Corsair CX600 ATX SMPS made by CWT problems, repair, need similar schematic

[lordvader88]

Corsair CX600 (600W) outsourced to CWT
not the V2 or 600M
http://www.overclock.net/forum/31-power-supplies/1431436-why-you-should-not-buy-corsair-cx430-500-600-a.html

I bought this used, for a little dual core PC (I have an EVGA SuperNOVA 850W G2 for my gaming rig). I was ripped off since it doesn't work....however now I get to have a go at fixing it  . I've never really worked on a SMPS before tho.

Google gives me near useless search results for stuff like this without published schematics. I'm pretty sure this is the original CX600, not the V2 or 600M, both of which were rated better. These things were only rated for 600W at 30 degrees C, so they failed fairly easily and some cheaper parts were used to meet the budget.

I don't want to just replace the caps at random, I want to find the problem.

Last fall I briefly checked around on it. It powers up, spins the fan for 1-2 seconds, and auto-shuts down. I didn't find anything obvious, maybe a few weak looking solder joints but I never retouched any. For the most part I think the diodes, BJTs and MOSFETs seemed ok unplugged, without knowing anything about the circuit really.

Then with it plugged in, and bad angle and lighting, started checking the PFC stage and shorted source to drain of a  K18A60V with the probe. And the fuse went right away, hopefully nothing else minded. So I have that here now. So I can replace that and hope it's back to where it was.


Can anyone point me to some modern, ATX PSUs with a PFC stage, and protection circuits that are fairly similar to big brand PC PSUs ?

Or any tips ? I'm not checking this with my scope yet when mains powered, until I know 100% where mains referenced stuff ends, and any high voltages are.

The main PWM driver chip is a ST9S429-PG14 . Its also in the CX430 (430W) and I just found this with most of the main parts listed, thats great, from a detailed review.
http://www.realhardtechx.com/index_archivos/Page17755.htm
http://www.realhardtechx.com/index_archivos/Page10506.htm

I'll drawing out the circuits again, check a few things again, and replace the MOSFET I zapped. Then start over with a SMPS troubleshooting guide I guess.

[Peeps]

You can Google atx psu schematic and the topologies will be pretty similar. There's a lot of big silicon parts in these so you should check them for shorts and report back. Theres a lot of capacitance and diodes scattered around so you may get false readings. Try to lookup the device you're testing and if you're really unsure remove it from the board and test.

Also since the fan spins I would assume the failure is on the low voltage side and not the 12V rail.

You could also watch each rail with a multimeter. Just probe a rail and turn it on and see if it reaches its specified voltage before it shuts off. This could help narrow it down.

[lordvader88]

I found out from here (http://www.overclockers.com/forums/archive/index.php/t-677256.html) that the SITRONIX ST9S429-PG14 director IC, is a rebranded Unisonic S3515
http://www.unisonic.com.tw/datasheet/S3515.pdf

I just had to dig up these datasheets again and some pictures. My c:\ drive died this morning. Lost all the PCB pictures I had all marked up with whats going where.

Got this SMPS troube shooting guide
https://readerspage.files.wordpress.com/2016/07/smpsrepairebook.pdf


The input/EMI seems fine, rectifier, main 400V cap is at 162.8V, it would be about 2x if it was working.

There's 2 daughter PCB boards on the Pri side. I don't know for sure yet which one has the CM6800, or whats on the other 1 . But both seem to start up and run off their power circuits.

The control voltage transformer primary winding is switching at 22.5kHz, thats coming right off 1 of the daughter boards, probably the IC itself but not sure what else might be on them). It's not using any transistor on the main PCB for that. So power is transferring over and powering the S3515 on the secondary side with 5V from that trans.

There's about a 1x1x1cm transformer, there's no voltage on it, don't what it's for yet.

The big transformer is not active. (there's 3 total on this PSU)

I've tried scoping the bases of the SMD BJTs that run the gates of the K18A60V, I didn't see anything, not over say 10kHz, I'll check some more soon.

The S3515 puts out the right Vref. I think the optoisolators work, never really saw anything on them when powering up. Nothing ever really appears at the 12V or 5V outputs, not on the DMM or oscilloscope's. I need to check on more scales tho.

Just about to dig in again, but without most my drawings. That's about what I got so far.

[floobydust]

The CX600 has been through a few revisions.
ST9S429 is a supervisor IC on the secondary side for OVP, OCP etc.
The active PFC done on a little PCB with Champion CM6800 which is common.

There's a few bits and pieces of schematics out there, just have to confirm main and daughter pcb markings; a pic helps.

Use caution and ensure the primary caps have discharged before working on the pcb.

[lordvader88]

For safety I'm using a current limiter 100W light bulb, and I drain the main cap with an LED.

Well After seeing nothing at all, with my little DSO138 scope from the CM6800 PCB, I pulled it. I see nothing like trying to start switching on any range/div.

I've applied 10.5V to the Vcc pin, and 2.5V to pin 15 VFB, I get nothing from the chip at all, and no Vref output. The PCB used 0.5mA. Vcc is on the gate of the 3M03X tri-MOSFET DIP.

There's a good internal diagram on page 5
http://www.bannerspan.com.tw/technical/Champion%20Micro/CM6800%20Design%20Guide.pdf

Unless they left something out shouldn't Vref be there ?




Back on the main PCB I could try more DC tests where I hooked 12VDC to the mains rectifier output, and try powering the mosfets with a square wave.


I'm not sure what the flyback transformer should be reading like. Right now it's like at lot of secondary side stuff is connected together with a multi tapped FYBT. There's only 100 or so Ohms between 12Vout and gnd. Is that normal ? I had just started mapping some of the sec. side when I lost my pictures and hardrive.

[floobydust]

Is it a 9 or 10 pin daughter board? The CM6800 does not sound happy. It should output +7.5Vref with 15V on Vcc and GND, alone.
The standby SMPS makes +5VSBY and typically powers the t.

The CM6800 also includes the main PWM section, so failures there... I always check the power semi's- AFPC and main drivers, as faults can work their way back to the IC.
So the CM6800 may be an open-circuit due to an overload.

Undervoltage lockout is around 13V, so it would like 15V. I would suspect a shorted mosfet or IGBT, and if so then their driver may be gone.

I'm not sure which transformer 12V you are referring to, the main 12V usually caps go bad there.

The CX600 may be similar to FSP OPS550-80GTLN, hard to tell since these OEM's are in bed with each other.
These schematics are all I can find, loosely drawn with the main sections.

[lordvader88]

The main PCB is a G08-CCX3160-PF00 rev1.8, the CM6800 PCB is TGD E351376 rev0.2 , it's 10pin and from 2013. Its similar to that 1 but not exactly. I fell asleep before taking a picture and mapping it out and checking the SMD parts.

No I meant I will input 12VDC after the mains recififer just to see if I can turn on the FETs. I'll do the same on the secondary just to see what pops out.

I have a MS5803 LCR meter now too, so I can check some caps too.

Well I better order a chip, at least they are cheap, hopefully they are 'real'

[casinada]

My Corsair CX600M stopped working around 6 months ago.  I replaced the Power supply with a new EVGA600BQ and put the Corsair in the pile of junk.
Today looking around found :


In short I soldered a 22K ohm SMD resistor in parallel with the existing one and the power supply is working again.
I'm glad I didn't have to spend the time and effort to find the problem 
I hope it helps somebody else with the same problem