Presumably that's where if first sorts out the H/SYNC for the different frequencies?
So if something went wrong there it might have the runaway issue you mentioned?
Okay so the H Signal seems to be fed into a QUAD 2 input SN74LS136 XOR gate IC chip , and flanked by 12 2SC945 NPN transistors. Though it could be anything else around there too.
More intrestingly the 136 IC is connected to a SN74LS123 which is a retriggerable monostable multivibrator , which sounds like it might have something to do with frequencies.
So looking up a video about the LS123 it mentioned it being used for time constants , the LA7850 IC that handles the Horizontal oscillation and the phase shifter has time constant pins connected to the horizontal adjustment and horizontal hold potentiometers.
In the video he mentioned the ICs being unreliable and drifting over time in value so I'm not necessarily saying the ic chip might have gone bad over nearly 40 years but its possible.
Also it seems the ic values are programmed via resistors and capacitors so those also may have gone bad and might need to be replaced.
There's also what appears to be a dead spider near the IC , probably not the culprit but would be kinda amusing if it was.
AppleSauce wrote on 2025-04-29, 06:54:I'm not sure if the sharp X68000 is at fault , (though it might be) because i remember also connecting the screen to a test socket 370 motherboard and having the same issue. So if two different systems have the same issue I'd assume the monitor is at fault.
AppleSauce wrote on 2025-04-29, 11:46:I just tested connecting the multisync to my DOS PC , had the collapse issue again.
OK, so this leaves the question why we see an increasing horizontal ouput signal frequency in Re: Nec Multisync horizontal collapse issue , but there is no loss of synchronization in Nec Multisync horizontal collapse issue.
Assuming the PC outputs video at a constant horizontal frequency of 31.5kHz, but the deflection output of the IC401 is increasing in frequency (up to 40kHz, and then the oscillator shuts down), there wouldn't be a stable image on the screen. This might have different causes:
As long as the picture doesn't lose horizontal synchronization, the horizontal oscillator works fine. You might have issues regarding the F/V converter and the subsequent comparators, op-amps and CMOS switches, though. This circuit is meant to configure the horizontal oscillator (both the oscillator in IC401 as well as parts of the kind-of resonant reflection circuit) depending on the input horizontal frequency. This means it would be helpful to validate the voltages in that section against the values given in the schematic. As the schematic has values for 30.5kHz (IBM PGA) instead of 31.5kHz (IBM VGA), values may slightly deviate. The output voltages of the comparators (entries 1-4 in the table on the top right corner near the "DEFlection PCB") is a first good indicator whether the "input frequency classification" works as intended. So we expect 16V ("logic" high) at pin 13, and 0V ("logic" low) at pins 1, 2 and 14 of IC553. As I understand it, the whole circuit around the F/V converter and IC401 operates on fixed voltages generated by the 85V/24V supply, and not by the variable B+ supply, so everything up to the input of IC401 should not be changing in response to the collapse, but if there is something wrong, that likely is the cause of the collapse. But please verify against the timing: If some voltages measure strange values only after the collapse, that's a sign they are not causing it.
mkarcher wrote on 2025-04-29, 18:35:OK, so this leaves the question why we see an increasing horizontal ouput signal frequency in Re: Nec Multisync horizontal colla […]
AppleSauce wrote on 2025-04-29, 06:54:I'm not sure if the sharp X68000 is at fault , (though it might be) because i remember also connecting the screen to a test socket 370 motherboard and having the same issue. So if two different systems have the same issue I'd assume the monitor is at fault.
AppleSauce wrote on 2025-04-29, 11:46:I just tested connecting the multisync to my DOS PC , had the collapse issue again.
OK, so this leaves the question why we see an increasing horizontal ouput signal frequency in Re: Nec Multisync horizontal collapse issue , but there is no loss of synchronization in Nec Multisync horizontal collapse issue.
Assuming the PC outputs video at a constant horizontal frequency of 31.5kHz, but the deflection output of the IC401 is increasing in frequency (up to 40kHz, and then the oscillator shuts down), there wouldn't be a stable image on the screen. This might have different causes:
- Maybe something changed between the original video that showed stable video when horizontal deflection collapsed, and the "this happens during shutdown" image. Can you confirm whether the monitor loses sync before collapsing now?
- Maybe may theory that the increasing oscillator frequency causes the shutdown is wrong, and in fact the oscillator runs up in response to B+ already having turned off. Can you see whether the increase of the horizontal oscillator frequency happens before or after the collapsing?
As long as the picture doesn't lose horizontal synchronization, the horizontal oscillator works fine. You might have issues regarding the F/V converter and the subsequent comparators, op-amps and CMOS switches, though. This circuit is meant to configure the horizontal oscillator (both the oscillator in IC401 as well as parts of the kind-of resonant reflection circuit) depending on the input horizontal frequency. This means it would be helpful to validate the voltages in that section against the values given in the schematic. As the schematic has values for 30.5kHz (IBM PGA) instead of 31.5kHz (IBM VGA), values may slightly deviate. The output voltages of the comparators (entries 1-4 in the table on the top right corner near the "DEFlection PCB") is a first good indicator whether the "input frequency classification" works as intended. So we expect 16V ("logic" high) at pin 13, and 0V ("logic" low) at pins 1, 2 and 14 of IC553. As I understand it, the whole circuit around the F/V converter and IC401 operates on fixed voltages generated by the 85V/24V supply, and not by the variable B+ supply, so everything up to the input of IC401 should not be changing in response to the collapse, but if there is something wrong, that likely is the cause of the collapse. But please verify against the timing: If some voltages measure strange values only after the collapse, that's a sign they are not causing it.
The frequency does increase as you turn the knob counter clockwise and move the image away from the far right to the left as the image distorts you can turn the knob back clockwise to pull it out of an unstable image but if you adjust the image to far to the left it will do the collapse.
Apart from 31 khz the collapse issue occurs at 25 khz when i run a game like Ys.
Only at 15 it wont , also i think the monitors idle frequency is like 14.5 or something when you don't connect the display to anything.
Anywho ill give the stuff you suggested a measure.
Also i here's some footage of the horizontal adjustment messing with the Horizontal frequency.
AppleSauce wrote on 2025-04-29, 18:45:The frequency does increase as you turn the knob counter clockwise and move the image away from the far right to the left as the image distorts you can turn the knob back clockwise to pull it out of an unstable image but if you adjust the image to far to the left it will do the collapse.
Oh, I totally missed the most important part of the problem description: The image doesn't collapse "just because", but only if you adjust the horizontal position. And while the frequency runs up, the image loses synchronization, though it is able to capture back the sync pulses when you move the position control more to the right.
So, revisiting the issues, it seems IC401 loses horizontal sync and runs up to high frequencies if you move the picture to the left. At 31kHz, both VR553 "SUB H CENTER" (no number) and VR554 "SUB H CENTER 3" should be inoperative, so what's left on the line is VR501 "SUB H CENTER 1", a 300k pot, and VR4 (the user control "H position"), a 10k pot. I fail to understand why the user-controlled put is such a low value compared to the service pots, and without any series resistor that prevents excessively low resistance values of VR4. On the other hand, the LA7850 datasheet also shows just 20K adjustable for the R/C time constant for the phase shifter as well, so 10K for the user control isn't the far off.
I recommend you to scope what you observe at pin 2 of IC401 when you shift the picture to the left, and compare the behaviour at that pin with the behaviour at 15kHz.
mkarcher wrote on 2025-04-29, 19:23:Oh, I totally missed the most important part of the problem description: The image doesn't collapse "just because", but only if […]
AppleSauce wrote on 2025-04-29, 18:45:The frequency does increase as you turn the knob counter clockwise and move the image away from the far right to the left as the image distorts you can turn the knob back clockwise to pull it out of an unstable image but if you adjust the image to far to the left it will do the collapse.
Oh, I totally missed the most important part of the problem description: The image doesn't collapse "just because", but only if you adjust the horizontal position. And while the frequency runs up, the image loses synchronization, though it is able to capture back the sync pulses when you move the position control more to the right.
So, revisiting the issues, it seems IC401 loses horizontal sync and runs up to high frequencies if you move the picture to the left. At 31kHz, both VR553 "SUB H CENTER" (no number) and VR554 "SUB H CENTER 3" should be inoperative, so what's left on the line is VR501 "SUB H CENTER 1", a 300k pot, and VR4 (the user control "H position"), a 10k pot. I fail to understand why the user-controlled put is such a low value compared to the service pots, and without any series resistor that prevents excessively low resistance values of VR4. On the other hand, the LA7850 datasheet also shows just 20K adjustable for the R/C time constant for the phase shifter as well, so 10K for the user control isn't the far off.
I recommend you to scope what you observe at pin 2 of IC401 when you shift the picture to the left, and compare the behaviour at that pin with the behaviour at 15kHz.
Yeah sorry I'm doing a horrible job explaining stuff , But wait just to confirm , so basically you don't think the SN74LS123 is a potential culprit then?
Also I'm pretty sure i measured pin 2 (thats the phase adjust time constant?), ill need to find the video i made ,
AppleSauce wrote on 2025-04-29, 19:26:Yeah sorry I'm doing a horrible job explaining stuff , But wait just to confirm , so basically you don't think the SN74LS123 is a potential culprit then?
Also I'm pretty sure i measured pin 2 (thats the phase adjust time constant?), ill need to find the video i made ,
You mean IC805? I excluded that chip, because the operation of that chip doesn't depend on the horizontal position control.
mkarcher wrote on 2025-04-29, 19:48:AppleSauce wrote on 2025-04-29, 19:26:Yeah sorry I'm doing a horrible job explaining stuff , But wait just to confirm , so basically you don't think the SN74LS123 is a potential culprit then?
Also I'm pretty sure i measured pin 2 (thats the phase adjust time constant?), ill need to find the video i made ,
You mean IC805? I excluded that chip, because the operation of that chip doesn't depend on the horizontal position control.
After some digging I found that a different monitor , a commodore 1901 also uses the LS123 IC.
It seems to be used in the horizontal shift circuit. Is it possible the NEC multisync also uses the same ic in a similar way to shape pulses and variate the position of them?
as does this phillips monitor , it might be nothing over all just wanted to point it out in case it is.
Okay so bit of an update , I ended up desoldering and socketing the ls123 chip with a replacement , I also replaced the capacitor and resistor that was used to set the values for the ic. But it didn't do anything , I put the ic in a rom flasher which can also test logic ICs and it came out fine.
After that I pulled the crt apart again , and decided to remove all the chips on it and to socket them. I got ic555 the 4066 switch since it was close to the area the issue was occurring and what do you know the chip tested bad. For a sanity check I tested ic 451 which is exactly the same ic , and it tested good. So this might be the smoking gun. I dont know why I didn't bother testing the ics before would have saved me a bunch of hassle and stress on the crt.
Okay another update , its apparently FIXED. So the chips arrived , I replaced them and all 3 sync modes work now and let me adjust the horizontal without collapse , ill run it a few more times just to make sure.
For some reason now the CRT makes a high pitch whine now though , hopefully the reflow didn't harm anything , or maybe the flyback is on the way out , luckily i have a spare one.
I'll probably just dial in the 3 different horizontal pos potentiometers for each sync as they might be a bit off , and hopefully that should be it for now. After that i can put the case back on.
