Bestec ATX-1523F

[Heihachi_73]

Re: Bestec ATX-1523F

Here's my version of the Bestec ATX-1523F (F for FAIL maybe?). Apologies for the poor quality photos, old dim CFL still hasn't been changed out.

[PC@live]

Re: Bestec ATX-1523F

I'd think so, even if it mostly dries up, there'd be enough leakage at 28V that it'd get rather hot.




Hmm...

Maybe "2.8V?"



You need to remove the riser and remove that glue, in addition to recapping.
The whole supply isn't going to work until you've got proper standby and aux voltages.

: Redface:Sorry, at the moment I am very busy, and I put in St.By the Bestec.
+ 5VSB 28V marks not!!!, but only 1,28V.
I checked and found OK the Q2, Q16 and Q17.

[Heihachi_73]

Re: Bestec ATX-1523F

Just recapped the HP 8800, swapping out the single failed KCC LXV with a 1500uF 10V KZG (yeah, I know!) pilfered from a dead IBM SFF toaster box, and threw in a lowly 64MB of PC133 just in case it went belly up and took the RAM with it (not going to waste 128MB or 256MB).

Got out my spare Cooler Master RS-500-PCAR-J3, plugged it in along with the monitor and Compaq PS/2 keyboard, shorted the board's power connector with a screwdriver and the HP came to life, resetting all the CMOS settings and defaulting the date to 01/01/1990. Success!

Luckily for this PC, the Bestec didn't make a roast dinner out of it. HP 1, Bestec 0.

Added one completely dead CR2032 battery into the bad cap bin for some future poppage.

[PC@live]

Re: Bestec ATX-1523F

Just recapped the HP 8800, swapping out the single failed KCC LXV with a 1500uF 10V KZG (yeah, I know!) pilfered from a dead IBM SFF toaster box, and threw in a lowly 64MB of PC133 just in case it went belly up and took the RAM with it (not going to waste 128MB or 256MB).

Got out my spare Cooler Master RS-500-PCAR-J3, plugged it in along with the monitor and Compaq PS/2 keyboard, shorted the board's power connector with a screwdriver and the HP came to life, resetting all the CMOS settings and defaulting the date to 01/01/1990. Success!

Luckily for this PC, the Bestec didn't make a roast dinner out of it. HP 1, Bestec 0.

Added one completely dead CR2032 battery into the bad cap bin for some future poppage.

OK now I repaired!
I replaced a total of 6 capacitors.
What was worse was the C14 10uF 50V, over 40 ohms of resistance!
I also changed the C12 (2.2uF 50V Jamicon) with a 2.2uF 100V.
Also changed:
C46 (10uF 50V Jamicon)
C45 (10uF 50V Teapo)
C35 (22uF 50V Jamicon)
C50 (220uF 25V Jamicon)
all replaced with Panasonic.

[technicalspecialist]

Re: Bestec ATX-1523F

I have a bestec ATX-1523F on the bench at the moment that I am trying to repair.
The problem is that the main supply just won't start. 5VSB supply is working, but it is a bit low at 4.85V unloaded. It drops when i apply a 33ohm load resistor.
When I short PS-ON to GND, the green LED goes out but there is no voltage at all on any of the main rails.
Its not even trying to start, no clicking or pulsing. 5VSB stays the same. It is drawing about 9W from the mains.

I have replaced the following electrolytics:

All replacement caps were tested with my Peak Atlas ESR60 and before soldering in .

5V and 3.3V: 2200uf 10V replaced with 2700uf 6.3V Rubycons from RROD xbox360 because they are the closest value low ESR caps I have that fit. ESR measured 0.02 or less.

12V: original 2200uF 16V caps were 0.02 ESR and at least 2200uF so i put them back in as I don't have any new spare 2200uf 16V caps in stock that fit.
1500uF 16V on xbox 360 board is the closest I have that'll fit but 1500uF is probably too low
(ripple would be too high).

5VSB: 680uf 16V replaced with 1000uf 6.3V Sanyos salvaged from dead skt 754 m'brd.
ESR is 0.01 and are at least 1000uF. 5VSB is still about 4.85V.

Also replaced all the small caps on the riser PCB.
0.1uf and 0.47uf caps replaced with 104 and 474 film caps.

10uF 50V replaced with 10uF 50V Nippon Chemicon from old PCB in my junk box. ESR is 0.80 and 10.5uF.

2.2uF caps measure about 1 ohm ESR and exactly 2.2uF.
I haven't replaced these as I haven't got any lower ESR 2.2uF caps in stock and I think 1 ohm ESR is ok for 2.2uF.

I haven't tested or replaced the 2 primary caps. They are not bulging and these rarely go bad anyway.

I have checked every diode for shorts and none are shorted.
I unsoldered and tested the primary MOSFET by applying 3.5V between G & S with my bench PSU then measuring the RDSon with my multimeter. RDSon measures 2ohms which is correct when vGS = 3.5V according to the data sheet.

There are two 8 pin DIP ICs on the riser PCB, TPS3510 and UC3842B. Understanding and tracing the circuit on the riser PCB is extremely difficult because of the SMD parts and white silk screen covering the tracks.


I have tested everything I can think of but now I am stuck. I think the fault is something other than bad caps now.
I want to get this PSU working. I have lots of spare working ATX PSUs on a shelf but none of these will fit in the case as this is microATX PSU.

Thanks

[momaka]

Re: Bestec ATX-1523F

Hi technicalspecialist, welcome to the forums!

Looks like you have eliminated most of the probable causes, which is very good.

Since the 5VSB seems a little low and the PSU is drawing around 9 Watts, I think that's reasonable grounds to investigate around the 5VSB circuit. In particular, the 5VSB in your PSU will most likely also have a second output: an auxiliary rail on the primary side. This auxiliary rail is used both for feedback and self-sustaining operation of the 5VSB as well as providing power to the PWM controller, the UC3842b IC.

When you send a signal to turn ON the PSU, pin 7 (Vcc) of the UC3842b IC should to rise to at least 16V or more to go out of UVLO (under-voltage lock-out) protection. Once the PSU is running, pin 7 (Vcc) will need to have at least 10 Volts to make the PSU continue to run.

Now, I see you mention that the TPS3510 and UC3842b are the only ICs in your PSU. As such, the 5VSB circuit could be a 2-transistor design, and those are known to often have a critical cap on the primary. The critical cap actually filters that auxiliary 5VSB rail on the primary side that I mentioned above. So check if the cap for that rail is good. If you have replaced all of the caps already, then check the voltage of the auxiliary rail of the 5VSB. Perhaps there may be something that is pulling it low. Then check the voltage on pin 7 (Vcc) of the UC3842b IC.

Let us know what you find.
Here's the datasheet for the UC3842b IC, in case you need it:
https://cdn.badcaps-static.com/pdfs/...020f09c147.pdf

[technicalspecialist]

Re: Bestec ATX-1523F

I managed to get it working successfully. One of the four opto-couplers was open, specifically the one that signals the UC3842 to turn on. There was 4.98V across the secondary side LED when the PS-ON wire was shorted to GND. Since LEDs work on about 2V, I figured that it might be open so I unsoldered it and measured the resistance with my DMM and sure enough it was open. I replaced it with one from an old scrap Antec TruePower 550 thats in my box of scrap boards that I use for parts. That uses the same topology (single transistor forward) and similar PWM chip (UC3844) to this Bestec 1523F so it had the same specs. I don't know why it had gone open but it did have a bit of corrosion on the pins. It did have some of that horrible light brown glue on it that goes conductive over time (it hadn't turned black or gone conductive though). I will attach some photos of the board in a second post.

[technicalspecialist]

Re: Bestec ATX-1523F

Photos.
I have thoroughly cleaned the board with an old tooth brush and isopropyl alcohol since taking these pics. I'm have edited the photos to show where components are. Best viewed full screen in a separate tab.

[momaka]

Re: Bestec ATX-1523F

Nice! Good to hear you got it working again.

So the problem was related to the PWM IC not getting power, after all.
Thanks for making those annotated images above. I'm sure they will actually come pretty helpful to anyone who has this PSU and is trying to repair it.

By the way, I see some United Chemicon SME capacitors on that circuit board. I don't suppose those were originally there. But either way, make sure to use 105°C -rated caps only, and from good Japanese manufacturers. SME is a general purpose 85°C series, and so those caps will likely have a rather short life in that PSU. Also, since you did note that there is a second transistors for the 5VSB, it is even more critical that you have good capacitors on that daughterboard. A 5VSB line a 7+ Volts is no good for your motherboard, and that can often happen with these 2-transistor 5VSB circuits when they develop bad caps.

[technicalspecialist]

Re: Bestec ATX-1523F

No, those SME caps are not original but they have lower ESR than the originals and lowest ESR (about 0.8ohm if i remember correctly) of any of the 10uF caps I could find in my capacitor drawer or on scrap PCBs. They are from the CPU board of an old UPS with fried MOSFETs and bloated battery's. It wasn't worth fixing.
Other 10uF caps i tested had at least 1.2 ohms or higher ESR.
I have moved that power resistor away from the caps with pieces of wire and hot glued it in place on top of the optos. It doesn't actually get too hot, only slightly warm. I can't work out what its purpose is though. Here are some more photos of the main board showing where the daughter board connects.