PDA

View Full Version : IBM 5150 Dead On Startup



interalia
August 10th, 2010, 09:08 AM
Hi

Tried to fire up the 5150 yesterday and although the PSU spun up ok. There was a burning smell and no POST.

I took the lid off and took the PSU out and the smell seems have emanated from the PSU vent.

Do I have to put a load on the PSU in order to put a multimeter on it?

Any other advice? I am right to assuming it will be a cap inside the PSU?

Regards

Neil

Surrey, UK

Lorne
August 10th, 2010, 12:37 PM
From what little I do know;

- you shouldn't need a load on a 5150 PSU to test it (and definately not on the black one)

- it could well be a cap, but you won't know until you open it up, so......

Chuck(G)
August 10th, 2010, 12:59 PM
Also note that if the first bank of RAM is bad and this is a 16-64K 5150, it will just stop. No beeps, nothing--goes to a HLT instruction and dies.

Chuckster_in_Jax
August 10th, 2010, 02:23 PM
Also note that if the first bank of RAM is bad and this is a 16-64K 5150, it will just stop. No beeps, nothing--goes to a HLT instruction and dies.

Would the same thing apply to a 64K -> 256K board? I got a machine off of Craigslist awhile back and I'm thinking the RAM may be a problem. It appears dead, and as you described; no beeps, nothing. The board has 3 rows of sockets for RAM that are all empty and 1 row of RAM chips soldered in.

Same problem on 2 IBM PCJr's Ive got. I'm suspecting RAM.

tezza
August 10th, 2010, 02:30 PM
Would the same thing apply to a 64K -> 256K board? I got a machine off of Craigslist awhile back and I'm thinking the RAM may be a problem. It appears dead, and as you described; no beeps, nothing. The board has 3 rows of sockets for RAM that are all empty and 1 row of RAM chips soldered in.

Same problem on 2 IBM PCJr's Ive got. I'm suspecting RAM.

Maybe not. I had a faulty RAM IC in a 64/256k board that very first row of chips (the row that was soldered in..grrr). The machine certainly gave me a POST message which helped identify it as a RAM issue. Incidently, piggybacking a good RAM chip over the suspect helped confirm it. The machine booted with no errors!

Tez

k2x4b524[
August 10th, 2010, 03:17 PM
if there were a burnt smell from the supply as you state, i would open it up and look for a blown cap, or some kind of short. Bad ram wouldn't cause it to produce such a smell..

james1095
August 11th, 2010, 10:38 AM
I would start by testing the power supply, or you could swap in a known good one.

The motherboards have a number of tantalum bypass capacitors on the power rails which have been known to short. The first time I powered up my XT after hauling it out of storage it had just this symptom, so I power cycled it. I thought I heard the hard drive try to spin so I power cycled again and there was a bang and sparks shot out from between some cards. One of the tantalum caps had shorted and was causing the power supply to shut down, then exploded. Replaced that and problem solved. IBM used unusual 3-legged capacitors to prevent backward insertion but they can be replaced with ordinary 2-leg tantalum caps.

interalia
August 22nd, 2010, 06:26 AM
Hi

First off, thanks to all the replies above. I have now pulled the PSU and these are my readings (no load):

P8
Yellow 11.9v
Green 11.54v
Purple 5.34v

P9
Red x3 5.36v
Blue 4.95v

P10
Green 11.54v
Red 5.34v

P11
Green 11.54v
Red 5.34v

Is this a healthy PSU? Is the 12v line a bit low?

If this is OK am I looking at pulling my 256K motherboard now?

Regards

Neil

modem7
August 23rd, 2010, 12:48 AM
Maybe not. I had a faulty RAM IC in a 64/256k board that very first row of chips (the row that was soldered in..grrr). The machine certainly gave me a POST message which helped identify it as a RAM issue. Incidently, piggybacking a good RAM chip over the suspect helped confirm it. The machine booted with no errors!
All three revisions of the 5150 BIOS test the first 16KB of RAM (before any on-screen display). As Chuck wrote, failure of that test results in the machine halting with no indication. The remainder of RAM is checked later, after the display is initialised.
On a 64/256KB board, the soldered-in bank is 64KB sized, and so your faulty RAM chip must have failed in such a way that the first 16K of it was still good.

modem7
August 23rd, 2010, 12:56 AM
I have now pulled the PSU and these are my readings (no load):

P8
Yellow 11.9v
Green 11.54v
Purple 5.34v

P9
Red x3 5.36v
Blue 4.95v

P10
Green 11.54v
Red 5.34v

P11
Green 11.54v
Red 5.34v

Is this a healthy PSU? Is the 12v line a bit low?
If this is OK am I looking at pulling my 256K motherboard now?

So, you've found that with the PSU disconnected from the motherboard, the +/- 12v , the +/- 5v, and the POWER GOOD line is as they should be. That highly suggests that the PSU is good. Don't worry about the 12v being slightly under. Maybe the odour you smelt in the PSU was due to the odour being sucked through the PSU.

Maybe the motherboard is shorting out one of the voltages (-12v, +12v, -5v, +5v) from the PSU. The scenario of a shorted tantalum capacitor on the motherboard, as described by James1095, is quite common. See http://members.dodo.com.au/~slappanel555/failure.htm
Note that just because the fan in the PSU is turning doesn't mean that the PSU must be working. The fan, for example, may be connected to the +12v within the PSU, and it is one of the other voltages that is shorted. To see if the motherboard is shorting out one of the PSU voltages, connect the motherboard to the PSU and do the same voltage measurements that you did earlier, excepting that you only need to measure on P8 and P9 (the connectors for the motherboard).
If there is a short, you'll see the POWER GOOD line, the purple wire on P8, stay near zero volts (or low, e.g. 1 volt), indicating that the 'power out of the PSU is not good'. And with that, you'll find that one of the voltages (-12v, +12v, -5v, +5v) is near zero volts.

Let us know what you measure.