Epower EP-600NE-XV Shuts down intermittently

[Pentium4]

Re: Epower EP-600NE-XV Shuts down intermittently

You should be able to power up the '494 with an external supply. 12-18V. Use secondary common for the return. Look around for either the cathode of an isolating diode or the cathode of a dedicated aux "pick-off" rectifier, such as the one connected to the top of C21 (22u/16V) in that diagram. If you connect to the right place, the entire control section will run, but none of the outputs, including +12, will come up. So the PWM runs at full (minus dead time) duty cycle.

Then, you can scope the bases on the primary side safely, since it's unconnected. You should have nice symmetrical waveforms.

When scoping those base drives, you must put the scope tip/gnd, respectively, across the base/emitter leads. Pull the switchers, too- the B-E junction will pull the positive excursions down.

Or, you can try with a DMM set to AC volts. You're not interested such a meter's high-freq accuracy- trying for a relative measurement here. Both transistors' B-E drive voltages should be the same, and also add up- they're driven 180 degrees out of phase. So if each one gives ~1.2V, checking base-to-base should read ~2.4V. Again, that's just a quick test to check for any action coming from the driver.

You can also check with a "freq count," if your DMM can do so.

Thanks for the write up! The SDC2921 is good, so the search continues...
I may retin some of the legs on the main toroid. That thing is heavy! It weighs 135 grams. Is there a reason that the color on one side of both cores is faded? Also, there was no insulation between the bottom of the main toroid and the diodes under it. Maybe they were shorting out?

Wow, they do look quite goofy. I wonder if one of them went open-circuited. It's a rare failure mode for BJTs, but it does happen. That could also explain the fan "blip" you see on the output.
If you still have them out of circuit, check their B-E and B-C PN junctions... i.e. set multimeter on diode test mode, put red (+) probe on the base (B), and black (-) probe on emitter (E), and then collector (C). You should get a healthy diode voltage drop on both - probably around 500-700 mV (usually around 600 mV).
- OR -
If you have a crappy PSU with 13007 BJTs (or something equivalent), try swapping those into this PSU.

Another thing I would suggest you check - the 3.3V rail mag amp circuit. In the schematic diagram that Kaboom uploaded (200W PSU, "DTK PTP-2038"), that would be on the lower right of the picture - you should have a circuit similar to that in your PSU. Should be a small transistor (Q13 in the PTP-2038 schematic), a diode (D32), and a 431 shunt IC.

So far I've found nothing bad in the mag amp circuit. Still testing though, quite a bit of components and I'm careful so I'm slow The switchers tested good but I replaced them anyways with these beefy 18A switchers:

different counterfeit parts with different markings.
but why?
bjt's are too cheap to be worth all that.

I know....check out the underside, they look like they were manufactured differently:

[Pentium4]

Re: Epower EP-600NE-XV Shuts down intermittently

Maybe check for short circuit between rails? - i.e. between 3.3V rail and 5V, rail, 12V rail and 5V rail, and etc. Perhaps something could be shorted on the large toroid.

I pulled the toroid to look at the tinning. It all looked fined, but I did put some insulation on the bottom of it because it looked like it was very close to, or touching the diodes underneath it. I then replaced the 5V and 3.3V rectifiers. The whole supply came up for 2 seconds before shutting down again and going back to the fan tick symptom.

According to the datasheet Functional Block Diagram, "REM" (your PS-ON signal) only appears to be an input. So I'm not sure if it is supposed to latch down to low (0V) like that. Check the VCC voltage on the chip and see if it's still there after this happens.

Either I was wrong or it changed. The PS_ON wire is always getting 5.16V now. I scoped the VCC on powering up and the voltage goes right up and stays up. No matter what though, the PG wire never gets any voltage, under any circumstances, even when the PSU initially tries to power up. It gets as far as VCC on the outptut. All the input voltages on the controller are as they should be.
VCC on power up:

I need to check the rest of the components connected to PG. There are a lot of resistors and a couple TO-92 BJT's I need to test. The zeners tested good.