[sdiy] Power Supply Start-up Issues

Tom Corbitt tom.corbitt at gmail.com
Thu Aug 27 04:11:30 CEST 2020


What if you added a switching circuit element (pick your favorite
relay/switch/fet/thyratron/etc.) that keeps the negative power from being
applied to the negative regulator until the positive rail has come up? I
don't see anything in the overall circuit that would be super unhappy with
that change at first glance.

If you wanted to take it up a notch, you could build in a zero crossing
detector and some control logic that would only allow the power to be
applied when it sees the AC input signal is heading positive. A quick
google search pops up a bunch of low cost chips that would be a good
starting point.

Just some shooting from the hip ideas,
Tom C


On Wed, Aug 26, 2020 at 7:26 AM ackolonges fds <ackolonges at hotmail.com>
wrote:

> Hi All,
>
> I've been having issues with a very standard +/- 12V power supply design
> for a modular synth. Every once in a while when powering up, the positive
> rail doesn't rise, and it needs to be reset to work. After some trial and
> error, and research, I found mention of this issue in Douglas Self's 'Small
> Signal Audio Design' book, where he shows a circuit almost identical to my
> power supply. A screen grab of the related page can be found here:
> https://pasteboard.co/Jo8PuIu.png
>
> This paragraph from that page describes the issue I'm having perfectly :
>
> *"**D2, D4 will prevent damage in this case, but leave the power supply
> vulnerable to start-up problems; if its output is being pulled down by the
> 15 V regulator, the 15 V regulator may refuse to start. This is actually a
> very dangerous situation, because it is quite easy to come up with a
> circuit where start-up will only fail one time in 20 or more, the incidence
> being apparently completely random, but presumably controlled by the exact
> point in the AC mains cycle where the supply is switched on, and other
> variables such as temperature, the residual charge left on the reservoir
> capacitors, and the phase of the moon. If even one start-up failure event
> is overlooked or dismissed as unimportant. then there is likely to be
> serious grief further down the line. Every power supply start-up failure
> must be taken seriously.**"*
>
> Unfortunately after stressing how important it is to fix such an issue, no
> solutions are mentioned...  I've tried removing all of the protection
> diodes, increasing and decreasing the value of the smoothing caps, tried
> many different makes of voltage regulators, and also tried an NTC
> thermistor to slow the in-rush current, all to no avail.
>
> I'm starting to run out of ideas, so if anyone on the list has come across
> a fix it would be great to hear about it.
>
>
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20200826/77e68e7f/attachment.htm>


More information about the Synth-diy mailing list