Help finding the fault in my lab power supply

**Downunder35m** · 23-09-14, 12:49 PM

Recently my lab power supply decided to blow the caps on me and now that it is all cleaned up and the faulty caps replaced I need help.
Initially I thought all is working fine but when I connected a load the current limiter went on and showed 0.8A although set to the full 2.5A.
Voltage also went down to 18V and with the load removed continued to go down to pretty much exactly 10V.
At first I suspected a faulty BD137 as the cause but it turned out that the damn thing does not get more than the 10V to the base.
Ok, must be the OP-amp but after replacing it with an LM type that needed to be rewired the symptoms were still the same

Checked the remaining capacitors, the potentiometers and even the shunts but all are fine.
Here is the circuit:

What am I missing?
Or is it time to bin it?

**Uncle Fester** · 23-09-14, 09:07 PM

Check that all diodes in the bridge rectifier are good and it is not acting like a half wave rectifier under load?

Sounds though like a ground or other voltage reference is floating around especially at the OPs.
Could just be a worn soldering joint especially if this is a single sided PCB which is what I assume from a 20 year old PSup.
Just re-solder the whole lot, shouldn't take longer than 10 minutes. Often the crack is very hard to see

**Downunder35m** · 23-09-14, 10:17 PM

Rectifier was not checked as I get a pretty clean DC out of it, but then again that does not mean much....
Will replace it tomorrow with a new one.
Solder joints should be fine as they are done old school and quite solid, but while I'm at it I do them all, even if it is just for the good looks

I only get very little going into the op-amp but it reacts nicely to the adjustment of the voltage knob.
Just seems the thing is bild for the full range of amps and the voltage.
It always goes back dow to pretty much exactly 10V.

Will report after the solder job, thanks for the tip!

**Reschs** · 23-09-14, 11:15 PM

Having trouble reading the part number for the IC.
Can you post the Part Number and what substitute you used.

**mickstv** · 23-09-14, 11:27 PM

Originally Posted by Reschs

Having trouble reading the part number for the IC.
Can you post the Part Number and what substitute you used.

The IC is a B2761d.

**TVguy** · 23-09-14, 11:33 PM

Originally Posted by Downunder35m

Rectifier was not checked as I get a pretty clean DC out of it, but then again that does not mean much....
Will replace it tomorrow with a new one.

An oscilloscope is very handy for fault finding in power supplies (linear or switching). Ripple due to leaky diodes cannot be detected with a DVM and will produce apparently stable DC readings due to the integrating action of the meter.
The fact that the main DC reservoir capacitor failed points to the bridge rectifier being defective. As this is a closed loop system, all errors will be reflected throughout the circuit.
N31 & N32 are curious devices indeed (DVM199) and my best guess is that they are opto-couplers. IC2 is equally mysterious and would appear to be a stable voltage reference, probably in a TO92 package with one pin tied to another?
The resistor network on the emitters of the 2N3055s forms a wheatstone bridge feeding the error amplifier IC1. The low value resistors are critical for correct operation and a good low range Ohmmeter should be used for checking.
BTW, I picked this circuit as a European design immediately. Reminded me of Philips circuits from the good old days when Eindhoven was producing beautifully designed discrete power supplies. The 2N3055 is almost as ubiquitous as the NE555 timer.

Good luck and don't give up.

**Reschs** · 23-09-14, 11:41 PM

The first winding determines the maximum output voltage. What is the voltage at pin 16 on the circuitboard ?
The second windings are to power the opamp. Voltage accross C1 & C2 they should be equal. Say 15V ?
R14 and D4 & D5 reduce the negative rail to IC 1 by about 4V. So the supply to IC 1 would be something like Pin 3 +15V and Pin 7 -11V.

IC 2 appears to be a precission voltage reference.

Pin 8 on IC 1 controls the voltage and Pin 6 controls the current.

**Reschs** · 23-09-14, 11:44 PM

Originally Posted by TVguy

An oscilloscope is very handy for fault finding in power supplies (linear or switching). Ripple due to leaky diodes cannot be detected with a DVM and will produce apparently stable DC readings due to the integrating action of the meter.
The fact that the main DC reservoir capacitor failed points to the bridge rectifier being defective. As this is a closed loop system, all errors will be reflected throughout the circuit.
N31 & N32 are curious devices indeed (DVM199) and my best guess is that they are opto-couplers. IC2 is equally mysterious and would appear to be a stable voltage reference, probably in a TO92 package with one pin tied to another?
The resistor network on the emitters of the 2N3055s forms a wheatstone bridge feeding the error amplifier IC1. The low value resistors are critical for correct operation and a good low range Ohmmeter should be used for checking.
BTW, I picked this circuit as a European design immediately. Reminded me of Philips circuits from the good old days when Eindhoven was producing beautifully designed discrete power supplies. The 2N3055 is almost as ubiquitous as the NE555 timer.

Good luck and don't give up.

The DVM199 Devices are the Panel Meters.

**Downunder35m** · 24-09-14, 12:42 AM

Jepp, the DVM199 are just the meters in the front.
IC2 is for temp compensation and a PTC, the hotter it gets the higher the resistance.
Pin 16 delivers close to 44V and yes 15V for the op-amps.
The op-amp was mainly used in old DDR devices and after the fall of the wall lot of companies used leftover stock to produce.
Replaced it with a LM393 but as I said with the same result.

Will go throgh all of it again tomorrow and do some more voltage readings and check the suggested things as well.
Don't really think the main cap failed due to AC as the whole inside looked badly corroded, the aluminum had white oxide areas all over the can.
Old tech like this should not be allowed to fail LOL

**Reschs** · 24-09-14, 01:09 AM

The LM393 is not a suitable replacement for IC 1.

**TVguy** · 24-09-14, 06:41 AM

Originally Posted by Reschs

The DVM199 Devices are the Panel Meters.

Thanks, I should avoid posting when I've been awake for 20 hours.

**Downunder35m** · 24-09-14, 01:34 PM

Originally Posted by Reschs

The LM393 is not a suitable replacement for IC 1.

Why is it not suitable?
Both are dual op-amps, only a bit of rewiring was needed to match the pinouts.
Official replacement is the but
a) almost impossible to find in AU
b) quite overpriced, cheapest option incl. postage I could find was getting close to 20bucks...

**mickstv** · 24-09-14, 01:47 PM

You could try one of the Chinese ones on EBay, for less than $10

**Downunder35m** · 24-09-14, 03:12 PM

With 3-5 weeks for the delivery from past experience with small stuff from china I was hoping to rule out other possibilities first.
So you are saying if transistors and, resistors and diodes are fine it comes down to the op-amp?
Still don't really understand why I can't replace the damn thing with something more common that easier to obtain.

**Reschs** · 24-09-14, 03:41 PM

The LM393 is a dual voltage comparator and not a linear opamp.
Choose a dual opamp. eg LM358

**Downunder35m** · 24-09-14, 04:55 PM

Damn it!
Of course you are right, was searching through my electronic ribbish for hours to find one and must have messed up

Ok, back to searchig again , otherwise I'l have to order a replacement to try.

Voltage I measured so far seem to be fine:
44V between pin 15 and 16
15V for the supply for the op-amp
10V for the display
Resistors and shunts are fine too with very similar values
Rectifier diodes are fine too but replaced the thing anyway for a try with no improovement

Really seems to go down to me being blond and using the wrong parts LOL
If I find another op-amp in my boxes I will try again,otherwise order one in.