Power Supply Musings

[Ray Cuellar]

Paul,
As I read your reply, I paused to take a quick look at the temp. rating on
"the big cap" of my PPG's power supply and you were right on the nose, 85C!

Thanks for the info!
Messages such as yours are why I love this group so much.

- Ray
-----Original Message-----
From: Paul Schreiber <synth1 at airmail.net>
To: 'Analogue Heaven' <analogue at hyperreal.org>; Caloroso, Michael E
<CalorosoME at corning.com>; Ray Cuellar <nsa at pacbell.net>
Cc: synthDIY <synth-diy at mailhost.bpa.nl>
Date: Tuesday, June 08, 1999 9:52 PM
Subject: Power Supply Musings


>>What are your thoughts on increasing the 7805 (1Amp) with a 78H05 (5Amp)?
>>Would this help or would it cause more problems?
>>
>
>I bet not!
>
>Here are my observations and opinions based on 28 years of messing with
this
>stuff:
>
>1) Designers have a habit of breadboarding and testing their prototypes
with
>a $3800 HP lab-grade bench
>supply (0.005% line/load regulation, <1mv ripple). "Works great, boss! Ship
>it!"
>
>2) The synth power supply is the *LAST* thing designed. "Hey Fred! Let's
use
>all those transformers
>from that last synth!" They should be OK!" Ha Ha HA!
>
>3) No one wants to do it: "boring", "low tech". But worse, no one knows HOW
>to do it!
>
>4) There are very few analog designers that REALLY understand power
>supplies. Things like a Schade curve
>(in order to pick the proper rectifier/capacitor combination, which was
>published in 1943 and *long* forgotten
>by most people), proper heat sinks (I got a news flash: bolting a regulator
>to the painted chassis is NOT a good idea!!),
>and understanding basic things like in the US the AC line voltage varies
>from 95VAC to 132VAC and you'd
>better take care of BOTH cases properly.
>
>5) News flash #2: the 78xxx and 79xxx are NOT GOOD ENOUGH without *proper
>and careful componen selection*
>to get you "studio quality" SNRs. You can get OK SNR, but not -90dB SNRs.
>They are too noisy. The ripple is too high.
>The line/load regulation is crappy. IF you have $$$ caps, a star ground,
and
>protection diodes you can get a "decent"
>supply. Has this ever occured in a synth? Not to my knowledge. The WORST
>supply in a synth I've seen is in the OB-8.
>
>6) Another factor: capacitor aging. Capacitors age over time. If your synth
>is over 15 years old, I'd STRONGLY SUGGEST
>replacing the big caps on your supply. Chances are, they are cheap
85C-rated
>caps. Get 105C, low ESR (effective
>series resistance) caps. Mouser and Digikey in the US are wonderful
sources.
>The cost will be $10 or less. Get a value
>equal to OR LARGER for your cap package (the case sizes  are standard,
>thankfully!!).
>
>7) Lastly: it's amazing (but not *surprising*) that National Semiconductor
>has published page after page of 78xxx regulator
>app notes and the synth designers ignored them TO SAVE 25 CENTS! You can
get
>PDF files of these, and then
>add the diodes and change the caps and beef up your rectifiers like it
>SHOULD have been done. You can unbolt the
>regulators from the black painted chassis (the black paint is a WONDERFUL
>heat insulator! This just bakes the regulators into toast)
>and spend a whopping 75 cents and get a PROPER heatsink that will drop the
>temperature by half (note that the HEAT
>is the same. As we say in EE101 "power is power" meaning the wattage
>dissapated is the same if you stuck it in
>liquid nitrogen, but the TEMPERATURE can drop!)
>
>Paul Schreiber
>Synthesis Technology
>www.synthtech.com  <<synth supplies done right
>
>
>