Should I keep this sketchy old power supply my uncle gave me?

[Parkman29]

In addition to the many useful electronics stuff given to me from my uncle, I got this power supply that is completely unlabled.
I suspect it was a hobby project, and I do not dare to plug it in due to the obvious issues pictured in the link.
Repairing it myself sounds like an interesting project but is it even worth it?
https://imgur.com/a/pFBkKrz

[Excavatoree]

Everything looks labeled to me.  I don't quite understand what you mean.   I'd plug it into my protected Autotransformer, just in case, but it looks OK to me from the outside.  Looks like a usefull supply if it works.  It would be a good learning project to fix it if it doesn't.

[Parkman29]

Lol just opened it up, there is no way in heck im messing with this. The only real problem is that the power cord is ruined, and in order to replace it I need to remove the board. I feel like touching that massive capactor, even though its likely discharged, is a bad idea.

[themadhippy]

I'd plug it into my protected Autotransformer,

only after changing the mains lead .The bottom switch marked 15 32 is a  most likely a range switch,saves the regulator circuitry having to work too hard at lower voltages.personally after changing the mains flex id plug it in cross me fingers and turn it on.

[Excavatoree]

only after changing the mains lead .The bottom switch marked 15 32 is a  most likely a range switch,saves the regulator circuitry having to work too hard at lower voltages.personally after changing the mains flex id plug it in cross me fingers and turn it on.

I didn't see the mains lead.   Eh, I'd cut the dodgy one off and use a cheater if I was unsure, but I just don't see it as the OP sees it.  Is there something in the internal picture I'm not seeing?  It looks fine to me.

[coppercone2]

looks well built and reliable just needs TLC and I think a ground.

if you can get a smaller modern capacitor to remove it from contact with the transformer it would increase its quality by alot. I don't like how big the can is, but its probobly easy to downsize now.. capacitor tech improved

you will be surprised how much smaller a modern 4000uF 50V cap is............. you could even increase the voltage by alot and get a smaller size.

I would say that if you resize the components for spacing reasons, its probobly damn good.

[Parkman29]

Everyone is saying that it looks fine, so ill have a go at fixing it. Im inexperienced when it comes to dealing with large amounts of power so that big boi capacitor just scared me. There is a resistor across its leads so i dont think it has a charge.

[coppercone2]

OK well look at the size of a modern 4000uF 50V capacitor in dimensions, measure that one and compare.

its scary because its old and size inefficent (improved film quality).

and 50 volts is not that much for a cap.

A small lead acid battery is way more dangerous.

Like look on digikey at capacitors for that size. its HIGHLY miniaturized tech now.

[coppercone2]

I recommend getting one with screw terminals on it (might be bit bigger), probobly you want a 63 volt cap or bigger because 50v is kinda odd size. and anything from 3500 to probobly 8000 uF would be fine from 50-100 volt ratings.

something like this. you could probobly find one that fits better
https://www.newark.com/cornell-dubilier/500c402t100ba2b/aluminum-electrolytic-capacitor/dp/31P6804

the hard part of this repair is usually if you can't find screw terminal cap and you need to make a screw to solder wire adapter (you will end up with something flimsy) . Screw is better IMO. Keep the hardware type the same. Screw in caps are extremely reliable. Solder in caps are very weak in comparison, if you push it to the side with a slipped tool it will yank the wire out of the cap lol

[Parkman29]

if you can get a smaller modern capacitor to remove it from contact with the transformer it would increase its quality by alot. I don't like how big the can is, but its probobly easy to downsize now.. capacitor tech improved

you will be surprised how much smaller a modern 4000uF 50V cap is............. you could even increase the voltage by alot and get a smaller size.

lol just took it out, and its as light as a pencil

[coppercone2]

its basically mostly like toilet paper in there. if you take them apart they are anything between acid damaged brown paper to white pristine paper (after 70 years) on the good ones. Slightly dampened with chemicals.

[Ian.M]

OTOH the large can size of those old capacitors made them better at dissipating heat, so you need to be careful that a modern replacement has a high enough ripple current rating for the application, and may even have to wire several in parallel to get enough ripple current rating.

Rewire the power cord and test it under load, checking the ripple at that big capacitor's terminal, before considering a modern replacement.

The chassis should be grounded, and if it works, its worth adding a green binding post for ground to the front panel.  Don't hard-wire ground to any part of the existing secondary side circuit, as bench power supplies are most versatile if their output is normally floating.

[Parkman29]

I took out the old power cord and am putting a new one in with ground. The transformer, long wire coil (I assume its an inductor of some sort) are attached to the case, is there any concern that i shouldn't ground them as well?

[Parkman29]

Ok! I finished swapping out the cord, and it seems to be functional. I'm not sure if this waveform is healthy DC current tho

[Ian.M]

So you've got 3.2 divisions of ripple there.  Whether that's good or bad depends on how many volts/div your scope is set to (including any probe attenuator multiplier), and the load current you measured the ripple at.   The DC voltage is also of interest, though mostly to get an idea how much 'headroom' there is for the regulator at max. output voltage.

[p.larner]

Everyone is saying that it looks fine, so ill have a go at fixing it. Im inexperienced when it comes to dealing with large amounts of power so that big boi capacitor just scared me. There is a resistor across its leads so i dont think it has a charge.

have a go at fixing it !,you havent even tested it FFS!.

[Parkman29]

Yeah, looking at it further, it ripples about 6mv performing better than my amazon power supply and nearly as good as my HP 6284A, so I think its safe to say that its working very well. Thank you guys for helping me out!

[RoGeorge]

The oscilloscope pic shows about 7mVpp ripple (oscilloscope was on AC, 2mV/div, right?).
Looks like a very good adjustable power supply, congrats. 

Test the power supply with a DMM on DC Volts, and see if the output can be adjusted by turning the 'Volts' potentiometer.  Check the potentiometer is still OK, in the sens that the output varies smoothly while you turn it.  The voltmeter on the front panel should indicate about the same voltage as the DMM.

Your power supply seems to have two voltage ranges, from 0 to 15V, and from 0 to 32V, selectable from the slide button above the outputs.  Some schematics do not start from 0V, but from 2-3V minimum, so the two ranges might be in fact 3..15V and 3..32V, which is not a defect, that's how some power supplies are.

The 'Amps' potentiometer sets the limit for max allowed current, in case of a short circuit, or in case of a too demanding load, so you won't see any change in voltage by changing 'Amps' (when measuring with a DMM on DC Volts), unless you attach a load and make the current too small for that load.

Regarding current protection, some schematics will turn off the power completly when overcurrent occurs, others will start to generate constant current (CC) instead of constant voltage (CV), that's why the name CV-CC power supply.  Put the 'Amps' potentiometer somewhere in the first quarter of a turn, and short-circuit the output.  If it's a CC-CV power supply, then the 'DC VOLTS' indicator on the front panel will go to zero, while the 'DC AMPS' indicator will tell the max current set by the 'Amps' potentiometer.  If you turn the 'Amps' potentiometer while keeping the short-circuit, the 'DC AMPS' indicator should follow the potentiometer.

You may want to test the 60Hz ripple again, by powering some load while with the oscilloscope in parallel at the output, for just in case the big 4000uF/50V lost its capacity (but most probably it's still OK, that is not ESR sensitive capacitor).

[coppercone2]

Everyone is saying that it looks fine, so ill have a go at fixing it. Im inexperienced when it comes to dealing with large amounts of power so that big boi capacitor just scared me. There is a resistor across its leads so i dont think it has a charge.

have a go at fixing it !,you havent even tested it FFS!.

look at pics, the power cord was a very clear broke ass shock and possibly fire hazard~!

[coppercone2]

I think the capacitor basically touching the transformer is sketchy

I also don't like how its not clamped from the top. Unless its clamped really good it can slam into the front of the chassis and make a short. Won't be spectacular but its not great. I like a band that goes all the way around if something is mounted like that. unless there is stops I don't see. I think a modern smaller cap and a pipe clamp would solve those minor concerns.

ALso what you are interested in PSU is ripple under load, not ripple when its doing nothing. Load regulation is what its for. its not a voltage reference usually.  To get a true comparison compare both power supplies you have while its loaded down at the max rating. For signal circuits probobly a 100mA load is generous.

The other parameter is transient response. That means you put a pulsed load on it and measure with a scope to see how it does with 'square wave' loads. Some power supplies will make alot of noise (acoustic) when you do this test. Harder to test though, need a good switch. A good response allows you to test circuit prototypes with less filter capacitance if the wiring is satisfactory, since the power supply takes care of things. Can save time.


The first time you play with a pump, solenoid, motor, etc.. you will appreciate having a robust easy to fix simple supply.

as for signs that its a high quality supply:
1) brass hardware
2) bolts
3) cable harness that looks good
4) big capacitors
5) beefy switch
6) thick metal chassis
7) high quality potentiometers

they don't build them like they used to!

a sketchy supply has random wires connected to a board thats held in to a plastic case by little plastic clips, pot metal hardware with a puny switch. don't let the OLED display and green circuit boards fool you...


for a full restoration it looks like it could use new thermal grease, cleaning of interconnect hardware (but it honestly looks pretty good), check or replacement on all the caps, cleaning, potentiometer greasing and maybe a paint job. but it will probobly work for many years without doing nothing to it (and alot of those things are perilous)

the most important thing though is to make sure that it s safe and calibrated because the incorrect readings make it a liability not an asset


another thing to check would be to see if the circuit boards have electrolyte on them (may require cleaning to be sure) from old caps that is potential short circuit hazard. like icky grease.

and as always check to see if the correct fuse is installed because its a fire safety feature. some people will throw in a 20 amp fuse in there because it burned out and its all they had, making the fuse useless. and make sure all the connections are tight because there might be a loose wire or screw some where

[T3sl4co1l]

Safety -- it's fine, well constructed I would say.  Painted metal enclosure is rather nice.  The exposed transistor (regulator?) isn't a touch hazard, strictly speaking (SELV is below 70V DC), though you wouldn't want to accidentally drop metal parts on it -- I'd be happy with a metal mesh covering it.  If it's just the TO-3 "hot", you may even find a plastic cap to snap in place.

I would check if the enclosure is tied with circuit ground: earthing it is an option, of course you need a three-wire cord to do that, and could perhaps install an IEC receptacle if you're feeling up to a little metal cutting.  The reason to check is really just so you know how it's wired.  If tied, you can't float it anywhere with respect to other earthed equipment (e.g. oscilloscope ground).  If circuitry is adequate isolation distance and everything, lifting it from chassis ground is acceptable, in which case you get the protection of an earthed metal enclosure plus a floating output you can wire to anything.

And if internal ground is lifted from chassis, then you can install a 3rd binding post (green) and shorting bar, so you can earth through the power supply directly, when desired.

The capacitor may need reforming; I guess since you've applied power, that cat's out of the bag.  If the capacitor got warm while you were testing, maybe consider taking it up slowly on a variac (or series lightbulb).  It might also be dry, but for capacitors of that type, I'd say it's hit or miss; many survive for a long time.  In any case, replacement is easy; comparable values are cheap, and much more compact.

I'm guessing the circuit is very basic.  A typical project is an LM317 -- it would be the -K version in TO-3 here -- with a transformer, rectifier, filter cap, and pot to vary output voltage.  The meters are convenient, and the range switch means more current capacity at low voltages before the regulator overheats (LM317 is internally thermally protected so it's okay to overload, just don't do it for a long time, it can eventually fail).

Compared to other options -- well, it's one supply you don't have to buy, if that's value to you then there you have it.  It's only a single output, it's analog movement (though you could patch in digital meters if you prefer), so it's not wildly functional or versatile or accurate.  It should be pretty stable, and it will be low noise as long as the regulator is doing its job -- in particular, you don't have the plague of switching noise from those cheap Chinese modules flooding the market these days.  It's probably unidirectional, i.e. it can only source current, not sink it; it's not a power amplifier, downprogrammer or SMU.  Mind, this only matters when connecting multiple supplies together, or doing things like regenerative braking where current can flow back into the supply.

Tim

[Ian.M]

For anyone interpreting the text on the scope CRT:
"2mV            A=50us"
As the OP is in the USA we know the period of the ripple, assuming full wave rectification is 8.33 ms.   The period on the trace is 3.6 divisions, so it is likely that the timebase is set for 2 ms/div but the var. timebase control is not set to 'Cal' so its running a bit slower at 2.3 ms/div.  A=50us is obviously unrelated to the timebase time/div setting, so probably refers to a cursor of some sort relative to the timebase trigger point.   It seems likely that the '2mV' refers to the voltage at the cursor, *NOT* the Y AMP Volts/div, hence my question about the scope settings above.

[Parkman29]

So this morning i tested it a bit more. When the switch is in the 15 range, I can bring the voltage up to almost exactly 35 volts. In the 32 range, I can bring it up to nearly 40 Volts! As for the big capacitor, it seems to be working pretty good and I forgot to mention earlier that it is in a thick plastic casing to shield it from the transformer.

I apologize for the confusion around the ocilloscope, but the 2mv Is related to the volts per division. The A=50us is a timebase setting since it changes when I turn the timebase knob, but there are buttons at the top (I assume they enable different cursor settings for the z axis) that change it to B, or both.

I installed a ground post and can verify that the case is separate from circuit ground.

Again, thank you guys for helping out and dealing with me

[Ian.M]

That clarifies things slightly - your scope may have a dual timebase (A/B, which can be used to 'zoom in' on some time interval of a trace, but to be reasonably sure of its settings, and thus the amplitude of the ripple and the meaning of the onscreen text, we'd need a good photo of the whole front panel (and its make & model no. if not shown).

Have you done any testing under load yet?  For the 15V range a 12V 10W incandescent auto bulb would probably be suitable.  For the 30V range use two in series.