across the back. 2" 
>thick copper rod if necessary. And in the case of mixers, I tend to try and 
>cluster all the Is and Os together on the back and make that my Star Point. 
>That's practical in a mixer but not terribly practical in a modular synth 
>unless all your jacks are clustered away from the actual modules.


>The only time I've seen good use of an artificial ground point was in a 
>single supply PSU run from a battery. Where there was no centre. The design 
>used a power op-amp to provide this reference.

I'm thinking along this lines to get something useful out of the 25V.

>And don't forget your Rs and Cs. You can get a lot of mileage out of 2 Rs 
>and one or two Cs. I can't remember the last time I powered an op-amp 
>direct from the PSU in an audio circus. It's always a matter of course for 
>me to feed them through a pair of 100ohm resistors and whack a 1uF to 10uF 
>cap across them. This may not be practical in every circuit in a synth but 
>it does multiply the op-amp's ripple rejection out of sight. 


>Jeez I just can't help myself can I.

Until now I've thought that PMPO testing was entirely done in the marketing 
department. Thanks for sheding some light onto that.



uzs159 at uni-bonn.de


More information about the Synth-diy mailing list