• Please review our updated Terms and Rules here

IBM 5151 monitor repair

Skip94

Member
Joined
Jul 23, 2021
Messages
21
Location
Somerset, UK
Hi all, Hope this is ok here.
I've got an IBM 5151 monitor that is exhibiting some strange behaviour.
It partially works and shows what it should for 1/2-1 second at a time, then the picture goes crazy, before settling down again. I'm genuinely unable to describe it properly, so have taken a video and will link it below.


I'm a relative beginner with electronics, but am comfortable safely discharging the CRT and with working around the mains voltage parts of the board.
So far I have visually inspect all the capacitors, had an initial look for bad solder joints and reflowed a few that I wasn't happy with. I have also tested the diodes in circuit. Just to check they have a voltage drop one way and nothing the other, I haven't actually check the specs. One seemed to test bad in circuit, so I desoldered it, but it then tested fine off the board, so I refitted it. This has made exactly zero improvement.
I have tested this on at least two different 5150's with different MDA/MGA cards. All of the PCs and cards work perfectly with my other 5151.
Has anyone come across a similar problem? I could just crack on and start randomly swapping components out, but I'm hoping someone may be able to point me to a specific area to start looking.
Cheers
Andrew
 
Try taking the back off and running it in a darkened room. Look for any arcing possibly around the yoke. That really looks like a corroded short in the horizontal yoke windings. You might have to pull the yoke off and inspect it, especially the windings that are touching the glass of the tube. If the monitor sits in storage for years, the tube glass gets condensation which starts corrosion on the copper yoke winding where it touches.

Larry G
 
Thanks Larry.
I'll try running it with the back off tomorrow, in case there are any obvious shorts. However I think I've ruled out the yoke.
I decided the easiest thing I could do was swap the main and neck boards from one 5151 to the other (which works well), so I did that this evening. The problem followed the boards, so I guess that rules out the tube its self, yoke, signal lead and main power input. Tomorrow I think I'll start by swapping just the neck board.
However, I turned the brightness down a bit to prevent too much wear on the tube and let it sit and run for 10 minutes or so.
After a little while it started to settle down and behave, to the point it was almost normal. There was just a barely noticeable flicker/movement, most obvious on the very right hand edge of the screen.
Something is obviously a little flakey, hopefully I can narrow it down to the main or the neck board, then start a closer inspection.
Cheers
Andrew
 
Found the issue...
Intermittent connection on a dry joint on the base lead of TR23, labeled as the horizontal output transistor in Sams computerfacts manual. Probing around with my multimeter for a couple hours while unoccupied at work and found I could only get a connection when applying significant pressure with the probe. Desoldered the joint to check the pad was ok, resoldered it and the monitor works spot on.
Andrew
 
Hi Skip94,

I was going to make some remarks on your monitor, but I see you have fixed it. The fault you have found is a plausible explanation of the fault you recorded on the screen, but there could be other latent causes.

Firstly though, one thing, please don't believe everything that has been posted or on youtube about the requirement to discharge a CRT before working on a VDU. The sad fact is the person/s who started this urban myth had little knowledge of the physics and anatomy of a CRT.

The electrical charge on the CRT's bulb, between the internal and external aquadag (conductive coatings on the glass) remains trapped there when the VDU is switched off. It cannot get to the CRT's base pins (as it is isolated by a vacuum) it cannot get to any of the LOPT's primary connections, as it is isolated by the EHT rectifier. You cannot come into contact with it in any way unless you poke something under the insulated anode cap or remove that cap. So for general VDU servicing, there is no need to discharge the CRT, unless you want to remove the CRT or LOPT.

But if you do want to remove the anode cap, what is the better way to discharge to CRT ?

Certainly not by shorting it out with a wire.

If you consider the CRT is charged to 13kV, and the arc (plasma that occurs as it flashes over has a low voltage drop, typically less than 1kV), you are discharging a a capacitance of around 500 to 1000pF into a near dead short, say 10kV/10 Ohm or 1000A peak current (albeit brief) This can destroy the internal Aquadag connections to the anode button in the CRT and shower Aquadag fragments inside the bulb.

The better way to discharge the CRT, if you feel compelled to do it, is with a high voltage rated resistor of many Meg Ohms. In a typical CRT EHT probe, is such a resistor, so ideally you would use one of these probes to discharge the CRT. Typically the resistors in them are 100 Meg Ohms or greater.

Enough on that topic, as I have posted on it many times, but those pervasive you tube videos telling people to discharge their CRT's "for safety" remain there like a mis-information disease.

Back to your fault:

If you look at the raster scan while the fault was there, you can see that the horizontal amplitude of the scan appears to be amplitude modulated by what looks like random noise, mainly, though some of the contours look a little sinusoidal.

The amplitude of the H scanning current depends on time. The horizontal output transistor is switched on by a rectangular drive wave, the current rises at a fairly linear rate of V/L amps per second, where V is the power supply rail voltage and L the inductance of the yoke. So any bad connections on any of the three terminals of the H output transistor could generate your fault.

However, as noted in posts above, shorted yoke turns can also do it as this alters the inductance L. Or of course, power supply fluctuations can also as this alters V. Or bad connections anywhere in the H scan output circuit.

There is one other thing that can do it, by causing high frequency instability in the power supply. On a number of computer VDU's the designers were a little remiss in that voltage regulator IC's can go into oscillation when electrolytic caps only are used on the input and output pins of the regulator IC's (ideally they should be ceramic or Tants of at least 2uF) and these aging electros go high ESR with age and this puts HF oscillations on the power supply rail.
 
Last edited:
Enough on that topic, as I have posted on it many times, but those pervasive you tube videos telling people to discharge their CRT's "for safety" remain there like a mis-information disease.
Yeah, sure. And every single service manual ever printed is also wrong about that, because your opinion is the only correct one. No offense, but you put people in danger by what you write here. I can only shake my head when reading such nonsense.

Do you also disable the airbag in your car because you know you won't have an accident while driving to work? No, you don't - because "things" can happen. And for the very same reason, you discharge the tube. Even if you don't plan to put your tools under the anode cap.

Oh, and btw:
https://www.youtube.com/watch?v=8cU9x23w4YE

Show me where he did what according to you is the only way to get a shock...
 
Last edited:
Yeah, sure. And every single service manual ever printed is also wrong about that, because your opinion is the only correct one. No offense, but you put people in danger by what you write here. I can only shake my head when reading such nonsense.

Do you also disable the airbag in your car because you know you won't have an accident while driving to work? No, you don't - because "things" can happen. And for the very same reason, you discharge the tube. Even if you don't plan to put your tools under the anode cap.

Oh, and btw:
https://www.youtube.com/watch?v=8cU9x23w4YE

Show me where he did what according to you is the only way to get a shock...


I'm not suggesting that anybody should deliberately try to go under the anode cap of a charged CRT to sustain a shock.

What I'm saying is, if you don't have to go under there for any reason, there is no risk of acquiring a shock from the charge on the CRT's bulb.

And the fact is, that for most CRT servicing, you don't need to go under there. Often, if you did, it would be with the set running and to measure the EHT with an EHT probe.

In addition, if you do the math (energy stored in a capacitance = CV^2/2) you will find that the energy stored in the charged bulbs of most CRT's is less than a single shock from a farmer's electric fence. It is "unpleasant" and your hands will recoil, but its highly unlikely fatal.

Most manufacturers recommended discharging the CRT before removing it from a set. The reason was, not risk of electrocution, but if you carry a charged CRT, and receive a shock, you will likely drop it on the floor.

In addition, manufacturers of CRT TV's such as Sony, discovered that the high peak currents from the CRT anode, to some point on the chassis resulted in high enough peak currents to destroy semiconductor devices on the pcb's, they detected this as early as 1962. So, if a CRT is discharged it is better done via a very high Ohmic resistance.

I can confidently say in this case my opinion is the correct one. Not only do I have a detailed knowledge of CRT's and their Physics, but I am also a Physician and I would never recommend anything that put a person's health at risk.

EDIT:

I could also add that the stupidity of routinely shorting out the CRT anode connection for routine servicing of the CRT's electronics (not removing the LOPT or CRT or anode cap) has these things wrong with it:

1) It exposes the inexperienced technician to the charge on the CRT's bulb by going under the anode cap, when that was not required for any of the tests on the un-powered pcb components, so, in fact, it is less "safe" in terms of whether a person receives a shock from that charge or not. So the people who suggested doing this "routinely" for servicing actually increased any absolute risk, and clearly didn't understand the problem.

2) As noted the initial peak currents from the discharge can be destructive to components, including the CRT;s internal anode button to the internal conductive coating's connection to that button.

If you are struggling to understand why the above is correct, you need to consider the design of the CRT. The two plates of the "capacitor" which stores the charge on the CRT's bulb are the external conductive coating and the internal conductive coating and or aluminization there, this is connected to the metal anode button in the CRT's bulb and to the final anode connection of the CRT's gun structure, this is isolated from all the other CRT electrodes that exit the CRT's base.

The glass of the bulb is the dielectric. The reason the CRT stays charged after the set is switched off is because the internal coating (one plate of the capacitor) effectively connects nowhere, because of the isolation provide by the EHT rectifier (which has very low reverse leakage current, if it did not the charge would quickly bleed away).

The only way you can acquire a shock from a charged capacitor, is if you can connect yourself across its two terminals. You cannot connect yourself to the CRT's internal coating unless you go under the anode cap with your finger or a tool. If you don't go under there, the charge on the bulb stays safely isolated there, Even if you removed the CRT and the LOPT together with the LOPT swinging off it by the EHT cable and cap, the charge is trapped there. Since semiconductor EHT rectifiers don't have completely zero leakage, this is why if you leave a CRT set turned off for some days or weeks, very slowly due to tiny leakage currents the charge does dissipate on its own.
 
Last edited:
Back
Top