Sujet : Re: power supply idea
De : jjSNIPlarkin (at) *nospam* highNONOlandtechnology.com (John Larkin)
Groupes : sci.electronics.designDate : 22. Apr 2024, 17:50:02
Autres entêtes
Organisation : Highland Tech
Message-ID : <el1d2jp7haikm58705nliicdgdbg4n6jhp@4ax.com>
References : 1 2
User-Agent : Forte Agent 3.1/32.783
On Mon, 22 Apr 2024 07:15:22 -0000 (UTC), piglet
<
erichpwagner@hotmail.com> wrote:
John Larkin <jjSNIPlarkin@highNONOlandtechnology.com> wrote:
If one had, say, a 48 volt power bus, you could hang a half-bridge
switcher to ground, and a lowpass filter out. If the drive has duty
cycle n, the output voltage is 48*n. So we have a programmable power
supply with no feedback, which will be stable into any load.
The load regulation will be mediocre, but we could almost sell it
as-is.
So now, sense the output voltage and compute the error against the
target, run through a slowish integrator, and tweak the PWM to get
zero output voltage error. Gross transient response is basically the
response of the output filter, with some modest drool from the
integrator.
We can constrain the influence range of the integrator, just enough to
give the regulation that we need. That limits output swing in case the
feedback is wrong, as one could get from a botched remote sense
connection.
https://www.dropbox.com/scl/fi/2fysyvkl4eim7vujhaobh/FFINT_PS_1.jpg?rlkey=rug6yi3cgemi9vvbz8apgboqi&raw=1
>
Looks like you have invented the buck converter.
I invented a control algorithm. All the buck chips that I know of are
all feedback driven, and will slam into either rail if the feedback
divider is broken. Blow things up.