BMH Online  Projects

 

[Back to Projects]

  PSUs...

[June 2021] I've been on a bit of a roll lately with repairing various computer parts. I recently recapped a motherboard [see here] and before that replaced a BIOS chip on a laptop motherboard [see here].

[PSU #1]

When I plugged in one of my spare computers and it went "pop", I immediately knew the Power Supply Unit had failed. It had been problematic some months ago when it could no longer cope with a graphics card, even though it was rated at 600W (it was some years old though).

I'd inspected the components and noticed some capacitors were failing.

At that time I didn't even entertain the idea of meddling with a power supply so I just swapped that unit into a less demanding computer where it worked ok. That was until it went pop!

Power supplies deal with high voltages and should not be meddled with, even when unplugged the components can still hold a dangerous charge. I recently got a lesson in this when I got a shock off an iMac's screen inverted board [see here].

I wasn't going to let that deter me, I didn't want to relegate yet another power supply to my ever-growing scrap pile for the sake of soldering in a few new components; I would just tread carefully... I also had another power supply (a 500W model) that had failed that I had designated to the scrap pile, but I now retrieved it to give it similar treatment.

I reopened the 600W PSU (after leaving it for a day to discharge - although this is no guarantee that the capacitors don't still hold a charge). In addition to the obvious capacitor issue there was fresh new damage from the component that had seemingly gone pop.

I wasn't sure what type of component this was, let alone its value. It looked like the damage to it could have obliterated any details on it, but fortunately it was one of a pair so I proceeded to remove both of them.

It turned out I could read the details on both of them: "SCK 053". While that was too vague for me I was able to find a data sheet online that revealed that SCK was a product code for an "NTC-Surge Current Killer" or Thermistor, and 053 indicated 5 ohms / 3A. I found I could order exact replacements, or spend a little less and perhaps get them a little sooner (still coming from China however) if I bought a set of different ones but with the same rating (I opted for the latter).

I tested the "thermistor" that appeared undamaged but the reading appeared haphazard with the resistance either being different each time and not matching 5 ohms, or not registering at all. Perhaps that was the nature of these components, or both were damaged.

Thermistors change resistance with temperature changes; they are temperature-dependent resistors. They're perfectly suited to scenarios where one specific temperature needs to be maintained, they're sensitive to small changes in temperature...

That would perhaps explain the variable reading, but the conclusion was that I couldn't really test the "probably ok" one without a known good one to compare with, but I'll likely replace them both anyway.

I also removed the obviously failing capacitors from each of the power supplies making a note of their values, and orientations; unlike the motherboard I recapped, the power supply boards didn't have the capacitor orientation indicated on them, and this is important. I'm not sure if the thermistors have an orientation but I have noted that also just in case.

I now await the replacement parts... I'll also need to hunt down the fan I pinched from that 500W power supply...

[Laptop Chargers] In addition to the desktop power supplies, I recently repaired a laptop charger; the lead had been damaged (by a vacuum cleaner going over it). The damage was near to the adapter so I figured I could reattach it without losing too much lengths. But how to get into it?

I tried various pry tools but it appears the only way into these is to break them so I hammered screwdriver carefully around the seam until the casing cracked open. I was able to reattach the lead. I had to slit the cable protector open to get the lead out. The end result certainly isn't pretty and the best I could do was strap the case back together with insulation tape and glue the cable protector closed.

I have another laptop charger that needs similar attention; the lead isn't entirely broken, and it still works, but it looks like a good few strands of wire are broken within and I'm not convinced this is entirely safe to use unattended like this.

Should I attempt a similar repair?

[PSU #2] The second PSU I attempted to repair had been delegated to my scrap pile some months previously, but I now wanted to give it a second look. Sure enough there were some suspect capacitors, but no other blown parts that I could see. I remembered how this power supply had stopped working; I'd plugged something into it I shouldn't have; I can't remember how it failed (I don't recall it going pop, at the most it just fizzled out when I plugged it in). Now it did nothing.

Since I had nothing else to work with I decided to replace the suspect capacitors. Sadly this hasn't got the PSU working, it still does nothing. I've watched some Youtube videos about repairing these and therefore intend to test some more components...

[PSU #3] I got overly confident, even before I repaired PSU #1, and I bought a faulty PSU cheap on ebay thinking I could fit it. This one appears new and there is not sign of failed components, just like my PSU #2 it just doesn't switch on... does it have the same fault?

[Shuttle PSU] I bought an old 'Shuttle PC' off ebay to tinker with. Upon visually inspecting the PSU I noticed the familiar sign of a bulging capacitor. It's of a difference value to the others I have replaced so far so I have ordered what I require. In the meantime I have serviced the fan.

[July Update]

I got busy today testing components in my pile of faulty power supplies [referring to this video]. In my gold-coloured 500W PSU I discovered what turned out to be a faulty Schottky diode (1N5844). I couldn't find an exact replacement but in referring to a data sheet I learned it was rated at 40V and 5A. This enabled me to find an alternative in the form of SR5100. These are rated at 5A and "up to 100V". I've ordered a job-lot since they will be coming from China.

In another PSU I discovered a faulty bridge rectifier, I first noticed the solder on the rear of the PCB to be discoloured and possibly fractured. I had to remove all three which were attached to the same heatsink and then I could test them individually. It was a MBR30100CK (30A 100V) that was faulty, but I could only find ones referred to as MBR30100CT and I was unable to find out what the CT/CK refers to. So I have ordered some of those, also coming from China.

[PSU #4 - 28/02/22] I'd bought this Enermax EMD625AWT Pro82 II 625w Power Supply off Ebay as working, and I'm pretty sure I tested it when it arrived, but it stopped working. Some investigating revealed that the internal 12A glass fuse had blown. From what I understand this will usually only fail if there is another fault but I tested it with another and it worked with my PSU tester. However, that fuse I installed for testing was too low a rating and I'd installed it in a plastic fuse holder that wouldn't fit when I came to reassemble, so I had to specifically purchase the correct fuses. Then when I came to do the final test with my PSU tester, I noticed that that was reporting that the 5V rail was low at only 4.5V (below the ATX standards). I have not yet tried the PSU in a computer; perhaps this is why the fuse blew. It will be beyond my ability and patience to determine what component is out of spec.

Helpful Videos:

https://www.youtube.com/watch?v=HcYFbCqM61g

https://www.youtube.com/watch?v=nlVGddo8DTA

https://www.youtube.com/watch?v=g8nWjpil9Hw

 

[Back to Top]