5150 won't power on. Dead?

[generic486]

And I thought the Quadboard was the worst of my troubles. It now seems to be that the 5150 has died completely. I have plugged it in, used different cables but not a sign of life. Does anybody knwo if there is a remedy for this?
EDIT: I've opened the system up. Devoid of power. The power supply fan will not spin up even with the motherboard power connector P8 removed. It also will not power on with just a Tandon full height floppy attached. I've come to the conclusion that the power supply is dead. What could be some remedies?

 

[modem7]

Follow the procedure at http://www.minuszerodegrees.net/5150_5160/MDC/minimum_diag_config.htm

 

[generic486]

That was the first thing I went through modem7. It is most likely the power supply. I have taken it apart and a little visually inspection shows that the fuse is not blown. The line suppression caps have not ruptured althought there is some some stress cracking (very minor) on both of them. However there is a bunch of about 5 electrolytic capacitors in about the center of the PSU. On four of them there is a blue coating? on them. Is this usual? One of these caps looks suspeciously bad. There looks like a pin sized hole in the top.

 

[modem7]

That was the first thing I went through modem7.

You should have mentioned that.

It is most likely the power supply.

You will be able to prove that by using temporarily using the power supply from your 5170.

However there is a bunch of about 5 electrolytic capacitors in about the center of the PSU. On four of them there is a blue coating? on them. Is this usual?

We would need to see a photo.

One of these caps looks suspeciously bad. There looks like a pin sized hole in the top.

A photo will help.

Before you invest a lot of time with the power supply, I suggest that you first prove it to yoursef that the 5150 power supply is at fault by using your 5170's power supply in its place.

 

[generic486]

As suspected, the computer works with the 5170 PSU.

You can't see the blue coating on top of the capacitors but you can see a small brown mark on one of the caps. I believe this is what may have failed.
Looks as if the picture is not coming up. I see what I can do.

 

[generic486]

Here is a video:

 

[modem7]

The video is very informative.

Regarding the aluminium electrolytic capacitors with the blue on top. To me it looks like those capacitors have started leaking. If it were my power supply, I would be replacing them, but I'm not confident that they are the cause of your current 'no-go' power supply problem.

Regarding your Quadram card. Does it have "Quadboard" marked on it anywhere? What worries me is that the locations of SW1 and SW2 on your card don't reflect what is shown in figure 1 of the the 256K Quadboard manual at http://www.minuszerodegrees.net/manuals.htm

 

[generic486]

It does not say Quadboard on my card. I simply assumed that it was. I might have made a big mistake. I thought that all Quadram memory expansion boards were Quadboards. There were a few differences between my card and the one in the manual. My card has the switches in a different spot and a few different electronic components are also moved. I only suspected that these changes were minor and that my card was like a revision of the one in the PDF. I hope the person I got the 5150 has the GTO book because maybe it will have the Quadram memory expansion upgrade manual in it. My card is a very early card with a black metal bracket and the card is dated 1982.

 

[modem7]

It does not say Quadboard on my card. I simply assumed that it was. I might have made a big mistake. I thought that all Quadram memory expansion boards were Quadboards. There were a few differences between my card and the one in the manual. My card has the switches in a different spot and a few different electronic components are also moved. I only suspected that these changes were minor and that my card was like a revision of the one in the PDF. I hope the person I got the 5150 has the GTO book because maybe it will have the Quadram memory expansion upgrade manual in it. My card is a very early card with a black metal bracket and the card is dated 1982.

So as you've realised, there is now the additional possibility that your card is fine - it's the switch settings that need to be determined.
An additional possibility: Verify that no tracks have been damaged where the battery was removed.

By the way. I've put photos of some Quadram cards at http://www.minuszerodegrees.net/quadram/quadram.htm

 

[generic486]

I'll be going past a Jaycar today.
The things I'll be getting are:
4x 2200uf 10V electrolytic caps for the power supply,
2x 0.1uf 275V X2 line suppression caps
also I'll still replace the 47pf ceramic cap though on the Quadram Mem expansion.

None of the traces look like they are damaged. There is some rippleing of the large ground traces on the underside but this does not affect the smaller tracese underside on the or the other side.
Also, I had a look at the Quadram cards but my card is like none of those. My card has the Quadram logo, Made in USA Copyright 1982, serial 2355B and it has a PN 8920-3. All of the card on that pages have the switches next to each other. Mine does not. The card has the date code of 83-06 so I assume it was made in Feb 1983.

 

[generic486]

Also don't know if this helps but when I first got the card, switch 4 was different to all the other dips in SW1. It was either on or off, I forget which but switch 4 was different to all other dips. So either switch 4 was off and 1-3 and 5-8 were on or switch 4 was on and 1-3,5-8 were off. I think it was swicth 4 that was on due to the dust on top of all the other dips except 4. SW2, I have not changed at all. It has always had 1-4,6-8 off and switch 5 on.

 

[generic486]

So I replaced the faulty caps in the power supply but it still won't go. Have you got any ideas of what it could be? Also, the gentleman who served me got me 0.1uF / 100nF 630 Volt Greencap Metallised Polyester Capacitors instead of the line suppression caps because they did not have any. Could I use use these? I have one 0.1 250VAC X2 line supression cap from an old PSU. Could I use this? The supression caps in the 5150 PSU are rated for 275VAC maximum.

 

[RickNel]

but it still won't go. .

Assuming you have a mains-rated multimeter, I would be checking voltages forward through the PSU circuit VERY CAREFULLY to avoid shorting mains voltage to anything. Start with any AC output taps from an input transformer, see if switching logic chip(s) are alive, see if the power-switching transistors are receiving and outputting anything. If "won't go" means no voltage on any of the connectors, then its more likely that the problem occurs up the chain before the voltages are split - eg switch logic is not getting powered or switching tranny has failed. Worst case would be a burnt-out transformer coil.

On the subject of switches - verify the master ON/OFF switch itself. They can fail with age and it's not always obvious. They can keep clicking without making the right contacts.

 

[generic486]

Okay, so testing the board so far gave these results. The power reached past the first board (the board attached to the side of the power supply) and the switch. The power was about 240VAC. It reached the mainboard and transformer. On side of the transformer read 240 and the other side (low voltage) read 20V. However, all of a sudden while retesting the high voltage side of the tranformer, the voltage jumped to 360 to 390V. A little bit odd. The power seems to reach the bottom board and I have tested up to the transformer's low voltage side.
EDIT: testing the ground shows this. When testing for continutity on the ground from the molex connectors, I was able to get continuity only to the middle of the board. The last point where I got last continuity was close to the Motorola SOC 604A chips.

 

[modem7]

So I replaced the faulty caps in the power supply but it still won't go.

They would have caused problems in the long run.

Have you got any ideas of what it could be?

If you haven't done so already, a very thorough visual inspection may reveal the faulty part. Using a magnifying glass is recommended. You could save hours of work by spotting a component that has visible indication of failure. Also inspect the solder side of the board.

In my experience, and what I've heard others say, the most likely (repeat: likely) components to fail in a switch mode power supply are aluminium electrolytic capacitors (big and small) and high power semiconductor devices (e.g. high power diodes, high power transistors).

If your 5150 shows evidence of being stored in a moisture prone environment, then take a good look at the glass encapsulated diodes. See http://www.vintage-computer.com/vcforum/showthread.php?20188

 

[modem7]

At some point (after much unsuccessful effort), perhaps consider what others here have done - replace the 'guts' of the power supply with non-IBM 'guts'.
In that way you still have what appears to be a geniune IBM power supply.

JDT's photos of such an operation are [here].

 

[modem7]

There's a 5150 PSU circuit diagram at http://www.minuszerodegrees.net/5150/5150_psu_from_sams_computerfacts.pdf
It may not match your 5150 PSU because IBM may have made design changes to the PSU during its lifetime.

CAUTION: Like most switch mode power supplies, there is no mains isolation transformer 'up front'.

 

[generic486]

They would have caused problems in the long run.


If you haven't done so already, a very thorough visual inspection may reveal the faulty part. Using a magnifying glass is recommended. You could save hours of work by spotting a component that has visible indication of failure. Also inspect the solder side of the board.

In my experience, and what I've heard others say, the most likely (repeat: likely) components to fail in a switch mode power supply are aluminium electrolytic capacitors (big and small) and high power semiconductor devices (e.g. high power diodes, high power transistors).

If your 5150 shows evidence of being stored in a moisture prone environment, then take a good look at the glass encapsulated diodes. See http://www.vintage-computer.com/vcforum/showthread.php?20188

It was exposed to a moisture prone environment. Both the AT and the PC as well as an XT that the gentleman had, were stored in his garage for 20 years. Oddly the AT shows virtually no signs of rust or other affects of moisture. The PC however does, the power supply case has what looks like the start of very minor rust but the PC case had some too but it was not deep and was easily removed. I will certainly look at the diodes. I will try anything to resurrect this computer. I was hoping to keep it IBM genuine if I can as I already have 2 XT clones. I may replace all the electrolytic caps on the board. The line suppression caps are okay which is good.
Well, maybe I am doing something wrong but the continuity of all the diodes on the board (glass and high voltage black diodes) appear to be bad. Firstly normally when a diode is mesured, the continuity should either be very high or complete continuity (000 on my multimeter) or no continuity (1 on my multimeter). All the glass diodes appear to do this:
When mesured, regardless of polarity, the multimeter will display a reading of about 40 to about 900. On my multimeter this tells me there is some partial resistance and neither changing the probes does anything (same results) . On the high voltage diodes, the readings are very close to 000 either way (either polarity or swtched probes). Perhaps, I have done something wrong but I thought this might be worth mentioning none the less.

 

[modem7]

Well, maybe I am doing something wrong but the continuity of all the diodes on the board (glass and high voltage black diodes) appear to be bad. Firstly normally when a diode is mesured, the continuity should either be very high or complete continuity (000 on my multimeter) or no continuity (1 on my multimeter). All the glass diodes appear to do this:
When mesured, regardless of polarity, the multimeter will display a reading of about 40 to about 900. On my multimeter this tells me there is some partial resistance and neither changing the probes does anything (same results) . On the high voltage diodes, the readings are very close to 000 either way (either polarity or swtched probes).

I prefer to check diodes using the 'diode check' functionality on the multimeter. It's explained at http://www.allaboutcircuits.com/vol_3/chpt_3/2.html

The forward voltage drop of diodes vary. A small germanium diode may measure about 0.2V, whereas a small silicon diode may measure about 0.6V Large power diodes may measure around the 1V mark.

Perhaps, I have done something wrong but I thought this might be worth mentioning none the less.

Note that some diodes will have other components in parallel with them, and those components could corrupt/affect what you measure. In such cases, you will need to unsolder one leg of the diode so as to make the measurement out-of-circuit.

 

[generic486]

Yes, I had a feeling that that would be the case. But I didn't know if that involved diodes. I thought that was only the case if you were mesuring current.
I used the ohm function and I still get similar results, however, erratic. Sometimes the diode does it's job and only allows current though one way. However other times, when remesuring, it will allow current though both ways. I will most likely replace them. I assume that those glass diodes are 1N4148 diodes. Would I be right? Also the high voltage diodes. Are they Zener or Schottky?