An old dinosaur bench top power supply - help needed

[Kg74]

I have had this old bench power supply which I have decided to look into repairing. I have no idea where it came from.

Attached is a picture. I cannot find much information on it to even figure out what it is rated for. The analog display goes to 25V, but the voltage regulator chip is rated for 5V (UA7805 - KC - F 8221 - Korea). The cap is rated a 100uf 50V. There is noting on the bottom but some traces. Pretty much what is seen is what is there.

Any insights into testing? Capability?

[R_J]

The control is used to vary the ground of the 7805 so the output voltage will vary from +5v to something higher 12~20v
There looks like a burnt resistor above the switch? or is it just a black shadow?
Something like this: The component values are different but it shows the general idea of how it works.

[Kg74]

Thanks RJ - sounds like it may be worth some attention. I measured the transformer output and it is 25V a/c. Diodes seem to check out.

Any suggestions on testing the UA7805? Can it be done on board or should I remove it?

[redwire]

It was a trick to make a fixed voltage regulator IC adjustable by putting a pot in series with its GND pin. So you could get 5-30V depending on the transformer voltage. It has a 50V filter cap, IC is 35V in max. Similar schematic below.

For the size of this, it needs much better heatsinking something I would add to the back panel. It won't make 1A for long without heating up too much.
If the IC has failed, I would convert it to use a LM317 on a heatsink at the back.

Why is the resistor burned on the brown wire?

[R_J]

Originally Posted by Kg74

Thanks RJ - sounds like it may be worth some attention. I measured the transformer output and it is 25V a/c. Diodes seem to check out.

Any suggestions on testing the UA7805? Can it be done on board or should I remove it?

Just measure the voltages, the regulator should have around +35v input and the output should vary depending on the position of the control, (you can also check the ground pin voltage on the ic) The output of the regulator should ajust from +5v to around +30v If it does not, the regulator ic is likely bad.
The burnt resistor looks like it is the current limiter for the led?

[Kg74]

Thanks redwire and thanks again RJ.

RJ - The regulator reads about 30V on one leg and nothing on the other, regardless of where the pot is set. So the 7805 is dead.

The browned resistor is for an LED and is working.

I will look into the LM317 conversion - Redwire I assume you are talking a regulator with three legs vs this current through the board 2 leg setup. Any recommendation of wire gauge? I figure I could pull the 7805 and solder jumper wires in the existing holes and run it to the back per your suggestion.

[R_J]

Do you measure any voltage on the output? If you get 0 volts the regulator must be open internally

[Kg74]

RJ- No voltage on the output.

I put the ground of my VOM on the heatsink and poked the probe under the board. One leg read 30V and the other gave no reading.

[R_J]

It could be that the ground circuit is open, check the resistance between the large capacitor negative and the regulator common(can/heatsink) then vary the control and see what resistance you measure. The pot might be open?

[Kg74]

Would I be able to use this eBay board? https://www.ebay.com/itm/LM317-Step-...4AAOSw-itXrY-G

Take out the existing board, mount the eBay board to align with the existing pot control, cut a hole for the LED Voltage display, remove onboard LED power light and run jumper wires to the panel and keep the analog scale and wire it somehow?

Turn the dinosaur into something modern, but keep the old school look and gain some function.

[Kg74]

RJ - I performed the test. VOM negative to heatsink and VOM positive to negative leg of the gray vertical 100uf 50v cap. As I turned the pot the resistance would increase and decrease.

I did the same for the really big 25000MFD cap and the resistance would vary, but not in a linear fashion.

[R_J]

Yes, that would work, I would remove the old analog meter and install the led readout in the hole.

[Kg74]

RJ - Thanks again. I will place the order. It will be weeks before it shows up.

Any idea how to get the analog meter to also work?

[R_J]

The analog meter is just connected across the output, so as long as it is still connected there it will still work. So there is no problem using both.

[Kg74]

Excellent. Thanks again for your help RJ. I will circle back in a few weeks when the board comes. Appreciate your and redwire's answers and guidance.

[stj]

that's actually a 3a or 5a regulator,
7805K

but heatsinking it to the case will be tricky because you need an insulator-kit AND a plastic cover to stop the case of the regulator touching anything on the table and linking to the chassis - instant 5v!

[eccerr0r]

7805's are no more than 2A or so whether it's TO-220 or TO-3, just determines how much heat it can dump. Swapping and rewiring for a LM317 is a good idea but keep in mind that you will still heat thermal limits fast with lower voltage outputs.

This looks like a homemade versus a mass produced PSU? Not sure.

oh boy that ebay ad pisses me off, do not say "step down" and especially not "buck" with linear regulators...aaah... (though marking a SMPS with "LM317" as search bait is equally annoying.)

[stj]

https://www.cricklewoodelectronics.com/78P05.html

i actually have 3a and 5a 7805's in one of my old parts drawers.
dont ask how big the heatsinks were on them back in the mid 80's!!

[eccerr0r]

Nonstandard 7805...

Just like LM317T vs LM317K, the "K" version is no more current - they want you to buy LM350's and LM338's ...

[sam_sam_sam]

Originally Posted by eccerr0r

Nonstandard 7805...

Just like LM317T vs LM317K, the "K" version is no more current - they want you to buy LM350's and LM338's ...

Your are right but I think that they can dissipate the heat better than the regular lm317 or the lm7805 could at least that what I found