Capacitor list for IBM 5150 16-64KB motherboard?

[Fire-Flare]

I have a 5150 that won't boot, and I'd like to try replacing the capacitors.

I can't find a list of which ones to order online, perhaps some of you know?

 

[twolazy]

Grab a multimeter, throw it on diode test, and check each cap. If its shorted it'll show it (usually).

 

[modem7]

Grab a multimeter, throw it on diode test, ...

Diode test ?

... and check each cap. If its shorted it'll show it (usually).

When you put the probes of a meter onto a filter capacitor that is still soldered to the circuit board, you are measuring not only that capacitor, but all other components on the same voltage line.

An example. Refer to [here]. The circled capacitors are on the +5V line. If I was to put my meter probes onto any one of those capacitors, I would be measuring all of the capacitors. And it is worse than that. The IC's are also on the +5V line, and so those would be measured as well.

 

[modem7]

I have a 5150 that won't boot, ...

Exactly what are the symptoms? "Does not boot" can be different things to different people. For example, if your IBM 5150 is displaying a 301 error (keyboard related problem), stopping further progress, a capacitor on the motherboard is highly unlikely to be the cause.

Note that the symptom of 'appears to be dead' has lots of possible causes. If that is the symptom, see how you go with the diagnostic procedure at [here]. If your motherboard has a short-circuit capacitor on it, the procedure is expected to direct you to it (that is what the procedure was design to do).

 

[na103]

I have a 5150 that won't boot, and I'd like to try replacing the capacitors.

I can't find a list of which ones to order online, perhaps some of you know?

triple lead 10uF tantalum capacitor, part number T398E106M016AT on mouser.
I suggest you a careful visual inspection of all the capacitors on the mainboard. sometimes they don't explode but there are clues that point to the shorted capacitor.
on my 5150 i swapped all these "pyrotechnic" capacitors just to be safe

 

[twolazy]

Diode test ?


When you put the probes of a meter onto a filter capacitor that is still soldered to the circuit board, you are measuring not only that capacitor, but all other components on the same voltage line.

An example. Refer to [here]. The circled capacitors are on the +5V line. If I was to put my meter probes onto any one of those capacitors, I would be measuring all of the capacitors. And it is worse than that. The IC's are also on the +5V line, and so those would be measured as well.

diode aka continuity testing mode.

All you are doing is looking for shorts. Even if its still in circuit, it shouldn't be a dead short. A dead short is enough reason to replace that cap in my opinion.

Sorry half awake, dumb toothache wont let me sleep.

 

[Fire-Flare]

Hello again.

na103: That is exactly what I wanted, but my motherboard is the 16-64KB version which uses two-leg capacitors instead of three-legs. Are the values still the same?

Modem7: The board does appear to be dead. I'll try your link, but I see a dark spot on one of the capacitors which makes me suspect it's gone bad.

twolazy: Sounds easy enough, I'll dig out my meter and see what it says.

 

[modem7]

diode aka continuity testing mode.

On my multimeter, there is both a diode test mode and a continuity test mode.

All you are doing is looking for shorts. Even if its still in circuit, it shouldn't be a dead short. A dead short is enough reason to replace that cap in my opinion.

Sorry half awake, dumb toothache wont let me sleep.

Yes, I still don't think that you are following what I wrote.

Modem7: The board does appear to be dead. I'll try your link, but I see a dark spot on one of the capacitors which makes me suspect it's gone bad.

The procedure at the link is a logical process.
Many with an IBM 5150/5160/5170 have used it to identify a faulty capacitor (or at least to a range of capacitors).
For example, it caters for the fact that a short (if one exists) may not be on the motherboard, but instead, on an expansion card.

na103: That is exactly what I wanted, but my motherboard is the 16-64KB version which uses two-leg capacitors instead of three-legs. Are the values still the same?

Yes.
See [here].

 

[twolazy]

My meter has both on the same setting. Why do you go out of your way to belittle others? Its a cheap harbor freight one till I can fix my Fluke (had voltage leaking to ground on my garage outlets, and zapped something in it).

Here is a video on it, showing ITS THE SAME FUNCTION ON THE METER. Now please if you got nothing helpful to say...

And as far as Modem 7, I understand what you are saying, down the line could be a short and not the cap. But If I knew there was a short down the line I replace the cap anyways... Ofc I would remove the cap, and test again to check for that. Once fixed, yes I would replace the cap.

 

[Fire-Flare]

I'm ready to place the order for new capacitors, but are the smaller yellow things between the RAM banks and near the KB/Cassette ports also capacitors?

 

[VERAULT]

capacitor list? pretty sure all you need to worry about are the 10uf 16v tantalums.. thats the only thing shorted that i repaired to bring my 16-64Kb 5150 back online.

 

[twolazy]

I'm ready to place the order for new capacitors, but are the smaller yellow things between the RAM banks and near the KB/Cassette ports also capacitors?

Yes they are but don't worry about them, they hardly ever go bad. As Verault stated, just replace all the 16v 10uf ones

 

[Fire-Flare]

Alright then, thanks. I'll let you guys know how things go.

 

[Malc]

Modem7: The board does appear to be dead. I'll try your link, but I see a dark spot on one of the capacitors which makes me suspect it's gone bad.

Did you follow the steps of the 'Diagnostic procedure' That Modem7 linked to on his website ?.

The suspect Tant is where on the motherboard ?, You could possibly snip one leg in such a place as to solder it back together again if it turned out to be good.

Keep in mind these boards are multi layer, Depending on your soldering / desoldering tools / Skills there is always the risk of causing damage when removing the Tants.

 

[Fire-Flare]

Did you follow the steps of the 'Diagnostic procedure' That Modem7 linked to on his website ?.

The suspect Tant is where on the motherboard ?, You could possibly snip one leg in such a place as to solder it back together again if it turned out to be good.

Keep in mind these boards are multi layer, Depending on your soldering / desoldering tools / Skills there is always the risk of causing damage when removing the Tants.

I did and got no beeps. My multimeter then said the PSUs I was trying had gone out of spec, so I ordered a new one of those also.

I think the suspect tant is in the lower-left corner in my uploaded picture. (I left it at work for the weekend since they have more tools and components for testing)

I read it's best to snip the old tantalums as high as possible and solder new legs to the old ones. And considering the age of this board, I think replacing all of them will be a good idea.

 

[Malc]

I did and got no beeps

And you followed the rest of the steps to try and find the rail the short is on ?

My multimeter then said the PSUs I was trying had gone out of spec, so I ordered a new one of those also.

I'm not sure what you mean, Do you mean you were doing the resistance checks on the motherboard and your DMM went over range / measured infinity ?

I think the suspect tant is in the lower-left corner in my uploaded picture.

So 'Possibly' a shorted tant on the +5v rail

I read it's best to snip the old tantalums as high as possible and solder new legs to the old ones

That's one way of doing it, Reduces the risk of damaging the board.

And considering the age of this board, I think replacing all of them will be a good idea.

I don't, IMO it's unnecessary, I would just replace the faulty tant/'s and leave the rest alone, The rest will likely go on for years to come.

 

[Fire-Flare]

And you followed the rest of the steps to try and find the rail the short is on ?


I'm not sure what you mean, Do you mean you were doing the resistance checks on the motherboard and your DMM went over range / measured infinity ?


So 'Possibly' a shorted tant on the +5v rail


That's one way of doing it, Reduces the risk of damaging the board.


I don't, IMO it's unnecessary, I would just replace the faulty tant/'s and leave the rest alone, The rest will likely go on for years to come.

My DMM said the +5v and +12v lines from the two AT power supplies I tried are out of spec, so I ordered a recently built Athena Power 300w AT PSU from Amazon.

I'll see what happens with the Athena when it arrives.

 

[modem7]

My DMM said the +5v and +12v lines from the two AT power supplies I tried are out of spec, so I ordered a recently built Athena Power 300w AT PSU from Amazon.

'Out of spec', not no voltage.

If you had a short-circuit capacitor on the motherboard, the power supply would have shut down, producing no voltage.

It seems unlikely to me that you have two faulty power supplies. You cannot test them elsewhere?

It is difficult for my procedure to cater for all scenarios. I am just now wondering whether the particular AT-class power supplies that you used require more of a load than simply an IBM 5150 motherboard. You could try perhaps, also connecting up the 5.25" floppy drive to the power supply in order to provide more loading (and possibly something else as well, such as an IDE drive).

(That is not in the procedure, because it can compromise the procedure, e.g. short-circuit capacitor in 5.25" drive, not motherboard.)

 

[modem7]

Further to last. Something to be aware of in regard to the 16KB-64KB version of the IBM 5150 motherboard. -5V is critical to that motherboard. If down the track, you think of using an ATX power supply with adapter, you cannot use an ATX power supply that lacks a -5V output. A lack of -5V to that motherboard results in what appears to be a 'dead' motherboard (one of many things that can cause that symptom).