Help with dead AT PC Power Supply

[naujoks]

The capacitance is 1000pF.

[Harry_22]

I think that Discontinuous Conduction Mode (DCM) is used here.
The reasons are as follows:
- Compact dimensions dictate the choice of a smaller transformer.
- High oscillator frequency of 144 kHz (Rt = 12 k, Ct = 1000 pF) which means controller using divide by 2.
-The construction utilizes additional filters and snubbers to reduce EMI.

So we are going to replace the broken controller with UC3844 that has 50% duty cycle.

Of course I may be wrong. Please correct me.

[naujoks]

Ok! I don't know anything about these things, so I will order one, put it in and then we will see what happens.
It will take a few days to arrive.
Thank you so much for all your help so far!

[Harry_22]

Hi!

Do not remove the faulty chip. Clean the pins and solder a new one on top as on the picture.


Before connecting repaired AT PS to the mains perform the following tests:

1. Connect a bench PSU to C8 capacitor (see the picture below). Measure voltage at the Reference point, slowly increase bench PSU output until 5V appear at the Reference point. The voltage of bench PSU should be around 16V.
Reduce the voltage until 5V disappear. The power supply readings are about 10 volts.

2. Turn on the controller raising the voltage up to 16V.
Connect oscilloscope to the Gate point and observe meander.
The frequency should be around 70kHz. Check it.

3. Connect another bench PSU with 5V to AT PS output as you did earlier (see the picture below). Slightly decrease and increase the voltage while observing with oscilloscope a signal changes at the Gate point. Keep in mind that when the voltage increases 5V, the OVP circuit (thyristor) will cause a short at this rail. Disconnect 5V PSU and the short will go away.


If all tests are successful, proceed to full-scale testing. Be careful. In case of MOSFET breakdown, pieces of its case, as well as controller, can injure you.