Sujet : Re: dc/dc startup
De : alien (at) *nospam* comet.invalid (Jan Panteltje)
Groupes : sci.electronics.designDate : 14. Jan 2025, 20:47:16
Autres entêtes
Message-ID : <vm6f05$3eo0$1@solani.org>
References : 1
User-Agent : NewsFleX-1.5.7.5 (Linux-5.15.32-v7l+)
On a sunny day (Tue, 14 Jan 2025 07:09:13 -0800) it happened john larkin
<
JL@gct.com> wrote in <
v5vcojt090tpv5h9a3ancgm1r7pc4p6bjc@4ax.com>:
>
https://www.dropbox.com/scl/fi/3hyvmc3mkvhrk6rc8dt72/P970_DC-DC.jpg?rlkey=npgerceq4t191hnjfd0907v6u&raw=1
>
That makes two isolated 56-volt power supplies. But the problem is
how to start it up.
>
I could just slam on the square wave drives and let the TI chip
current limit until the caps charge up. That might work.
>
I was thiking I could start the two half-bridges with in-phase square
waves and then slowly slide them out to 180 degrees. My FPGA kids
could do that. That would be a nuisance to Spice.
>
We did a similar supply a while back with discrete fets, no current
limits. Startup made 300 amp spikes that made all sorts of problems.
Just drive the thing from some Microchip PIC and use cycle by cycle current limiting
feedback via curent transformer.
No weard chip needed.
Why full bridge?
One power trahsistor should do.