PowerMust 600 dead

[alindumitru46]

I have this Mustek PowerMust 600 UPS for years, received as defective.
Now, waiting for some parts to repair motherboards/videocards already presented, I decided to seek competent help here to fix it.
The fuse was blowed but I replaced with a similar one.
UPS battery was drained, but I replaced it with a good battery, charged to 12V.
I inspected visually the board and there are no over hot traces of circuits or swollen capacitors; verifying CET / CEP50N06-680FZ not look shorting.
Unfortunately it does not start at all. No beeps, no led light. No communications of the UPS with PC.
I did measurements on CET mosfets and found 2.87V for top3 and ~ 7V for last mosfet(gate to drain, gate to source). The LM 317T MOSFET does not look voltages at all.
Any ideas?

[momaka]

Re: PowerMust 600 dead

The LM 317T MOSFET does not look voltages at all.

That LM317 is a linear voltage regulator, NOT a MOSFET.
Pin 1 = Adjust
Pin 2 = Output Voltage
Pin 3 = Input Voltage

Check datasheet for details. That LM317 is perhaps used to power the logic circuits? And with no voltage on the input pin, maybe that's why the UPS is dead?
Try tracing the input and output of that LM317 and see where they go to. Perhaps that might give us a clue. As would pictures of the solder side so we can help you trace stuff.

Lastly. check the small TO-92 transistors. When severely overloaded, they can sometimes go partially or completely open-circuit without looking burned at all. So definitely check the Base-Collector and Base-Emitter voltage drops with a multimeter set to continuity/diode test. Do this with battery and power DISCONNECTED from the UPS.

[alindumitru46]

Re: PowerMust 600 dead

I checked all CET-CEP50N06-680FZ for continuity/diode test, I think all are ok.
And the transistors seems to be ok, no continuity. I checked and all diodes, additionally, looks in good condition.
The values from spreadsheet are measured by diode test.

[momaka]

Re: PowerMust 600 dead

Like I said, start with the LM317. That is a linear regulator and likely used to provide a clean voltage for the logic circuitry. Pin 3 is the input voltage and Pin 2 is the output. First see where the input voltage is supposed to come from and then see where the output goes to.

As for the 2SC1815 TO-92 transistors, those are NPN BJTs. Here's how to test them:
1) Set multimeter to diode test
2) Put red (+) multimeter probe on Base pin (pin 1)
3) When you put black (-) multimeter probe on Emitter pin (pin 3), you should get a reading of about 600 to 800 mV (or 0.6 to 0.8V depending on multimeter models)
4) When you put black (-) multimeter probe on Collector pin (pin 2), you should get a reading of about 600 to 800 mV (or 0.6 to 0.8V depending on multimeter models)

For PNP transistors, simply reverse the multimeter probes and repeat the same steps above.

That said, are you sure that "MPS3 097A" is not a MPS3904? Because if so, that would be another NPN transistor.

Finally, the "MC19L 05ACP"... check the number again. It could be a MC79L05ACP. The "79L05" is a common -5V regulator. Comes in various cases - from TO-220 to TO-92.

Oh, and check all larger resistors for open-circuit or higher than 5% tolerance. Some may be fusible, so they will go O/C when overloaded severely.

[alindumitru46]

Re: PowerMust 600 dead

The only correction is for MC79L, otherwise everything is exactly as I wrote on each row.
I completed xls file with the values obtained and links for the datasheet.
For backside no pictures until now.

[Behemot]

Re: PowerMust 600 dead

Check small caps, many of them tend to go bad in these craps. Be extremely carefull when soldering this - they make the traces from some shit and it is incredibly easy to tear them off. Often you remove a cap and part of the trace just disappears.

[alindumitru46]

Re: PowerMust 600 dead

Yep, sir!
So fast the true problem with routes arise... one of the smallest capacitor is victim already.
I measured and put the results. Obviously I'll try to change them quickly as I can.
I have 2 solid capacitors, is a good option?!

[Behemot]

Re: PowerMust 600 dead

Seems like they are already going bad, though not SO bad it should not work, yet.

These are usually crappy 85° C generral purpose, polymers won't do good. Use 105°C GP caps like Chemi-con KMG, Nichicon VZ, Rubycon PX.

[momaka]

Re: PowerMust 600 dead

Seems like they are already going bad, though not SO bad it should not work, yet.

It's not unusual for small general purpose caps to show "high" ESR in the range of 1-2 Ohms. Virtually every 100 uF, 16V cap out of the Xbox 360 motherboards I've tested has read at least 1 Ohm ESR (these are usually Chemicon KMG, Nichicon VZ, and/or Rubycon YXA).

As I stated in my earlier posts, find out why that LM317 voltage regulator doesn't have any voltage going to it - that is, if you want to fix/troubleshoot this UPS.