Samlex SEC-14 linear power supply 13.8VDC 10A repair

[redwire]

Repairing a Samlex SEC-14 linear power supply, rated 13.8VDC "10-14A". Hong Kong company.

Dumpster rescue, main pc board is RPS-1203 with two filter cap pc boards labelled RPS1220. The same control board is used in many of their models.
The big 4,700uF Hangda electrolytics all took the piss, but no bulging as the bung has a vent.

Inside I expected an LM723 but instead is a few transistors, couple 2N3055's and overall low tech. It's the cheapest power transformer, just loose windings no tape, a bit bigger than a MOT. No safety approvals. PC boards were riveted in.
I drew a schematic and I'm not sure if this is worth five new caps ($15). No current limiting, so I can't use it for charging batteries and realistically it's good for ~6A continous due to the tiny heatsink.

Part of me want to bin it. Any ideas? Or maybe I could get some low cost caps and use a TL431 or add current-limiting. Hmmm.

[eccerr0r]

At least it has short circuit protection...

Also it should have fairly decent regulation (instead of just having a simple emitter follower regulator. No op amps or LM723 but it does appear to have worked around it well enough. No need to use a TL431 unless the zener broke.

Adding current limiting would be good if the short circuit protection isn't sufficient.

[eccerr0r]

IMHO:
The DNP capacitor should take the other 4.7mF capacitor on the control 22VDC rail, and that now depopulated 4.7mF capacitor can safely go down to 2.2mF. Perhaps this would be slightly cheaper to recap this way...

Then again how bad are the caps barring the ones that spilled their guts? Capacitance? ESR?

Wait...what the ... ? Are you sticking in low ESR 105°C caps in this? (Mouser seems to be more like $1.5/pop for bottom of barrel 60Hz 85°C GP caps, so less than $10 to recap the 5 caps...?)

[redwire]

The power supply works- the capacitors are still alive. Not sure why it got thrown out, maybe it just stunk up the place I wouldn't run those caps, worried they would vent.
Here are the capacitance and ESR readings. Note the AN8008 DMM is crap for testing electrolytic capacitors, the numbers are always useless from that Hycon single chip set- if there is any leakage current and these top out at maybe 1-2MEG ohms, it reads way high. Other numbers from Anatek Blue ESR meter and Amrel LCR meter, both of which I trust. Dissipation factor was high or infinity on most.

The PSU has no current-limiting or short-circuit protection, no battery-charging protection diode, and I don't like any of that.
It also has some strange AC mains fail/UVLO logic with the extra transistor which seems to latch off unless AC pulses are seen- but the diodes look wrong.

Nowadays, you'd put in an RD Eng module.
With the pandemic, my rule is to only use what parts I have lying around in the junkpile.
I have a $5 0-30V 2mA-3A chinese kit power supply board, which ain't the greatest but if it can run on a center-tapped transformer, I might mod it to go to 10A and plop it in. I really like adjustable output current.

I looked at new quality 4,700uF 35V caps (no need for low ESR here) Digikey is all around $CDN $3.00 each. From Aliexpress, including shipping around $1 each and like going to the casino.

[eccerr0r]

Yes it does have short circuit protection, you drew it. The "UVLO logic" indeed is short circuit protection. You probably can easily use that circuitry or even repurpose the components there to build a foldback current limiting, but yeah it will be some work if you want adjustable current limiting.

And ah... $CDN funny money... now that makes sense... though you probably can get by with quality 25V capacitors.