[colug-432] mostly dead switch — diagnostics for fun

Rick Hornsby richardjhornsby at gmail.com
Fri Jul 20 21:26:43 EDT 2018


On July 17, 2018 at 10:55:56, Jeff Frontz (jeff.frontz at gmail.com) wrote:

I'm not an EE, but it seems like your momentary "dead short" is a cap
charging (one perhaps intended to smooth out spikes/sags from cheap wall
warts).


And I think you want to measure your wall-wart's output with a load on it
(otherwise, it seems like the circuitry could be hunting around to figure
out what is really required downstream).

I was about to give up and chuck it when for some reason I started
wondering if the board’s problem was not enough current. My bench power
supply has a knob for current and one for voltage. When I turn the voltage
knob, the little display immediately reflects my change to the output
voltage.

When I turn the current knob, usually nothing happens. So I rarely mess
with it. Up until now, I’ve never had to. I know there’s such thing as
“constant current” but I’m not sure what it is and I don’t think I can
select between CC and CV, at least on this PSU.

I gradually turned up the current, and lo and behold the switch started
going through its startup diagnostics. It was pulling a little under 1A.
The switch’s power supply is rated for 2A. I plugged that busted power
supply back into the switch, and it is only supplying about 35mA — not
nearly enough current and far below its rating.

https://4.bp.blogspot.com/-MIZ_PDXfj9U/T6AJIuFal4I/AAAAAAAAAnU/RE6CdCdZnHc/s1600/theres-your-problem.jpg

Bottom line, power supply is toast but the switch is probably just fine.

To answer the other question, it always seemed like a hardware rather than
software problem. In either case, I have no idea what version of the Linux
kernel it might be running. The datasheet (which the mfg tries really hard
to keep anyone from getting) seems to say that the CPU is a SoC (Switch on
a Chip). I doubt it’s running any form of Linux as we know it.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.colug.net/pipermail/colug-432/attachments/20180720/0e0062f5/attachment.html 


More information about the colug-432 mailing list