• Please review our updated Terms and Rules here

Bad experience replacing a power supply on a XT

inakito

inakito

Experienced Member
Joined
Oct 5, 2011
Messages
297
Location
Spain
The power supply in my old 5150 was dying. From time to time the PC was shutting down by itself. I decided to replace the psu with a new one. I checked out the new one had the -5v supply. Then i wired the 20 atx connector to two p8 and p9 connectors. I tested then with a meter and every voltage was correct and in place. I had my doubts about two pins: the power good and the power select. According to what i have read power good pin must be at 5v when everything is already ok. However some people claim there are power supplies that do not provide this functionality. As with regards to power select pin i bridged it to gnd. I checked for two times every pin with a meter without the p8 and p9 connected. All ok.

Then i connected p8 and p9 to the motherboard. Then it was time to press the red big lever. I was nervous.
Red lever to the ON position and a zap sound of dead came out of the power supply. Right away i placed the lever on the off position again. A smoke smelling came from somewhere in the inners of the pc.
I suppose now i have a defunct pc. Ok, i will bury it with honors.

However, my big question is: Why?
What happened ? What was wrong? Was anything missing? Is it related to the power good pin or perhaps to the power select that i bridged? I accept the dead of my beloved IBM PC but I want to know why he dead?

Thank you for your condolences.
 
Don't Bury it! It's possible that the PC isn't dead. A lot of those old power supplies have paper filter caps that go bad. Here is a good thread for you to look at.
 
Power Good should definitely be +5V

You need to find what failed, it could've just been a titanium cap on the motherboard (which isn't always a concern) - but if it was a fault in the PSU or a short then it needs to be rectified. I certainly would NOT bury the PC until you've found the answer.

Smoke means something blew, which means there will be marks or something broken, your eyes should be able to spot it.
 
We have found that a bad capacitor on the PC XT/PPC motherboard (C56) can give two symptoms. If it has gone bad it can look like a bad Power Supply. (The other thing that it does is go bang when power is applied.) If you get the former problem, just snip the legs of C56 and remove it.
 
dpatten said:
A lot of those old power supplies have paper filter caps that go bad.
Click to expand...
Sorry if I'm being too pedantic, but I do think that paper caps are unlikely to be found in computers. :) I suspect you are referring to electolytics. I haven't checked the few old paper caps I have left, but I doubt they actually go bad.
 
I have isolated the problem to the hard disk or hard disk controller. It is a Type 31 20MB. Apparently the motor is working but it is not detected, so it might be a controller problem. The first time i connected the power supply and detected a burning smell,it was located near the hard disk controller. So it is possible the controller is dead now. The remaining components do work. The system reads the floppy and is able to enter Basic when no boot device is detected. By optical inspection i cannot see a burnt component on the controller board, it looks clean. The click i heard was coming out the hard disk itself, i figure out it is heads being getting out of park position or perhaps some relay within the disk?. It is amazing that with the case closed i never paid attention to that click noise.
 
Last edited:
Ole Juul said:
Sorry if I'm being too pedantic, but I do think that paper caps are unlikely to be found in computers. :) I suspect you are referring to electolytics. I haven't checked the few old paper caps I have left, but I doubt they actually go bad.
Click to expand...

You're right. I was talking about caps in the power supply. I hadn't fully comprehended the OPs message where he said he was using a modified ATX power supply as opposed to replacing it with another vintage supply.
 
I haven't opened an early IBM PSU yet which still has the paper caps in one piece. They work fine without them, just whoever is using it when they first blow might need to open a few windows. Haven't had an electrolytic fail on an IBM, but have had them fail on modern power supplies (annoying).


OP:
Check the motherboard around where the hard drive controller is, would be good to find the fault. Could be the HDD controller or drive, but it is quite common for the caps on those motherboards to die, and two common ones I've seen blow are right at the back of the machine around where the controller would be.

Good luck with the hard drive situation, before changing power supplies when was the last time you used it?
 
Last edited:
Last time i used the machine was about six months ago. It was failing from time to time but i was able to work for an hour or so before the system failed. Then i had to wait for a couple of minutes before restarting. I have tried to move the controller board to a different slot, without success. I have taken a close look to all tantalum capacitors. There are 4 that look pretty old with a wrinkled surface but no signs of burnings. The question now is whether to try or not a new controller. Is there any determinant check to rule out hard disk or controller?
 
This is how my adapter is wired:

dsc00923mf.jpg
 
inakito said:
Last time i used the machine was about six months ago. It was failing from time to time but i was able to work for an hour or so before the system failed. Then i had to wait for a couple of minutes before restarting. I have tried to move the controller board to a different slot, without success. I have taken a close look to all tantalum capacitors. There are 4 that look pretty old with a wrinkled surface but no signs of burnings. The question now is whether to try or not a new controller. Is there any determinant check to rule out hard disk or controller?
Click to expand...

A fair test for the controller and drive would be to see if you can perform a low level format.
The downside to this, is that you'd lose everything on the drive.
 
inakito said:
I have isolated the problem to the hard disk or hard disk controller.
Click to expand...
If they are removed from the 5150 and taken into a different room, does one smell much worse than the other?

inakito said:
Is there any determinant check to rule out hard disk or controller?
Click to expand...
So, without a spare controller or hard drive on hand.

The XT-class controllers have self-test functionality. It is crude (does not test all of the controller) but can be useful. You could run something that invokes the self-test and look at the result:

PASS = The 5150 is able to communicate with the controller. Some (not all) of the controller is known to be working.
FAIL = Either: 1. 5150 is not able to communicate with the controller; or 2. Controller reported that its self-test failed.

And so if you see a fail, then the controller is probably the problem.

[SpeedStor] software is one way of invoking the self-test (select 'Diagnostics' then 'Controller').
 
You must log in or register to reply here.
Back
Top