Real Power 235W Model SR-235FS

[torin3]

I'll post some pictures later, but I just recapped this PSU. It has several swollen caps. I didn't replace the primary caps, as they didn't look bad and were Rubycons.

This is an older PSU, 20pin motherboard connector. 8A on the 12V rail. The 3.3V and the 5V rails test pretty close to right on, but the 12V rail tests out to 11.4V

That is right at the edge of the specification. Should I look for a way to bring it closer to 12V, and if so, where should I start looking?

This PSU was running an old Athlon XP system that I also recapped.

[Toasty]

Re: Real Power 235W Model SR-235FS

>>I didn't replace the primary caps, as they didn't look bad and were Rubycons.<<
Did you replace all other caps though?

Does it have TL494 PWM and LM339 Q-Comparator IC's in it?

Check for failed resistors near them.

What exactly are the 3.3v & 5v readings?

Toast

[370forlife]

Re: Real Power 235W Model SR-235FS

This *sounds* normal for a older lower watt power supply. Put a load on it, particularly some more 12v loading than a athlon xp board, and it should go back up. 11.4v is on the very bottom of the ATX spec for the 12v.

[torin3]

Re: Real Power 235W Model SR-235FS

>>I didn't replace the primary caps, as they didn't look bad and were Rubycons.<<
Did you replace all other caps though?

I replace all of them except for a small 10μF 50V that I apparently missed when I was putting together the order.

Does it have TL494 PWM and LM339 Q-Comparator IC's in it?

Not that I could see. There are 2 8pin ICs and 3 4pin ICs. I'm attaching a pic of a closeup of the area

Check for failed resistors near them.

What exactly are the 3.3v & 5v readings?

Toast

3.29V 5.15V

And if I remember right, I can check the resistiors in the circuit, and rated resistance or lower should be ok, but higher is a sign I should pull and check resistance out of the circuit?

Also, I note that the board has a Seasonic logo that has been sharpied out.

Btw, sorry I'm slow in replying. I don't have much computer time at work anymore, and my favorite online game is closing tomorrow, so I've been involved with all the end of game stuff related to it.

[Toasty]

Re: Real Power 235W Model SR-235FS

It's a Seasonic. Look at the brand mark on the transformer.

I marked 2 resistors with yellow arrows that appear a little cooked. Slight discoloration/browning.

270 ohm? and a 22 ohm. The 270 ohm below (green arrow) appears undamaged and when compared to the ? one, appears the same in value.

Chip to the right is KA3843A PWM controller.
What are the numbers on the chip marked with blue arrow?

Check the blue and the green ceramic power resistors (mounted upright) in the other picture between the transformer and the heatsink next to the mains caps.

>>And if I remember right, I can check the resistiors in the circuit, and rated resistance or lower should be ok, but higher is a sign I should pull and check resistance out of the circuit?<<

Correct (mostly). Significantly higher than marked tolerance are bad.
1000 ohm 5%(gold band) reading 1050 is okay (marginally high), 1100+ is bad.

Not a hard and fast rule, but pay more attention to higher wattage resistors than lower for initial troubleshooting. 1/8w-1/4w are "usually" fine. 1/2w and up are the more likely failures.

As 370 said, load it and see how the voltages respond. An old hard drive or 2 is a good way.

Happy hunting!
Toast

[everell]

Re: Real Power 235W Model SR-235FS

Looks just like my SeaSonic. His has pc board date 1999 whereas mine has pc board date 2001.05.22

The other chip is TSP3510

Check the 5vsb. Mine was 12 volts when I started working on it.

[torin3]

Re: Real Power 235W Model SR-235FS

It's a Seasonic. Look at the brand mark on the transformer.

Yeah, I'm just amused by how little the sharpie covers it up.

I marked 2 resistors with yellow arrows that appear a little cooked. Slight discoloration/browning.

270 ohm? and a 22 ohm. The 270 ohm below (green arrow) appears undamaged and when compared to the ? one, appears the same in value.

270 ohm reads 180, the 22 ohm reads 21.8. The green arrow 270 ohm one reads 262.

Chip to the right is KA3843A PWM controller.
What are the numbers on the chip marked with blue arrow?

It is a TI TPS5510P
http://pdf1.alldatasheet.com/datashe.../TPS5510P.html

Check the blue and the green ceramic power resistors (mounted upright) in the other picture between the transformer and the heatsink next to the mains caps.

>>And if I remember right, I can check the resistiors in the circuit, and rated resistance or lower should be ok, but higher is a sign I should pull and check resistance out of the circuit?<<

Correct (mostly). Significantly higher than marked tolerance are bad.
1000 ohm 5%(gold band) reading 1050 is okay (marginally high), 1100+ is bad.

Not a hard and fast rule, but pay more attention to higher wattage resistors than lower for initial troubleshooting. 1/8w-1/4w are "usually" fine. 1/2w and up are the more likely failures.

As 370 said, load it and see how the voltages respond. An old hard drive or 2 is a good way.

Happy hunting!
Toast

A hard drive on it bumps the 12V line to 11.84V

I'm going to put the computer and the PSU back together and see if it comes on.

[Toasty]

Re: Real Power 235W Model SR-235FS

11.84v is in spec. It should be fine. Congrats!

[torin3]

Re: Real Power 235W Model SR-235FS

Well, I fired it up in the computer and there doesn't appear to be any power problems. However, I noticed the HSF isn't spinning. I'm not sure if the header is dead, or if the fan is dead. I've got replacement fans I can put on if the fan is dead. I'll just plug it directly into the PSU if the header is dead.

[torin3]

Re: Real Power 235W Model SR-235FS

Dead fan, it overheated and bubble the label. I replaced the fan and it is working fine. 12.15V on the 12V rail and 4.99V on the 5V rail.