• Please review our updated Terms and Rules here

IBM 5154 repair and recapping

B

BakaOsaka

Member
Joined
May 5, 2020
Messages
24
Location
United Kingdom
Hi,

I recently acquired an IBM 5154 in great looking condition. It was marked as "untested" but a number of other things I bought from the same seller were also "untested" and worked fine, so I'm fairly confident that I wasn't ripped off.

I plugged the 5154 in to power and turned it on. The power LED didn't illuminate and instead the monitor just made a loud whining noise. I wasn't sure if this was the actual CRT or something else, so I disassembled the monitor and tried again. The whining is clearly coming from the PSU.

I disassembled the PSU (Bloody rivets!) and saw some residue on the inner casing. I assumed this was from an exploded capacitor but they all appear to be fine - No obviously blown tops. I'm planning to replace the caps at C11, C13 and C14 as well as the two large caps in the middle of the PSU.

I've uploaded a video of the buzzing here:
https://drive.google.com/file/d/1d1Qquj2EvSD9DmZIgxtJT2NoUbpibcyx/view?usp=sharing

I've also uploaded some photos of the board and the residue here:
https://imgur.com/a/Oi59sYl
 
Hi,

The first thing to do is not to jump to conclusions about what could be wrong and start shotgun replacing parts, like capacitors. Every time you make an intervention you add some risk of introducing a new fault and you could end up in a real mess with many replaced parts and still no further ahead. Most likely for example the large sized two pale blue main filter capacitors will be ok, often these physically bigger types are much longer lasting than smaller types and possibly even better quality than some new replacements you might buy.

The first thing is to get the 5154 manual and to start collecting information on the components in the set.

With the psu removed from the unit, or with the power off initially, check the ESR of all the electrolytic caps, in circuit, with an ESR tester, a good one like a Bob Parker designed Blue Anatek type, these put such a low level across the capacitor that in circuit tests are not corrupted by associated components.

Then with a meter check all of the semiconductors you can for shorts or being circuit. Depending on the circuitry, occasionally a leg or two of a device needs to be disconnected, the least destructive way to do this is to remove the solder around the lead and pcb pad with solder wick. Most of these older phenolic pcb's do not have plated through holes, so its quite easy.

Keep a record of anything tested. If you cannot identify any obviously failed part initially, the next move is to go to testing with the unit powered up, using both the meter and scope, to track down the fault that way. Its fairly safe to use the meter on the primary side for testing say the output of the bridge rectifier and main filter caps, because the meter is isolated, however if a scope is used here its risky, unless its a special scope like the Tek222ps or you isolate the power supply and or scope with an isolating transformer. The normal scope is fine for anything on the secondary side of the power supply. The voltages here should be checked with the meter too.

I cannot hear the audio on the computer I'm using, but will listen to it later. There is plenty of scope for whining noises coming from either the psu or the scan stages in these sets. CRT's them do not often generate any sound, the inside of the CRT is a vacuum and the metalwork of the gun is non-magnetic on the whole , occasionally there can be tiny sounds tick like sounds emitted as the heaters warm up. That is one of the first things to check too, are the CRT's heater's running ?
 
I have listened to the sound.

You need to try to identify the physical place where this sound is coming from. It has some features of high voltage arcing, which is worrying.

(If a CRT has gas in it, or let down to air, it can get internal arcing, but usually if that was the case there would be a spectacular blue light show in the CRT's base and neck, so its probably not that and generally to get that requires the exhaust tip is snapped off the CRT near the socket or a cracked neck and its fairly obvious).

The high voltage points are on the CRT's final anode and the EHT cable leading to that from the high voltage module and to a lesser extent on the base assembly where the focus voltage is introduced.

If the noise is coming from the power supply unit, one possible explanation is that is running way low in frequency and the sound is coming from the transformer's core.
 
Thanks for the info. I'm going to try and do more investigation - The noise is definitely coming from the PSU rather than the CRT. I also can't see the CRT heater lighting - The lack of power LED on the front makes me think that the PSU isn't providing any power to the rest of the monitor.

I've got an ESR tester that I will use to check the capacitors and I'll replace any that give suspicious values. I take photographs as I go along so hopefully that ticks the "documentation" box. Unfortunately I don't have an oscilloscope yet, so I will cross that bridge when I get there.

One of the things that made me suspect the capacitors is that when powering on the monitor it squeaks for a second or so in the same way that my 5160's failing PSU also squeaks, caused by old capacitors. I suspect they've dried out and that could be the cause, I know jumping to conclusions is never the best solution! However, minuszerodegrees.net mentions that both the LED and white raster would be visible following power up, and that capacitor failures in the PSU is a known issue. I've also found that this source states that a chirping noise can be caused by the capacitors.

I will investigate further and keep you updated.
 
The X1 line suppression cap in my 5154 exploded, as they often do, but there was nothing wrong with the electrolytics. Your X1s are still clearly unexploded
 
While I agree with Hugo Holden, I think replacing 35+ years old caps is a good idea. It's unlikely that replacing the caps will introduce new issues. If at all, you can rule out one probable cause of the actual issue you have with the monitor.
 
BakaOsaka said:
Thanks for the info. I'm going to try and do more investigation - The noise is definitely coming from the PSU rather than the CRT. I also can't see the CRT heater lighting - The lack of power LED on the front makes me think that the PSU isn't providing any power to the rest of the monitor.

One of the things that made me suspect the capacitors is that when powering on the monitor it squeaks for a second or so in the same way that my 5160's failing PSU also squeaks, caused by old capacitors.
I will investigate further and keep you updated.
Click to expand...

Sounds sensible.

That squeak for a second may possibly be the power supply attempting to start but being overloaded, say, from circuits in the monitor overloading it (the typical failure here being a shorted H deflection transistor) and as I recall for the 5154, there are two transistors here that are possibilities, because they split the job of EHT generation and H deflection in two, to allow a stable EHT when the monitor changes from CGA to EGA scan rates. So the noise could represent some rapid cycling in a PSU shutdown mode.

While I also generally agree with Timo W. about 35+ year old electrolytic capacitors , it is not the actual birth date of the capacitor that determines if its likely to need replacing, it is the numbers of 1000's of hours the capacitor has been in use with the monitor, with the monitor running. The fact that your particular 5154 looks so clean and free from the usual dust and soot that is attracted to places like the EHT cap by electrostatic forces and around the focus module, and that none of the small electros on the PSU have dome tops (indicating gas pressure), I suspect that most likely nearly all the electros in your monitor are ok (just a hunch, but I could be wrong), except perhaps the one that is right near the vertical deflection output IC that unfortunately dies very early due to heating in the 5154.
 
Last edited:
Caps have a shelf-life as well. Even if unused, the rubber gasket won't stay intact forever causing the cap to dry out eventually. At least they need to be checked - and while this can be done partly in-circuit, you only get meaningful results when testing the caps unsoldered. But then you have already done 50% of the work needed for a full re-cap. In the end, if you want to preserve the monitor, one would do a re-cap anyway.

Now don't get me wrong, I have systems as old as 1983 that are still using their original caps and I don't have any reason to renew them. That is, as long as these systems still work. As soon as they show some fault, replacing the caps will be the first thing I'm going to do, just because of maintenance.
 
Timo W. said:
Caps have a shelf-life as well. Even if unused, the rubber gasket won't stay intact forever causing the cap to dry out eventually. At least they need to be checked - and while this can be done partly in-circuit, you only get meaningful results when testing the caps unsoldered. But then you have already done 50% of the work needed for a full re-cap. In the end, if you want to preserve the monitor, one would do a re-cap anyway.

Now don't get me wrong, I have systems as old as 1983 that are still using their original caps and I don't have any reason to renew them. That is, as long as these systems still work. As soon as they show some fault, replacing the caps will be the first thing I'm going to do, just because of maintenance.
Click to expand...

I don't think there is anything wrong with your philosophy about replacing old electros, though I put a different set of conditions on doing it, one is a time line slice. I replace most of them in semiconductor based equipment if they are older than 1970. That includes TV's monitors radios etc.

However, if the item is 1980 or newer (currently) I decide on how many hours the unit might have done. A little detective work helps here (as Mr Monk would say, I might be wrong but I don't think so), looking for signs of age in use. In the 5154, the video output transistors and the vertical output IC run hot, so, if the phenolic boards are darkened in these areas, you know the monitor has had a lot of use, it not likely has very low hours in that case.

Take for example the 5154 that the OP has. It is very clean inside and I think has very low on time hours. I know these monitors very well. It will be interesting to see how many electrolytic caps are actually defective in this one, I would guess only one or two at worst. So if this were my monitor I would only seek to find, on testing with ESR and my capacitance meters, any that were definitely defective or out of spec. I wouldn't on this set do a global re-cap.

If the set was a decade or two older I would consider a global re-cap and use high quality Japanese Nichicon caps to do it.

But there is no golden rule here, each piece of electronic apparatus you restore requires individual assessment.

Also, many professional SMPS apparatus now, take the VFD (variable frequency drive) for AC motors, the CPU monitors the on time, to give the owner a recommended time to replace the main filter caps, so the notion of capacitor life vs ON time is well established in the electronics industry.
 
Last edited:
I've not been able to do further diagnostics on it yet, however I did take the monitor apart to try and test the PSU in isolation.
Disconnecting it from the main CRT board and running it with just the two wire connectors gave the same results - Hopefully this is fine? - so I believe it's not due to anything on the CRT board. I also removed the PSU's case and powered it up, recording another video that shows the sound better:
https://drive.google.com/file/d/1diTBIwt4-AWE6qvQPR-UctBKwx9ynqQz/view?usp=sharing

I can't identify exactly what component the sound is coming from, and I'm trying not to power it up too much to avoid damaging it too much. However, there is a strong smell of burning electronics coming from it when it's running and rewatching that video it looks like there's a small waft of smoke coming from the bottom-left of the board.
 
Last edited:
Smoke/heating of a component possibly D18 lower right on the supply's output side. Check the diodes and capacitors there for shorts. I will look at the schematic and make some suggestions.
 
Thanks, will open it up again over the weekend and check the values.
For reference this is a UK 5154002 model - I don't know if the schematic is different from the one you're referencing.
 
Check diodes D17, D18 and D19 & D20 for shorts.


Check C25, C29 , C27, C31, C33. Also check the small value 330pF capacitors that are in parallel with D16, D17 & D18.

There should be no power drawn from the output with that unplugged. So if something is heating up, it could either be a shorted diode (applying AC to the filter cap) or a shorted filter cap. Its also possible but less likely that IC12, the 7812 regulator has a shorted input, but not impossible. You should be able to find the issue with the Ohm meter.

Electrolytic capacitors can go shorted as a failure mode. All it takes is a small amount of metallic debris at the factory that gets into the paper insulation between the foils and they short many years later. I have a good photo of this somewhere..attached.
 

Attachments

  • capshort.jpg
    capshort.jpg
    94.7 KB · Views: 8
Last edited:
From the lack of dust and darkening of the power supply board, that monitor appears to be in good condition, and wasn't heavily used. I would start by looking for shorts on any of the power supply rails. An overload will cause the power supply to shut down.

It wouldn't hurt to replace all of the small electrolytic caps in the power supply and video modules since they have a history of failre, but I have a feeling it won't fix this problem. The large caps are probably fine, but check underneath them for signs of leakage.

Try monitoring each output voltage from the power supply when you turn it on. Look for any that either don't come up, or jump too high. Power supply shut down is usually caused by either an overload, or a loss of regulation.
 
Hugo Holden said:
Check diodes D17, D18 and D19 & D20 for shorts.
Click to expand...

Hello,

I'm coming to this thread because I have a 5154002 with a very similar issue (the noise is very close as far as I can tell).

I'm a complete noob to this (I'm trying to make the old PC from my childhood work again). I think D17 and D18 are shorted, but I'm not sure about the specs for replacement parts. Should I just remove them and
find out, or is there a part list available somewhere?

IMG_20201226_142804911_HDR.jpg
 
I actually proceeded and replaced D17, D18, C26 and C28. It did not work but I realized C25 and C27 were shorted to ground and it actually power on now. The screen struggles to sync with the EGA output of my 5162 but it gets OKish after a while, so I will try to not mess with it too much now.

Thanks to Hugo Holden for setting me on the right path, that was fun.
 
"Olosta" You'd have to remove them to check for a short or just one leg at least. No way of telling in circuit if the diodes are shorted or something else.
Replace them with off the shelf 1N4004 if it a less than 250V PSU. I think a short would almost certainly blow that fuse though.
What's wrong with the power board anyway ? I don't think you should playing around with that board with powered either :|
 
Last edited:
You must log in or register to reply here.
Back
Top