TEC SMPS Repair - Primary Side Advice, Schematic Attached -Resolved

[RadioAct]

Hello All,

I have a SMPS from a wine fridge and I am looking for some advice on the primary side. The power supply is run by a TL494 and using some neat feedback trick on the secondary side the output voltage runs from 8-12Vdc. This supply uses two transformers for the isolation, one is obviously a power transformer and the other is connected to control the transistors. The PWM chip lives on the secondary side and drives the feedback transformer. Attached is a schematic I made of the primary side.

When I got the supply, both pass transistors were blown along with the line fuse. Replace those parts and I plugged in the supply. It briefly came to life and quickly died. None of the parts I replaced popped plus I can not identify any other bad parts on the primary side. I tried some new transistors again just in case but this time plug in a no sign of life. I checked the HV supply and there is +/-164VDC and it looks nice and clean.

The current state:
I have disconnected the secondary side of the power transformer and attached two high power resistors as a dummy load. I am powering the secondary side from a bench supply and the PWM circuit is alive. I have tweaked the feedback voltage to make the switcher IC think the voltage is low and it is running at about 40% duty cycle driving the control transformer. Probing the primary side of the control transformer with no AC and I can see some resemblance of the PWM waveform but it is distorted, I suspect that is expected since its traversing a transformer.

I plugged in the AC hoping the circuit would come alive and drive power into my dummy load but no luck. The primary transistors do not switch even though the secondary side PWM up and running. Q6 appears to turn on and stays that way. I am not sure where to look from this point, I have extracted and tested most every component on the primary side but I can not find any fault. I don't have a very good understanding on how the transformers work in this topology, I was not expecting them to be connected together like the are. It's a different deal then the opto-isolators type feedback SMPS circuits that I have played with before.

Can anyone enlighten me how this circuit should work and/or suggest any likely culprits to test?

Thanks in advance
Ty

[SeanB]

Replace the 150k resistors R32 and R41, and check the diodes D12 D10 are not leaky. the resistors are the start up supply to get it running so the secondary will drive it. Also replace both C2 capacitors, do not bother to check them. You have checked C16 is close to its marked value.

[RadioAct]

Thank you for the suggestions.

I went through and replaced all the parts you mentioned but unfortunately I did not see any difference. Testing the parts out of circuit, everything was in spec. I have considered trying the bad auto mechanic methold and just start replacing parts until it works but there is not much left to replace. I am wondering if there is something wrong with the output from the transformer, not enough drive level etc.

I am going to try and capture the base drive signals at each transistor and see what I can see.

Ty

[RadioAct]

I probed the bases of each transistor and poked around at the various points on the transformer but I didn't see anything with significant amplitude. There is a small waveform that seems related to the pwm signal but it's very small in amplitude, under 500mv. It could be my probing technique, it's hard to find a spot to ground a scope on an ac circuit.


I did have a bit of luck and I got my hands on a nearly identical power supply with a different issue on the secondary side. The secondary PWM circuit has some bad caps and the feedback was not working right. I was able to bypass/replace all that and get it working. I tried some comparison between the two and the one thing I noticed is on the good supply when plugged in the primary transistors would self oscillate to generate some bootstrap power to the secondary side. It was generating out 5v on the 12v rail but that's enough to get the PWM to kick in and take over. On my original board when I plug it in I do not see any self oscillation. There is no bootstrap power on the secondary side. Given my original testing methold I was powering the PWM from a bench supply so even with the lack of bootstrap supply, the primary side should have started to switch.

My question now is what causes the primary side to oscillate before the PWM kicks in? Is it related to the construction of the transformers? If so then could the transformers be damaged/shorted and this be causing all my issues?

Thanks
Ty

[RadioAct]

Another update:

On a whim I changed C17 and now all of a sudden the primary side is now self oscillating. Apparently that cap across the transformer coil is important and was damaged.

Now the supply is making a slight clicking sound but that's progress. A quick blind test and it was only producing 0.5V on the secondary side, not enough to get the PWM circuit running so next I will try to track down the source of the clicking sound.

Ty

[RadioAct]

Ok repair complete, board working now. Here are the details for future generations...

First the noise was coming from the main transformer due to the switching frequency running way to low due to only one transistor switching.
Some poking around and I stumbled across an intermittent base drive resistor R35, poke it and it measures up, poke it again and its open circuit. Ripped both base resistors off and replaced with new ones and the board came to life. Popped it back in and the TEC is working again.

Moral of the repair, poke it and test it again.

-Ty