Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Computer PSU Problems

Status
Not open for further replies.

Beholder88

New Member
Hello, I have a PSU that I have had sitting around, and wanted to turn it into a test bench PSU. I originally stopped using it because it wouldn't power up my PC, but would work outside the case. Since pulling it out of the closet, I've been messing with it trying to get it to function again. The PSU is a Coolmax model CUL-850B. Not the best, I know, but not going to be using it in a PC anyway. I opened the case and checked all the caps and circuits to make sure nothing was burnt or bulging, and we're good there. I checked all the screws securing the PCB to the case, and made sure there wasn't anything shorting on any of the wires or on the PCB itself. The power switch went bad, so I bypassed it completely and use a power strip with a switch on it to power it on. When flipping said switch, the fan lights up and twitches a little, then the whole unit shuts down. I put a load on it with an HDD and a few fans, and they don't even budge, but the light on the fan stays on a few seconds longer and I get a bit of a hum sound coming from the PSU. For the life of me I can't figure it out, and the most I could find elsewhere online was nothing but suggestions to buy a new one. As a bit of an electronics hobbyist, that answer does not sit well with me. Any suggestions?
 
The suggestion of replacing it is probably the most cost effective solution. Switch mode power supplies are potentially dangerous so unless you fully understand the mode of operation and are well used to working with high voltages I would advise not to work on it. To be able to do any diagnostic tests on it with an oscilloscope it MUST BE POWERED VIA AN ISOLATING TRANSFORMER. The way your question is worded suggests to me that your level of knowledge is not sufficient for you to work on it. The fact that you have not even measured the voltage on the fan to see if it is the correct voltage for the fans rating makes me think your level of knowledge is not sufficient to safely work on it. The fact that it attempts to work suggests that one of the safety circuits is coming into play. If it is a three wire fan it could even be the fan that is faulty causing the power supply to shut down due to lack of up to speed pulses. The only safe test you could do is to leave it for a few days to allow the capacitors to discharge then check all the capacitors with an ESR meter

Les.
 
So to Les, that response is about the same that I see most often. "You're an idiot, you need a PhD to work on these, go buy $1000 pieces of equipment to test it..." All I'm looking for is if someone has had this experience with a PSU before and has somewhat of an idea of what it may be. The fan I am referencing is the fan that is built in to the PSU. Anything that I have plugged in to any of the leads gets no power. Also, you kindly pointed out that "...you have not even measured the voltage...". Well, I have. As I described in my post, the PSU does not stay powered on. For a few brief seconds I get readings that bounce around and do not give anything accurate, after that, nothing. If it is a safety circuit of some sort that's tripping and powering the unit down, there must be a cause.
 
Sounds to me like 1) You may have a short somewhere. Likely the filter caps are no good (just because you can't see them bulging doesn't mean they haven't gone bad. Especially if they're cheap ones, like Samxon or Capxon. 2) You have something that's heating up too much causing a thermal shutoff to trigger. Usually heating is due to excessive current, which is often caused by a short. So see #1.

Many of those types of PSUs have over-current protection, as well as under-current. If you don't have just the right load connected, it may not register that it's connected and may shut down. This is less likely--I haven't come across this much myself so I couldn't tell you for sure.
 
I rummaged around and found two DVD drives and another fan with a molex connector. I also got a HDD with a SATA power connector. I switched the supply on and got a little but of coil whine at the start but it all kicked on and stayed on. I was able to measure the voltages reliably and this is what I got: On the 3.3V rail - 3.27V, on the 5V rail pins I got between 4.8 and 5.1V. On the +12V I got 12.4V, which is on the high end but not out of tolerance from what I could find. The -12V side got me 12.1. I let it sit and run for about 20 minutes and nothing seemed out of sorts. I'm going to try to put something together that will pull more of a load on it just to be sure, but it looks like the reason it wasn't firing up before was due to there not being enough of a load on it to begin with.
 
There is a Green wire on Pin 14 of the 20 / 24 pin connector that needs to be grounded for the PSU to turn on.
 
In your OP you mentioned a HDD and a few fans. So?
 
So to Les, that response is about the same that I see most often. "You're an idiot, you need a PhD to work on these, go buy $1000 pieces of equipment to test it..."
Les is only looking out for you..... When someone that has joined the forum, we have no idea of experience...

What Les has said can be seen as "Leave well alone" or Continue with caution.. Switched mode power supplies are next to impossible when you have a PHD... The 400V is lethal that accumulates in the caps on the hot side...

I have found that these caps are usually the main source of switched mode power supply failures, BUT!! they hold the charge for quite some time.... The last time I tried to drain them by shorting them out lost me one of my favorite screwdriver ( I know we all do stupid things )..

As well as a resolve you can learn here...
 
As Ian says, caps are a frequent cause of trouble in PSUs like this. Now that you have proved that the thing basically works, you may wish to consider replacing them anyway if you can get replacements at reasonable cost.

That said, as you know, sometimes these supplies won't run off load. I believe that the reason for this is that not all outputs are regulated independently - but they may well be protected by over-voltage shutdown. This means that an un-loaded output can climb too high in voltage, causing the entire PSU to shut down temporarily. You may like to experiment with which output(s) need to be loaded, and by how much, in order for the PSU to run.

And as everyone has said, please be careful. Always check for voltage on the mains-side capacitors before handling the board.
 
Besides the Power On Jumper to ground indicated by Ramussons,
Older PSU's need a 5% ~10% load on the primary regulated output (5V) only.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top