[ARC5] Troubleshooting Tips from the Late Nick Broline, #2

[David Stinson]

Back in 2021, Nick and I were experimenting with
those little yellow "0-400VDC" power supplies you
see on Amazon and Ebay.
There has been some recent discussion concerning them,
so those with an interest might find Nick's preliminary
tests interesting.
----------------------------------

I had ordered 4 of the yellow colored 45-395V boost converters a year 
ago, then promptly misplaced them. Took three days to find!
(Lord, I know that feeling. DS)

I put the supply feeding a 20W wire-wound w 3.5K load. Ran it off a 13.5 
VDC fixed output enclosed Lambda enclosed supply.

That looks like the following:

@180V Io = 51ma. W = 9.25W

@220V Io = 63ma W = 13.8W

@250V Io = 71 ma W = 17.9W

What I found was that the 180V performance was bulletproof into the R 
load. It the switcher could start with that resistive load, and was 
happy connecting the load and re-applying it while running. This would 
be perfect for an ARC-5 or ARC-12 receiver under all circumstances.
The 220 output was about the same.

The 250V output handled the load well, but when I tried to start the 
supply from the 12V source and the switcher loaded it would ramp up to 
~180V and stop. Disconnecting the load, then reapplying it, would let 
the supply recover to 250V output. This told me the supply would run a 
receiver well, but it you interrupted the source with the filament hot 
(and operating on 12V), the supply wouldn't come up all the way.
Switching to a higher input voltage, like 24V, seemed to work perfectly.

The switcher IC is a TI UC3843A which is a controller that charges an 
inductor until the current reaches a value specified by the voltage 
sense circuit. It make a current measurement every cycle. What I'm 
thinking is that when starting under load the chip runs into its current 
limit on each cycle and can't deliver any more energy into the inductor 
(and load, of course), and the control loop can't coax any more output. 
Looks like the converter gets "stuck." I don't know if anything is 
getting hot in the supply.

I suppose this says if I want to use the converter at a higher output 
voltage and current, but still within its limits, it has to be with a 
higher source voltage to get it to start under load, or needs to always 
start working into a low current load.

Next time I get back to the bench I'm going to look at the converter 
with a scope and see if its bumping into another "chip" limitation.

I also noted that the output of the supply overshoots when the 71ma load 
is removed, and I think that's because they are using a very large 
output cap (100 ufd, I think) for the application. I seem to recall the 
normal output cap is a large ceramic, like maybe 2uFd.??????  memory 
again.....but its been ~25 years since i dealt with these things.

73
Nick Broline


-- 
This email has been checked for viruses by Avast antivirus software.
www.avast.com