I would stat with checking all of the electrolytic caps on the neck board (the board attached to the CRT tube, or "A" board, as Sony usually calls them on their SMs). In particular, the small caps on there are the most likely to have dried up, since they are in a rather hot zone. The A board is also where all of the signals are handled, so I expect the issue to be there. Apart from that, do you have any pictures of the inside of the monitor? I'm curious if Sony used all-Japanese caps here or if they cheaped out like they did on a Multiscan E100 (CPD-E100p) that I have, which has a lot of tiny Jamicon caps everywhere.

Of course, the issue could be elsewhere too. Half a year ago, I did a "quick recap" on a small 15" HP 56 (M/N: P4795) that had the same (or similar?) smearing issues you described, and the recap didn't actually fix it. But I didn't do a full recap on that one due to being pressed for time, so I don't know if perhaps I didn't miss something else.

I agree with the idea that the fault is most likely on the neck board. Brightness is determined by the cathode drive voltage and the high voltage used for acceleration. There are multiple reasons the fault at hand is not caused by the high voltage supply: First, the high voltage does not only influence brightness, but the picture size as well. An issue in the high voltage supply would not just cause smearing, but also generate severe geometry issues. Second, the CRT itselfs acts as a capacitor on the high voltage. High-voltage issues to not appear out of nowhere from one whte pixel to the neighbouring dark pixel. You can't (dis)charge the CRT that quick, so this leaves the cathode drive circuit as primary suspect.

The inability to show sharp horizontal brightness changes hint towards insufficient supply to the cathode drive (voltage amplification) circuit. And the primary components ensuring a stable supply that can handle quick energy demands for sharp contrasts are electrolytic capacitors. An important hint in the opening post is that there is a luminance fault with no color problems - so all electrolytics that are dedicated to a single color channel are unlikely to be the issue. This can reduce the amount of candidates a lot, especially if you have a service manual that helps you identify which caps are per-channel. If you can make out capacItors that are used to buffer the cathode drive supply voltage (usually around 75-100V) or are used in dynamic brightness tuning/stabilization circuits (if present), those should be the first ones to check.

momaka wrote on 2024-07-11, 17:30:

I would stat with checking all of the electrolytic caps on the neck board (the board attached to the CRT tube, or "A" board, as Sony usually calls them on their SMs). In particular, the small caps on there are the most likely to have dried up, since they are in a rather hot zone. The A board is also where all of the signals are handled, so I expect the issue to be there. Apart from that, do you have any pictures of the inside of the monitor? I'm curious if Sony used all-Japanese caps here or if they cheaped out like they did on a Multiscan E100 (CPD-E100p) that I have, which has a lot of tiny Jamicon caps everywhere.

Of course, the issue could be elsewhere too. Half a year ago, I did a "quick recap" on a small 15" HP 56 (M/N: P4795) that had the same (or similar?) smearing issues you described, and the recap didn't actually fix it. But I didn't do a full recap on that one due to being pressed for time, so I don't know if perhaps I didn't miss something else.

I'm not a cap expert, but I recognize brands and label patterns from Mitsumi/Panasonic, Chemi-Con and Nichicon on the A board. The big filter cap in the 120V section is a Rubycon as well. I would be able to test (and might already have the right caps in my parts bin) but the RF shield is soldered on in about eight places, one of which is right underneath the CRT socket.

mkarcher wrote on 2024-07-11, 18:21:

I agree with the idea that the fault is most likely on the neck board. Brightness is determined by the cathode drive voltage and the high voltage used for acceleration. There are multiple reasons the fault at hand is not caused by the high voltage supply: First, the high voltage does not only influence brightness, but the picture size as well. An issue in the high voltage supply would not just cause smearing, but also generate severe geometry issues. Second, the CRT itselfs acts as a capacitor on the high voltage. High-voltage issues to not appear out of nowhere from one whte pixel to the neighbouring dark pixel. You can't (dis)charge the CRT that quick, so this leaves the cathode drive circuit as primary suspect.

The inability to show sharp horizontal brightness changes hint towards insufficient supply to the cathode drive (voltage amplification) circuit. And the primary components ensuring a stable supply that can handle quick energy demands for sharp contrasts are electrolytic capacitors. An important hint in the opening post is that there is a luminance fault with no color problems - so all electrolytics that are dedicated to a single color channel are unlikely to be the issue. This can reduce the amount of candidates a lot, especially if you have a service manual that helps you identify which caps are per-channel. If you can make out capacItors that are used to buffer the cathode drive supply voltage (usually around 75-100V) or are used in dynamic brightness tuning/stabilization circuits (if present), those should be the first ones to check.

Yeah, picture is good (or as good as can be expected for a high-quality tube with a lot of hours behind it) except for this smearing issue. Looking at the service manual there's a +80V going to the RGB amp and +150V going to the cutoff amp, but in both cases there doesn't appear to be any caps in line, only bypass caps to ground. Is this what you're talking about?

terrorinstinct wrote on 2024-07-11, 21:04:

Looking at the service manual there's a +80V going to the RGB amp and +150V going to the cutoff amp, but in both cases there doesn't appear to be any caps in line, only bypass caps to ground. Is this what you're talking about?

The attachment Screenshot_2024-07-11_15-53-18.png is no longer available

Exactly, I am talking about electrolytic bypass caps. The +150 part for the cutoff is mostly cropped on your attachment. For the +80V line, it is C014. If I understand the circuit correctly, the "cutoff amp" is responsible for setting the base brightness level, and the RGB amp only produces the AC part of the video signal. So any problem close to the cutoff amp is possible to cause an issue as you see it.

mkarcher wrote on 2024-07-11, 21:37:

Exactly, I am talking about electrolytic bypass caps. The +150 part for the cutoff is mostly cropped on your attachment. For the +80V line, it is C014. If I understand the circuit correctly, the "cutoff amp" is responsible for setting the base brightness level, and the RGB amp only produces the AC part of the video signal. So any problem close to the cutoff amp is possible to cause an issue as you see it.

Whoops, here's a better picture of the full path:

I went ahead and ordered the electrolytic caps on the 150V and 80V lines, plus the filter caps used on the RGB preamp IC's 12V rails for good measure:

Looks like Sony had an excess of those "orange/brown chiclet" ceramic caps when they were designing this one, almost every line has one, either inline or as a bypass to ground. Also, what's with those "PROT" diodes that seem to feed into the 80V rail? Why does the 150V rail seem to be tied to the RGB cathode lines in parallel to the cutoff amp's outputs? I'm terrible at reading schematics and would like to get better at it so I can stop asking silly questions like these.

That monitor circuit is quite convoluted, because it seems to separate DC bias and AC signal on the cathode. I'm not completely sure I understand everything about the circuit, but I think I got the basics. A simplified version of the same idea is shown in the application suggestion at the end of the LM2405 datasheet. The datasheet claims a usable output amplitude of that chip of 50Vpp (the dfference between the highest and the lowest voltage produced) and a gain of -14. The negative sign means that the output voltage goes down when the input voltage goes up. To cause a change of the output voltage by 50V, the input needs to change around 3.6V. The datasheet suggests that the input voltage should be centered around 2.6V (i.e. it changes between 0.8V and 4.4V) and the output voltage will then be centered around 50V, i.e. be in a range of 25..75V. Outside of that range, distortion will occur. On the other hand, the Sony schematic and the application suggestion in the data sheet suggest that the CRT needs to operate at a higher voltage level, with the high bound being around 100-130V, so some circuit is needed to shift the output voltage upwards. A higher voltage at the tube (at the cathode) means less bright pixels, so the highest voltage at the tube is considered "black". You need the beam to be "off" during the retrace - this is called "cut-off", and should be "blacker than black", i.e. a voltage level that makes sure the beam is still invisible during retrace even if the brightness is cranked up quite high. The ideal cut-off voltage depends on CRT manufacturing tolerances and are tuned during factory adjustment. The cut-off circuit in a CRT driver controls the DC level in a way that the video output signal is presented to the CRT in a way a controlled driving voltage is achieved. Assume that the "perfect cut-off voltage" is 120V: you would want to have the cathode voltage between 70V and 120V in that case, so the output of the LM2405 needs to be shifted up by 45V. This is done by capacitors in the signal path. C25-C27 in the datasheet circuit; Cx06 (x=1,2,3) in the Sony monitor.

The voltage across these capacitors needs to be controlled in some way. As the beam consists of electrons leaving the cathode, it means negative charge goes away, which will provide a current that will raise the voltage level at the cathode. This current needs to be dissipated somewhere (usually GND) to keep the cut-off voltage stable. That's where the diodes Dx06 kick in: Everytime the video goes black, the voltage at those diodes will be maximal. Extra charge from the beam, that might have increased the voltage across Cx06 will then cause the voltage at the anode of Dx06 to be that high that it starts conducting towards the cutoff amplifier until the desired DC level has been re-established. The diodes makes sure that the cut-off circuit only influences the video signal while the video signal is (temporarily) black - so the cut-off circuit appears "as if it is not present" during the active time of the picture, and does not provide extra load to the RGB amplifier. During retrace, any excessive charge that has accumulated during the scan line will then be removed, so the voltage across the capacitors Cx06 stays (approximately) constant. As the cut-off circuit is only able to pull down, and you don't want to rely solely on the beam current to provide the counteracting pull-up action, Rx09 continously provide extra positive charge to keep the circuit at its operating point. Diodes Dx04 and Dx05 are indeed just for protection: On a CRT, you might get undesired voltage excursions, e.g. during power-up and power-down when the different parts of the tube get charged to their desired target voltage (or discharged on power-down), as well if arcing occurs in the tube. The diodes are meant to funnel excessive voltages into GND or +80V without that voltage passing through the LM2405 chip (which could damage it).

terrorinstinct wrote on 2024-07-11, 21:04:

I would be able to test (and might already have the right caps in my parts bin) but the RF shield is soldered on in about eight places, one of which is right underneath the CRT socket.

Yeah, I know that may sometimes look like a difficult job, but it's actually not that bad at all.
Usually what I do is I take a small flat heat screw driver and start putting a bit of pressure near each soldered tap and then heat it up with the soldering iron. On the first few, I just get maybe a mm or 2 max of separation. After that, the others allow the shield to be lifted more when I heat those taps. Eventually, I circle back to the first few tabs I only desoldered a few mm from and lift them further. I do this several times and the shield comes off after that. The soldered tab on the back of the tube socket is the last one I do / desolder.
You could also use braid or just heat up the solder from each tap and try to drag it away... but I personally find the "lift each tab a little at a time" a little easier.

terrorinstinct wrote on 2024-07-12, 19:52:

Also, what's with those "PROT" diodes that seem to feed into the 80V rail?

Those are there so that in case the output pins on the RGB amp (Vout1, Vout2, and Vout3) go higher than 80V due to a fault (i.e. C106, C206, and C306 go short-circuit), then the diodes will clamp the voltage down to no more than 80V + ~1V max for the diode voltage drop of the protection diodes (i.e. about 81V total.) So in essence, those "prot" diodes are for over-voltage protection.

terrorinstinct wrote on 2024-07-12, 19:52:

Why does the 150V rail seem to be tied to the RGB cathode lines in parallel to the cutoff amp's outputs?

Well, the 150V line is not directly tied to the RGB cathode lines. Rather, it pulls up the RGB cathode lines up to 150V DC through the 1 MegaOhm resistors R109, R209, and R309.
Of course, the Cut-Off amp may decrease that DC value to something lower by slightly sinking current on those lines in order to control the base brightness, as mkarcher mentioned.
Finally, you have the RBG video signal coming in from the RBG amp, which is an AC signal and is coupled to the DC voltage above through capacitors C106, C206, and C306). The resultant signal/voltage going to the RGB cathodes is a positive DC voltage bias along with an AC voltage signal on top.

terrorinstinct wrote on 2024-07-12, 19:52:

I'm terrible at reading schematics and would like to get better at it so I can stop asking silly questions like these.

Hey, no worries.
Asking "silly" questions is the right way to learn.
Besides, I'm sure you know the saying: there no such thing as a silly question, only a silly answer. 😉

I replaced the caps yesterday. Smearing is still there, but from my memory (I hadn't used this monitor in a while) it's less severe than it was before the recap. It's a very subtle rise or dip in brightness (rise after a long run of bright pixels, dip after dark pixels) that slowly recovers over the rest of the scanline. You have to squint to see it on a static image but it's obvious when you start moving things on the Windows desktop. Given the cable has gained a few small kinks over the years, I'm guessing the root cause is signal reflections at this point and I'll just have to live with it. It wasn't a total waste of time though, the picture is a lot brighter now than it was before the recap. When I first pulled this thing out of storage I assumed the dim picture was due to tube hours, so having to turn brightness down from the 80s to the 50s to keep my eyes from burning was a pleasant surprise.

Thanks to everyone for the help. Someday somebody is going to need to gather all this wisdom into one place and help save more of these beasts from the landfill. Maybe that somebody will be me.