Corsair SMPS repair.

[IvoS]

I have TX650 Corsair desktop power supply that I am trying to repair. This is really nice power supply with all Nippon Chemi-Con capacitors inside. I found one faulty diode and the PFC/PWM control chip dead.  The CM6800G integrated circuit is mounted on small separate board with lot of stuff on it. The schematic is as attached on the picture. I also found shattered C18 that is SMD and I have no idea what the value of C18 should be. By looking at the schematic, it looks like C18 and R43 form a low pass filter and also holds the reference voltage on pin6. Pin6 goes to collector of photo transistor. I don't know how critical this capacitor value is on the power supply performance as it may affect the switching response and noise? I need at least a ballpark value. Can someone experienced help? Thanks.

[uncle_bob]

Hi

Where do the pins (1,2,3,4...) off of that board go?

Bob

[IvoS]

Unfortunately, I don't have a full schematic, so I don't know.

[uncle_bob]

Unfortunately, I don't have a full schematic, so I don't know.

Hi

I would suggest tracing it out. It appears to be a loop compensation filter, but maybe not.

Bob

[tautech]

Be sure to check C16, the local bulk VCC cap that is normally supplied from both a resistor dropper from the HV supply and a flyback winding on the primary side of the transformer. Check flyback diodes too.

[IvoS]

I found some schematic. It is not exactly the same as the Corsair is designed, but basically using same parts and concept. So, pretty much the same.
The schematic can be seen in PDF file page15. You can see more stuff on the PCF/PWM control board, there also is a LM358 opamp doing something, but everything else is basically in the same connection. There are more passives added, as the CAP for example I am trying to find out its value.

[tautech]

C18 appears to be part of a snubber/smoothing arrangement so I'd try .1uF and larger.

[IvoS]

It is in a RC network of a VDC pin6, which is parallel to feedback of optotransistors. I thought the value should be much smaller, like 1nF, just to snubber the switching peeks off. But I can be wrong.

[tautech]

It is in a RC network of a VDC pin6, which is parallel to feedback of optotransistors. I thought the value should be much smaller, like 1nF, just to snubber the switching peeks off. But I can be wrong.

Try 1nF, i thought a larger value that are usually in series with a resistor.