2 Simples power supply 110VAC to 36VDC fixing help, both with same failure

[Axxel]

Hi friends!

I'm trying to fix those 2 power supply from 110VAC to 36VDC.

Both of those power supplies had a problem that I fixed, one has the PWM driver and his mosfet dead, I changed them.

The other one had the rectifier input bridge from the primary was dead and I replaced it.

And the two power supplies gives me this result:

Downloas video -> https://gofile.io/d/m8LRO8

You can see the failure, it looks it want to start but it "stops"... I can hear a little noise when it cuts in the transformer.

I replaced the PC817 optocouplers too it doesn't fix. I replaced the LM431 voltages references, same.

When I quit the PC817, the protection test lamp turns on, the power supply do the same but with a lot more power (That's why I think the feedback side is correct?)

I checked all diodes and resistors, everything looks to be okay...

More info:

The driver is the CR6842
The mosfet is the 16N60
The schottky rectifiers diodes are MBRF202000CT
The optocoupler is the PC817
The voltage reference is the LM431



What I have forgotten to see or what do you think those power supplies problems are? Hope they can help me, sometimes I had this problem on power supplies and I never managed to fix it...

[tunk]

I think one of the first rules of thumb, is to check the capacitors.

[Axxel]

I think one of the first rules of thumb, is to check the capacitors.

I tested them on the output of the power supply I have 2000uF same on both. The output power discharge resistor of 2K is OK and not shorted.

The main output of the power supply looks not shorted...

[madires]

Don't forget the small electrolytic cap near the controller IC.

[Axxel]

Don't forget the small electrolytic cap near the controller IC. When I measure the voltage at the power pins of the IC it's not stable it oscillate between 11V and 15V at the rythm of the flashing led output. (I can do a video of the multimeter measuring it if it helps)...

I changed it it doesn't fix. One thing that could be interesting, the voltage on the power pins of the controller IC is not stable. It gos from 11V to 15V synchronously with the low dimming led output.

I think one of the first rules of thumb, is to check the capacitors.

I test with others, it doesn't fix.

[madires]

At power-on the controller IC is powered by the rectified mains via a high value resistor and then by the auxiliary winding. If everything on the primary side is fine, then check the diodes at the secondary side.

[Axxel]

At power-on the controller IC is powered by the rectified mains via a high value resistor and then by the auxiliary winding. If everything on the primary side is fine, then check the diodes at the secondary side.

The schottky rectifiers diodes MBRF202000CT looks to be okay, not shorted and multimeter in diode mode gives me a umbral value of 0.2V approx in each par which looks a correct value for a schottky.

[madires]

I forgot that you already checked all diodes. Let's go back to square one. What tools do you have (besides the DMM)?

[Axxel]

I forgot that you already checked all diodes. Let's go back to square one. What tools do you have (besides the DMM)?

Analog Osciloscope, regulable DC homemade 0...5A 0...30V bench single power supply with current limitation, RLC Meter 

[madires]

Looking at the images of the PCB I see a wire bridge at the mains input, but no fuse. Was there a fuse? And have you replaced it with the wire bridge?

[Axxel]

Looking at the images of the PCB I see a wire bridge at the mains input, but no fuse. Was there a fuse? And have you replaced it with the wire bridge?

I don't have fuse like them so I bridged directly. But to protect I use a serial incandescence protection lamp. Obviously if I get to fix the power supply, I will buy original fuses.

[madires]

I see. Now, measure carefully (mains!!!) the voltage at the filter caps of the primary side. Since it's rectified mains it should be sqrt(2) * V_mains.

[Axxel]

I see. Now, measure carefully (mains!!!) the voltage at the filter caps of the primary side. Since it's rectified mains it should be sqrt(2) * V_mains.

I have 177VDC Stable on both.

(I have 127 VAC 60Hz input - I live in Mexico)

[madires]

Do you know how to set up two scope channels as differential probes?

[Axxel]

I don't have differential probe...

[madires]

I think we should stop here for safety reasons. It would be better to ask someone more experienced to repair the SMPSUs.

[Axxel]

I think we should stop here for safety reasons. It would be better to ask someone more experienced to repair the SMPSUs.

I'm there to learn, I bought new 36VDC power supplies to fix the problem I got and I fixed from long ago... but the objective there is really to learn HOW to fix it by myself, NO FIXING them paying it...

What de you want to measure with a differential probe?

[madires]

In that case we shall continue. The next step is to check the controller's VDD fed by the start-up resistor and the auxiliary winding. To do that we use the scope with differential probes. But before that we set up two channels as differential probes and check the setup with a low voltage, e.g. a 9V battery. Take two probes and remove the ground clips. Set both channels to the same Volts/Div, invert the second channel and then add the channels. Now you should have two probes without ground clips and one trace on the scope. The first channel/probe is plus, and the second one minus. Connect the probes (just the tips) with the battery and check if you see the voltage. If you swap the probes the voltage should be inverted too.

[Ian.M]

Your photos aren't good enough to see what the part number of the controller chip on the primary side of the PSU is. I assume its the socketed DIP-8 IC.

You'll need its datasheet to identify its pins and, if you don't have a schematic for the PSU,  the typical application circuit for it.   Print out a large copy of the typical application circuit (I suggest A4 landscape) and compare your PSU to it.  Its quite likely to be fairly similar to the typical application circuit, and you should be able to annotate it with the differences and eventually draw up a reasonably accurate schematic, concentrating on the primary side, and the secondary side feedback network to the optocoupler.   

Once you have a reasonable idea of the circuit layout, you'll be able to make measurements round the control chip to determine why its repeatedly shutting down and restarting.   e.g. Is it getting its sustaining supply from a primary side auxillary winding? Is the feedback signal driving it into shutdown?  Is it detecting excessive switch current after startup?

Personally I wouldn't consider working on a line powered SMPSU that cold checks and 'shotgunning' part replacement haven't fixed, without an isolating transformer, and preferably an appropriately rated isolated differential scope probe, as you are one slip of the hand away from a very nasty electric shock.  Even with all the right safety gear, I strongly recommend soldering small loops of wire track side you can clip the probes to, so you can set up each test with the power off, and *NOT* have to handle the board while making measurements.

The SMPSU section of 'Silicon' Sam's Repair FAQ is well worth your time to read: https://www.repairfaq.org/sam/smpsfaq.htm