IBM XT 286 5162 finding short on motherboard

[ia2115]

Hey

When i measure the motherboard alone, with continuity between power connector +5v and to ground i get a continius tone like there is a short same with other +5v and with -5v, the 12v seems ok not short there. how would i find this short the best way, without having to desolder all tantalems and measure dem out with esr meter?

 

[modem7]

If there is indeed a short-circuited tantalum capacitor on your 5162 motherboard, then the motherboard will overload the power supply, resulting in the power supply shutting down, the power supply providing no +5V/-5V/+12V/-12V output. But sometimes, an underload (which includes no load) will result in no +5V/-5V/+12V/-12V volts.

Is it the case, that with your 5162 motherboard connected to a known good power supply, the +5V line does not measure about +5V ?

If you do measure about +5V, there is no short-circuit tantalum capacitor.

If instead you measure 0V (with only the 5162 motherboard connected to the known good power supply), then we need to work out why.

When i measure the motherboard alone, ...

I know that you used the word "alone", but to be sure, note that the power supply must NOT be connected to the motherboard when you measure the +5V rail of the motherboard.

... with continuity between power connector +5v and to ground i get a continius tone like there is a short

The purpose of the continuity setting on your multimeter is not to alert you to the presence of a short-circuit. Instead, you should be using the resistance setting.

The resistance of the +5V rail on a 5162 motherboard will vary from motherboard to motherboard, and you may even get a different reading each time that you measure it, but it will probably be in the order of a few hundred ohms (based on what I measure on a good 5170 motherboard of mine). If instead, you measure between 0 and a few ohms on the +5V rail of your 5162 motherboard (power supply not connected), then that would indicate a short circuit.

 

[Hugo Holden]

Hey

When i measure the motherboard alone, with continuity between power connector +5v and to ground i get a continius tone like there is a short same with other +5v and with -5v, the 12v seems ok not short there. how would i find this short the best way, without having to desolder all tantalems and measure dem out with esr meter?

The data presented here suggests there may or may not be any problem. Most digital meters sound a tone when the voltage drop across their terminals is lower than a couple of hundred millivolts, deemed to be that from the current generated by the meter and the apparent load resistance.

On a meter like this, it might in fact be normal when you test across the board power supply terminals.

So beeps from a meter are not very diagnostic.

 

[Ruud]

What I miss in your story: did you try to power on the board? If not, try that. Tip: do it outside your house. If there is a short, the PS will shut down itself. If there is a weak tantaal, it will go BANG and you have found your culprit.
In caste the PS shuts down, desolder the tanaals one by one and power the system up after each removal. In this case you have chance you will find it before having to remove all tantaals.

Side thought: is it possible to find a weak tantaal with just measuring its resistance?

 

[ia2115]

with ohm i measured 14.7 ohms on ground to +5v (P4-5-6 on P9) and i cant measure any ohms on -5v. on the P9- Pin3 connector

 

[daver2]

Read the resistance with your multimeter probes both ways round.

5V into 14.7 Ohms equates to only 5/14.7 = 0.340A worth of current (assuming a dc linear relationship). This current is nothing when compared to what the card should draw under normal circumstances...

Give it a go...

The worst that could happen is either the PSU shuts down (due to an over current condition) or a tantalum capacitor goes into volcano mode! Or the card is already dead...

Dave

 

[Ruud]

"Volcano mode", I must remember that

And with "measuring its resistance" I meant measuring it AFTER the tantaal has been desoldered. The first seconds you will see something because the multi-meter is loading the tantaal. But it should be an increasing value. After that you should see the "infinite" value.
Second thoughts: it is possible that the weak voltage won't do anything with the tantaal but that 5V is already enough to make it BANG. Third thought: test it with 5 or 12V after it has been desoldered. You can even go as far as the voltage stated on the tantaal minus 20%.

 

[daver2]

Or a “Smurf Grenade” as Falter calls them !

I had a white hot piece of tantalum whizzing around an enclosed cubicle once when a solid tantalum bead capacitor had the power supply accidentally reversed to it. That was fairly spectacular!

Dave

 

[Malc]

with ohm i measured 14.7 ohms on ground to +5v (P4-5-6 on P9) and i cant measure any ohms on -5v. on the P9- Pin3 connector

14.7 ohms on the +5V rail is fine, The -5V rail should have a reading in the M ohm's range but a lot of DMM's go "Out of range" as they can't go high enough, Usually you see the scale rise in the M ohm range and then drop to 0 or 1 or OL or whatever dictates "out of range" for your meter, If the caps are charged you may not see it move at all.

It's a good idea to first check the resistance of your DMM leads, Some of these cheapo leads can have a lot of resistance in them and you need to deduct the lead resistance from the Ohm's value you got when you measured from the board connectors. So for example say your DMM leads has 14.7 ohms resistance and you measured 14.7 ohms at the board, 14.7 - 14.7 = 0 ohms which would indicate a definite short on that rail.

Here's the measurements i got from my Good working 5162 ...
All cards removed and everything disconnected.
P9 connections at the motherboard
Pin 6-5-4 +5v = 14.65 ohms
pin 3 -5v = >20 Mohm's. The meter i used went out of range when it reached it's max.

P8 connections at the motherboard
pin 4 -12v and pin 3 +12v = >20 Mohm's, The meter i used went out of range when it reached it's max.
pin 2 +5v = 14.65 ohm's
pin 1 PWR GOOD = a bit over 10K ohm's

 

[hbsoftware]

if in doubt i would use a current regulated bench supply to supply 5v to the 5v rail ...slowly raise the current and if nothing explodes or gets hot or starts current limiting by dropping the voltage then its good to go.... but maybe not worth it since the 14 ohms is the same reading as a good board apparently

 

[ia2115]

i put the power on the motherboard, and indeed all the "problems" i tought i had, wasnt anything and the board works as i see it flawless. thanks though for all the inputs as they can be reused on next project.