I always feel stupid mentioning this because we are all adults but the only difference from a normal electronic device and a crt is very high voltage , I worked on them as a teen very carefully while they were live, plastic tools and caution will ensure you don’t get shocked…

Not the correct way to fix this but if your brightness knob works you could turn it down then internally mess with the cathode voltage and other pots.

There is a small possibility that your contrast dial is oxidized, internally you may be able to test it and apply deoxit, the fact it’s random the way it is , it’s very possible a component other than the contrast dial is either failing or has a cracked solder joint.

Only way to know is to try manually bypassing the contrast dial to see what happens then poking around tracing the circuit from it to see if anything strange is going on.

I’ve had cases where if I (using appropriate protections) touch a component with a probe on a live screen the colors would flake out, then would find the damage.

Sometimes with the screen off and discharge a simple but detailed manual examination of the board and components will find the damage, traces, joints, caps etc.

Good luck, hopefully it’s just some oxidation

Oldies but similar general crt stuff

IBM 5151 Monitor Repair

https://www.repairfaq.org/REPAIR/F_monfaq4.html#MONFAQ_003

I'm unsure whether it is a photo issue, but it seems your issue is isolated to red. Your photo shows nice colors for blue, green and cyan (which is blue+green), but very dark red, pink and brown. Note that IBM brown has its normal red level, but reduced green, to make it a more reddish brown instead of a dirty looking "dark yellow". All the bright colors look blown out in your photo, but this is either because you increased the brightness/contrast to an insane level so you could see red at all, or it is due to the photo being overexposed, both of whicht are not problems in your monitor.

Anyway, my conclusion of "only red is low" is that this is very likely not a contrast pot issue, because that would affect all three color channels the same way. While audio amplifiers often have a stereo pot for volume, in which one of the two channels might be broken, while the other one still works, disturbing only one channel, monitors do not use three-channel pots that operate on R,G,B separately, but they influence the reference voltages the digital-to-analog conversion in the monitor operates on, and there is one shared reference voltage for all three color channels. So the issue is "anywhere downstream in the signal processing chain", which is to be frank, not very specific. In the 5153 monitor, which likely has a similar construction, the digital RGBI inputs end up being processed by two 7406 open-collector inverters, which then drive the analog electronics that generates the colors, which are then amplified by the cathode driver. Anything behind the output for "red" on the 7406 might be wrong on the red channel.

Assuming your issue is a bad solder joint (which is a reasonable assumption, but not certainly the cause), I suspect you should be able to coarsly locate the fault by tapping the monitor PCBs with a plastic rod if you operate it with an open case. It's also possible that this is caused by a bad connection which is not a solder joint, but a connector, which can not only be cable connectors inside that monitor, but also the connection of the neck board to the tube. The neck board is the PCB that is plugged to the end of the tube.

There is high voltage in the monitor, so don't use conductive tools, and don't touch anything while the monitor is running. Don't disconnect the big red anode cable from the CRT, that's not your problem, and that's where the highest voltage is. Other parts still have several 100 volts, and the cathode driver circuit that might be at fault still operates at around 100V in typical color monitor designs, so there definitely is a moderate chance to get very unpleasant or even injuring shocks if you are super careless, though.

Oh wow. Thanks for all the detailed answers rmay635703 and mkarcher! Much appreciated.

Going through rmay635703's attached materials, I definitely should have a close look at the IC201 and the capacitors in the neck board. And mkarcher, you are correct, the most significant color issues deal with red, which is especially apparent in games, where it comes off looking more like dark brown. But I think there are still issues with blue, especially when the monitor has been off for a couple months and I turn it on cold with only 64 kb of memory to load. It's virtually black, along with most of the other dark colors.

I'm going to include a picture of that, along with several more pictures of the problematic monitor, but this time photographed alongside my second PCjr monitor, which doesn't have the same issues. I've tried to be more mindful of the colors this time. It's funny that I didn't think about color accuracy when posting my last picture, despite expressly needing help on my monitor's colors, 🤣.

Anyhow, Comparing them side by side, I've noticed that even when the screen is white, the monitor colors are noticeably different. And when the basic screen loads, the problematic monitor shows gray text, while the normal white is bright white.

Looking at the newer pictures, it seems like rmay635703 is on something, the issue indeed looks like excessive contrast, and the contrast pot not working fits in as well. In this kind of digital monitors, the brightness typically affects both "bright" and "dim" colors, while the contrast setting sets how much of the brightness remains for "dim" colors, but leaves the "bright" colors mostly alone. Your first picture shows the "bright" colors approximately the same on both monitors (by the way, including the working monitor as reference to know what issues are picture artifacts was a great idea!), while the "dim" colors are too dim.

Your observation that the issue is worse after the monitor was off for several months hints to capacitors as a possible root issue. Capacitors are supposed to not conduct DC current, but if they are worn, they may get electrically leaking (I don't talk about physical leaking of electrolyte!), and develop some "leakage current" as if a resistor is in parallel with the capacitor. I found http://bitsavers.informatik.uni-stuttgart.de/ … rence_Nov83.pdf , which has the schematics for your monitor on page B-29 (PDF page 527), in which at least the color input circuit looks nearly indistinguishable from the IBM 5153 color display. I did not do a side-by-side comparison, but at first glance, I did not notice anything that differs between the color input circuit on the 5153 and your monitor, including the mostly, but not fully consistent offset of "30" between the components numbers used on the three color channels.Possibly parts of the schematics, or even everything except the case is identical. Looking at those schematics, there is no potcapacitor near the contrast pot, so while it looks like a capacitor issue, it doesn't seem to be a capacitor issue.

The idea of the contrast circuit in that monitor is that the intensity signal from the PC arrives at pin 9 of the IC called Q253 at the left edge in the top half of the schematic. If intensity is "high", the IC pulls pin 8 to ground, so there is effectively no voltage at the contrast pot, and transistor Q251 is turned fully off. The output of pin 8 of IC Q253 is also fed back to input at pin 5 of that IC. If pin 5 is pulled low, pin 6 is "high impedance", i.e. the IC does nothing to the voltage at pin 6, so resistor R216, which has the +5V supply on top (quite difficult to follow) turns transistor Q252 partially on. For low contrast, this is exactly the opposite way. As Q251 and Q252 have their collector connected directly to the +5V rail, this is a "common collector" configuration, also known as "emitter follower" configuration, and the higher base voltage at Q251 or Q252 determines the emitter voltage at R232, which is then fed through the service switch before it reaches the R DRIVE, B DRIVE and G DRIVE pots. At the same time, the output of pin 6 of the IC Q253 is also forwarded directly into the transistor circuit generating the colors via the diodes D501, D531 and D561, so there will likely be some contrast even if the circuit around Q251/Q252 does nothing.

Only the Q251/Q252 circuit depends on the contrast setting, so if the contrast pot "does nothing", the issue is likely very close to this circuit. As there are no possibly leaky electrolytic caps near this circuit, I don't think capacitor failure is the primary issue. As I do not yet undestand how the color generation circuit works in detail, I am unsure what problems will occur if the service switch S301 gets flaky. As that is a mechanical part with metal "just touching" to provide connectivity, it is prone to oxidation, and it is located in the circuit that is meant to respond to the contrast pot. You should be able to verify whether S301 provides "good connectivity", i.e. less than 1 ohm, with the monitor turned off, assuming you have a multimeter. If you don't have one, diagnosing will be way more difficult, but you can try toggling that switch multiple times anyway and see whether this helps with your issue.

The "low red" issue I observed is likely a different issue: All colors are too low at low intensity (which is the intended behaviour at contrast set to the maximum), and possibly red is generally too low compared to the other colors, which also explains the blueish tint of the white screen. This is likely to be an entirely different issue than the excessive contrast, and should be fixable by adjusting the "R.BKG" pot. There is dangerous voltage around the bkg pots, so make sure you use an isolated tool when you adjust these pots while the monitor is running.

EDIT: fixed the claim "there is no capacitor near the contrast pot", which originally read "there is no pot near the contrast pot".

Oh my goodness, thanks for the thorough response mkarcher! You've given me quite a lot to work with, and a bit to study up on too.

It's funny. I actually have that technical manual, but I never saw the monitor schematics inside. I remember looking for it once, but gave up, figuring it wasn't there because none of the diagrams said "monitor" when you flipped through it. Turns out it was folded up inside a different schematic. I wouldn't have bothered looking there if you hadn't of told me the location, 🤣!

Fortunately, I have a multimeter, so I should be able to test out those components. Unfortunately, I don't think I have any plastic or insulated tools like you two suggested, so I may need to hold off from any serious work until I get some. I'm going to check with a friend who does some computer work to see if he has anything.

In the meantime, I might just open up the monitor to visually inspect the inside and see what's easily accessible and what isn't. I've seen a couple videos of people taking the PCjr monitor apart, but it always seems a bit different when you're doing it yourself, 🤣! Thanks again for all the help!

Muckrake wrote on 2026-02-19, 02:47:

Fortunately, I have a multimeter, so I should be able to test out those components. Unfortunately, I don't think I have any plastic or insulated tools like you two suggested, so I may need to hold off from any serious work until I get some. I'm going to check with a friend who does some computer work to see if he has anything.

You can easily test whether the connections from the main board to the contrast pot and the service switch make good contact with your meter while the monitor is turned off. Keep in mind that a pot turned to one extreme is supposed to read a very low value between two of the pins. The contrast pot is a 500 ohm pot, and thus at the extremes, it should have less than 5 ohms between the wiper and the corresponding end of the carbon track. Thus it should be perfectly safe and quite easy to test my current hypothesis, that the contrast issue is caused by a bad connection around the contrast pot.

Sorry it's taken me so long to get back.

I opened my PCjr monitor Saturday and found a couple of surprises. First, the service switch, which was easily accessible, wasn't labeled S301 like it was in the schematic, rather it was labeled S691. I'm assuming that shouldn't make too much of a difference in our testing. But second, there appears to be two RGB drive pots for red and blue, with the red labeled VR6R1 and VR6R2. If that is the case, which one of those should I adjust (with the insulated tool) to fix the red issue? Or have I just misidentified these?

In potentially more productive news, I did test the service switch with my multimeter, getting the result of 19 ohms. I placed the multimeter's probes at the middle and left pins (see Pic3) of the service switch. No other combination produced any results. The multimeter's ohms setting was placed at 2,000k. At 200k I got 18.8 ohms. Do these readings make any sense? Did I do this right? Let me know.

Anyhow, the only thing I am certain of is my discovery of the horizontal centering switch. I didn't mention it before, but the monitor's picture was off center, probably by a couple inches, positioned almost all the way to the screen's right edge. The switch sent it more toward the center, but not perfectly so. I don't believe that monitor is capable of perfect centering.

Did you clean up the contrast dial?

A word of warning Old pots can get stuck or have hot glue applied.

Make sure to clean them a bit before you mess with them then ease in as I’ve broken a few.

Next you will want to run a program with a color test pattern.

after having your contrast and brightness set into a neutral position you will want to start leveling out the drive voltage first , there are prescribed ways to do brightness and contrast but I prefer to set just below the black level illuminating (depending on how tired your tube is)

Once your white intensity and black level are set then you can fine tune the color pots.

You likely have nothing to reference but could use your other screen with the same pattern set.

If red is the issue start there until you align its appearance to the other screen, remember that a dying screen may not be able to come back into spec so you do your best. RGB are all relative dials so you can intensify red also by turning down GB.

If gain goes too high on one color you can strange side effects.

Good luck, analog it’s all relative

One of the fbook groups I belong to has had several members find box fulls of contrast dials

https://www.facebook.com/groups/ElectronicPar … S40aB7S9Ucbxw6v

Thanks for the input and resources rmay635703!

Unfortunately, I haven't cleaned the contrast dial yet. It's at the front of the monitor, so to get access to it, I'd have to do a fair amount of disassembly. I don't think there's a separate contrast pot on the neck or main board -- at least that I can find. On the front, the contrast dial turns fine. It just doesn't do anything.

I plan to access it as some point, but the insides of a monitor are little more complicated than what I'm used to, so I'm proceeding cautiously. That's why I wanted to test what I could before I got there.

I had a monitor with a similar issue, but it was a more modern CRT from 1998. I didn't get to really fix it before it went totally kaput for a different reason (power supply issues) and gave it away. Nearest I could tell it seemed to be an issue with G2/screen regulation. After being off for a while it would act like the screen control was turned way down, and after warming up it would reach a point where it started "blinking" into proper operation then stabilise - It wasn't a gradual increase. Occasionally when I turned on the monitor the image was blank. I got the feeling a component was dying and reforming somewhat after being exposed to voltage again.

The interesting part was that since it was a more modern screen it had standby, and as long as the screen was kept in standby it would turn on instantly like normal. That indicated to me it was a failing component that received power while in standby, but this older screen probably gets hard switched every time.

Thanks for the info NeoG_! A failing component is probably what I'll ultimately be dealing with.

So I've been researching some more information on RGB color pots on CRTs and found out that the three color pots under the "cut-off" label (as pictured in my earlier post) on my PCjr monitor are probably for bias (black level), and the two color pots under the "drive" label are for gain (white level). So I'm assuming it's the red drive pot that I should be adjusting. Does that sound correct mkarcher? Also, did the 19 ohms fit with hour hypothesis?

And rmay635703, I noticed when I clicked on your link for the contrast pots that those looked different from what's on the front of my jr. In fact, those dials look like what I have in back of my jr to control the vertical height and hold. Just thought I'd mention it in case that would cause an issue. When I do make it to the contrast pot in front, how do you recommend I clean it? I imagine I'd have to desolder it from the board first, but I know you mentioned just spraying it with deoxit before trying anything else.

Thanks again for your help guys!

Muckrake wrote on 2026-02-24, 02:56:

So I've been researching some more information on RGB color pots on CRTs and found out that the three color pots under the "cut-off" label (as pictured in my earlier post) on my PCjr monitor are probably for bias (black level), and the two color pots under the "drive" label are for gain (white level). So I'm assuming it's the red drive pot that I should be adjusting. Does that sound correct mkarcher? Also, did the 19 ohms fit with hour hypothesis?

I expect that you mailny need to adjust the cut-off pot. The "drive" pot is used if bright red isn't bright enough, but dark red is bright enough. The cut-off pot is used if both bright and dark red is too dark. 19 ohms is not really OK. A mechanical switch is supposed to make "good contact" which is below 1 ohm. Before jumping to conclusions too early, please do a null check first: if directly connect the probes (that is, you hold the together), the value should be "close to zero". This is the zero level of your meter. For typical meters, the zero level is around 0.7 ohms. Professional meters allow to calibrate that level. If the contacts of the probe leads or the jacks in the meter are corroded, you get a random zero level, which might even be above 40 ohms, and low ohm measurements are completely unreliable.

The service switch is a dual-pole dual throw switch. If you label the pins like this

11 2 3
24 5 6

in one position, pins 1-2 and 4-5 should be connected (below 1 ohm), and in the other positions, pins 2-3 and 5-6 should be connected. If this isn't the case, try toggling the switch multiple times to scratch away internal corrosion.

Well, I got some answers, but I don't know if they're the right ones.

First I did a null check, and crossed the two multimeter probe needles. This got me 0.00, so I think that's OK.

Second, I toggled the service switch to the left (when looking from above like in my earlier pic), however, it didn't register anything when I touched pins 1-2 and 2-3. But I did get a 19 when I touched pins 1 and 3 with the probes. When I turned the computer on, it acted as normal. I couldn't tell the difference. Everything operated the same as when the switch was toggled in the middle.

Third, I toggled the service switch to the right, however this time I didn't get any reading no matter which combination of pins I touched. When I turned the computer on, the monitor produced one horizontal blue line in the center.

I don't know if this helps, but I found some more accurate schematics for the monitor online. From what I can tell, the numbers match what I see inside.

Thanks again!

Muckrake wrote on 2026-02-25, 02:50:

I don't know if this helps, but I found some more accurate schematics for the monitor online. From what I can tell, the numbers match what I see inside.

Oh, that's interesting. These schematics show a similar, but clearly different monitor than the one in the technical reference manual. Remember how I said I didn't notice any difference between the PCjr technical reference manual and the IBM 5153 regarding the locations I looked at? I guess this is because they just put the IBM5153 schematics in that manual, possibly before they decided to build a specific PCjr monitor.

Forget the service switch. The service swithc in the schematics you posted seems to only act on the deflection circuit, but not on the color circuit. But most importantly, I said the "contrast issue looks like an electrically leaking cap", but I had to reject the theory because the schematics from the technical reference did not show a cap near the contrast pot. But now, look at this schematic: There is C203 exactly at the position that would cause the broken contrast issue you observe! If you have a replacement capacitor at hand, I suggest you replace C203. Specific value is not important. While this cap is specified as 220µF/10V, I suspect the monitor would work equally well with anything between 100µF and 470µF. Higher voltage is also fine, so if you have somehting like 100µF/25V or 330µF/16V at hand, you can try those as replacement for C203 as well.

Oh, wow. Thanks again mkarcher!

I should only have to discharge the monitor, disconnect the neck board, remove the shield, and desolder and replace the capacitor. Much easier than what I thought I was going to have to do, which was taking the whole monitor apart and digging out the front circuit board.

Now, about the capacitor, the closest I have is 100µF/50V. Think that will work?

Man, I really appreciate your help. You really went above and beyond analyzing the schematics like that!

Muckrake wrote on 2026-02-26, 00:38:

Now, about the capacitor, the closest I have is 100µF/50V. Think that will work?

If it fits mechanically and you take care to not insert it backwards, it will work. Electrolytic capacitors only work well in one polarity, so it is important to connect the "-" and "+" connections the same way as the old capacitor was installed. I can't guarantee that swapping C203 will work, but if C203 lost its forming you would get symptoms you observe with your monitor, so I think it is worth a shot to swap it.

Thanks for the reminder!

I'm probably going to be a little delayed in swapping it out, as I have to figure out the best way to remove the metal shield first. I came across this post https://microcorelabs.com/2023/12/17/ibm-pcjr … or-display-fix/ the other day from someone who accidentally ripped a trace taking it off, so I want to make sure I don't make more work for myself. What's more interesting is he had an issue with brightness, which he solved by swapping out cap C205. He tested this first by snipping one of the cap's legs, which brought the colors back and confirmed it needed to be changed. I wonder if snipping a leg on C203 would also confirm whether it's the problem.

Muckrake wrote on 2026-02-27, 02:31:

I wonder if snipping a leg on C203 would also confirm whether it's the problem.

Yes, this will work. The job of C203 is to stabilize the contrast control voltage generated by the contrast potentiometer. If you snip it, you might get slightly flickery low-intensity colors, or you might get slight horizontal gradients in brightness, but these issues will be minor compared to a completely missing contrast control voltage (your issue when the monitor is not working), and the reason is known. So, I think there is no reason not to test it the way you suggest unless you are keen about keeping the original C203 intact if it turns out that your issue has a different root cause.

Couldn't you measure the resistance across C203 with a multimeter to determine leakage? If you put the red multimeter lead on the negative side of the capacitor, the transistor is reverse-biased, so the only current path is directly through the capacitor. Depending on what else is upstream, a low reading could result in a false positive, but if it's high, the capacitor is likely good.

If I'm reading this right, VC1, VC2 and VC3 go into the contrast potentiometer, and VC2 goes to the cap. VC1 and VC3 have a 330 and a 1000 ohm parallel resistance to ground, so with the pot at its minimum setting and a good cap, you should get a reading of at least 250 ohms.