VOGONS


Quandchrome EGA screen failure.

Topic actions

First post, by SVD

User metadata
Rank Member
Rank
Member

I have gotten a Quadchrome EGA screen. I plan on using it with my IBM 5162 with a CHIPS based EGA card and that is what i have tested with.

Ufortunatly the screen wont work. Initially I got it to display some scrambled red text when CGA was selected on the video card. But now it goes completely dark\dead when hooking up the monitor (except the green led) in both EGA and CGA. When not connected the monitor has this vertical band as in the photo.
When opening it i smelled something burt in the power supply unit. Can a bad powersupply capacitor make the screen behave like this? I am a bit reluctant to open up the power supply. But at least it is not riveted in this unit but screwed. Unlike the IBM 5154.

The brightness and contrast seem to work when adjusting them when the monitor is disconnected.

Attachments

Reply 1 of 22, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Do not run in this condition! Horizontal circuit if vital parts is very hard to replace these day if it finally blows under stress in that malfunctioning condition.

You'll need to find someone who is expert in monitor and TV repair when CRT was in it's prime.

Cheers,

Great Northern aka Canada.

Reply 2 of 22, by Caluser2000

User metadata
Rank l33t
Rank
l33t

Sark over in SLASHNet #retrodreams irc channel should be able to help you out. He in a CRT Guru located in the States.

There is bound to be someone at the vcfed.org forums who can help out as well.

There's a glitch in the matrix.
A founding member of the 286 appreciation society.
Apparently 32-bit is dead and nobody likes P4s.
Of course, as always, I'm open to correction...😉

Reply 3 of 22, by SVD

User metadata
Rank Member
Rank
Member

Thanks for the replies guys. I have not had the screen on for much time of course. but I did need to see if it worked and if i could get it to work in CGA and so on.
I have tried two retro\TV reps locally but none answers... You do not believe in going over the power supply?

Reply 4 of 22, by retardware

User metadata
Rank Oldbie
Rank
Oldbie
SVD wrote on 2021-10-27, 11:12:

You do not believe in going over the power supply?

Definitely worth to check the PSU voltages (voltage, maybe ripple too).
These have usually a number of voltages and I'd consider the possibility that one of these failed not unlikely.

If you don't have the schematic, better make a diagram/table yourself and mark the test points on the boards, so you can do these tests quickly and don't run the monitor longer than necessary in this state.
Usually there are prints on the PCB solder and components sides that help a bit.

Reply 5 of 22, by Benedikt

User metadata
Rank Oldbie
Rank
Oldbie

If you don't want to open the PSU, you could still try to take pictures of the PSU board with a mobile phone camera through the ventilation holes.
Maybe you can spot something that looks obviously broken. I used that technique on an ATX PSU once.

Reply 6 of 22, by SVD

User metadata
Rank Member
Rank
Member

I think I will open the PSU then, hopefully the fault is obvious, like a blown capacitor or something. And if the board marks what voltage the different points should have I can check them, but otherwise I have no idea. I have had no luck finding any schematics on this. Its hard to find any info at all. x)
I kinda feel like I just should replace all the electrolytic capacitors in this thing. but then I probably have to adjust the 100 something potmeters in it.

Reply 7 of 22, by Caluser2000

User metadata
Rank l33t
Rank
l33t

Be careful it may full of dead cockroaches 😀

There's a glitch in the matrix.
A founding member of the 286 appreciation society.
Apparently 32-bit is dead and nobody likes P4s.
Of course, as always, I'm open to correction...😉

Reply 8 of 22, by kdr

User metadata
Rank Member
Rank
Member
SVD wrote on 2021-10-26, 21:50:

Ufortunatly the screen wont work. Initially I got it to display some scrambled red text when CGA was selected on the video card. But now it goes completely dark\dead when hooking up the monitor (except the green led) in both EGA and CGA. When not connected the monitor has this vertical band as in the photo.

Disclaimer: I still don't have a lot of experience fixing CRTs yet

It's fairly strange to see the monitor showing a bright blue image when nothing is connected to the inputs. That makes me think something is wrong with the video input circuitry. The narrow width of the image indicates that there isn't enough voltage in the horizontal yoke (or the hsync frequency is too high).

Since you mentioned a possible fault in the power supply, I'm thinking that it could be all of the voltages being output from the PSU are too low. There's likely to be a +12V output that powers the video input section as well as the B+ output at anywhere from 70V ~ 120V (varies depending on the monitor design) for the horizontal output and flyback section. If you're lucky the PCB will have some silkscreening that identifies the output voltages. Anyway it's not impossible that recapping the PSU section could fix the monitor, so it's worth a try!

Reply 9 of 22, by SVD

User metadata
Rank Member
Rank
Member

I finally opened up the power supply. It seems there is a load resistor on this monitor as there was no charge on the tube at least.
The power supply looks as if it went of the production line not long ago to me. No caps are bulging, except the two big ones, but i guess they are supposed to be like that?

Attachments

Reply 10 of 22, by SVD

User metadata
Rank Member
Rank
Member

On the video input circuitry there seem to be a SN74522 chip that has been changed. As the board looks to be heated on the back, and possible flux residue. Or is it fried?

Edit: I disconnected the input circuitry and turned on the monitor, its just as black and dead as when the EGA card is connected. I measured the voltages coming in to the circuit. its 6.8V and 16.8V.
I measured again and now its 7.7v and 18.8v... The red wire is the ground, the blue 7.7v and black 18.8v. Is it normal to have the colours switched like that?

Edit 2:

And then its confirmed, the SN74522 has been recieving close to 8v and its max spec is as far as i found 5.25v Its probably been fried before and changed, maybe its fried again or some other chip is.

Attachments

Reply 11 of 22, by mkarcher

User metadata
Rank l33t
Rank
l33t
kdr wrote on 2021-10-27, 19:56:

Disclaimer: I still don't have a lot of experience fixing CRTs yet

It's fairly strange to see the monitor showing a bright blue image when nothing is connected to the inputs.

It's quite common for monitors of that era. A usual quick test for monitors in the 80s was to disconnect the input and turn up the brightness to 80%. You expected a light gray screen with small black borders left and right. The black borders in the OP are too wide, so the picture is suspect.

The reason for that effect is that TTL inputs appear "high" if they are left open, so a monitor not connected to a graphics card "sees" the color "white" on the input, even if it is working fine. This changed with VGA when we switched to analog inputs that are terminated using 75Ω resistors to ground, so an unconnected monitor sees "black" as color.

Most CGA and and EGA monitors have a horizontal oscillator that oscillates even if there is no hsync. That's why you get deflection even without any input signal. This is in contrast with the IBM 5151 MDA monitor that directly takes hsync pulses from the graphics card to trigger the horizontal retrace, so it doesn't run at all without an input signal.

Reply 12 of 22, by kdr

User metadata
Rank Member
Rank
Member
mkarcher wrote on 2021-11-28, 18:15:

It's quite common for monitors of that era. A usual quick test for monitors in the 80s was to disconnect the input and turn up the brightness to 80%. You expected a light gray screen with small black borders left and right. The black borders in the OP are too wide, so the picture is suspect.

The reason for that effect is that TTL inputs appear "high" if they are left open, so a monitor not connected to a graphics card "sees" the color "white" on the input, even if it is working fine. This changed with VGA when we switched to analog inputs that are terminated using 75Ω resistors to ground, so an unconnected monitor sees "black" as color.

Ah, thanks for the correction!

Reply 13 of 22, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Still, need to check the horizontal deflection circuits and their driver circuit. Usually the capacitors are the issue. Also make sure width adjustment circuit is good.

I use ESR meter and multimeter to check them.

Cheers,

Great Northern aka Canada.

Reply 15 of 22, by SVD

User metadata
Rank Member
Rank
Member

I looked for the horizontal deflection circuit, I think i found it, but i dont know much about these things. The circuit looks good. But the coil has some corrosion on its grounding, i dont know if its so much that it matters somehow tough. Since none of the TV-repair shops around answers my call, I really hope someone can help me with this x) Thanks for the tips on until now BTW, its much appreciated 😀 I think I need to get an ESR meter...

Attachments

Reply 16 of 22, by kdr

User metadata
Rank Member
Rank
Member
SVD wrote on 2021-11-29, 19:19:

Since none of the TV-repair shops around answers my call, I really hope someone can help me with this

You can also try posting at the VCF forums (vcfed.org) - there are a couple of members over there with a lot of CRT troubleshooting experience. (I'm thinking of Hugo in particular.)

Reply 17 of 22, by mkarcher

User metadata
Rank l33t
Rank
l33t
SVD wrote on 2021-11-29, 13:46:

You do not think the 8V have burnt the chips on the board?
BTW what makes the power supply give that high of a voltage to begin with?

If a 74-series logic chip is powered from a certain output on the power supply, that output is supposed to be 5V. If that output rises to 8V, that chip is likely toast. This doesn't just apply to the 74522 you were writing about, but to all logic chips designed for 5V supply. Depending on the design of your monitor, you might have only one 5V chip in it, because a lot of CRT control chips are made with classic analog TV design in mind and often run at 10-15V with a nominal 12V supply.

TV and monitor power supplies sometimes output strange voltages. These devices don't need to interoperate with anything, so they may design their stuff to whatever voltage they like. A +8V output isn't unheard of, and is a perfectly valid design choice as the input voltage for a 7805-type regulator. It is of course completely invalid to just connect a +8V output to the supply of a 5V logic chip. So maybe it's not your power supply that's broken, but a voltage regulator that should provide stable 5V from the coarse "around 8V" has failed and passes the full input voltage. We can't know that without the schematics of your monitor.

If the supply is supposed to output 5V, but does in fact output 8V, it is either broken, or the monitor is mistreating it in some way. As you found out, your supply outputs two voltages, around 8V and 18V. If it is a switch-mode supply, the output voltages are in sync. Just suppose for the moment that your supply is meant to provide +24V and +5V, and the +24V output is overloaded because something in the monitor is shorted. This would cause the power supply to pump a lot of energy into the monitor trying to get the +24V output up to the desired voltage. This excessive energy pumped into the supply will for some part end up on the +5V output which is not overloaded and bring this output too high. The strategy for regulating such a multi-voltage supply can be implemented in different ways: You could try to just regulate one of the outputs, and the other will be "good enough" unless the supply is overloaded, or you might regulate the average output voltage. There are a lot of possibilities, making a lot of theories plausible, but the only thing that can help diagnose your supply is knowing how it is supposed to work, so again a schematic would be very helpful.

Reply 18 of 22, by SVD

User metadata
Rank Member
Rank
Member
mkarcher wrote on 2021-11-29, 21:07:
If a 74-series logic chip is powered from a certain output on the power supply, that output is supposed to be 5V. If that output […]
Show full quote
SVD wrote on 2021-11-29, 13:46:

You do not think the 8V have burnt the chips on the board?
BTW what makes the power supply give that high of a voltage to begin with?

If a 74-series logic chip is powered from a certain output on the power supply, that output is supposed to be 5V. If that output rises to 8V, that chip is likely toast. This doesn't just apply to the 74522 you were writing about, but to all logic chips designed for 5V supply. Depending on the design of your monitor, you might have only one 5V chip in it, because a lot of CRT control chips are made with classic analog TV design in mind and often run at 10-15V with a nominal 12V supply.

TV and monitor power supplies sometimes output strange voltages. These devices don't need to interoperate with anything, so they may design their stuff to whatever voltage they like. A +8V output isn't unheard of, and is a perfectly valid design choice as the input voltage for a 7805-type regulator. It is of course completely invalid to just connect a +8V output to the supply of a 5V logic chip. So maybe it's not your power supply that's broken, but a voltage regulator that should provide stable 5V from the coarse "around 8V" has failed and passes the full input voltage. We can't know that without the schematics of your monitor.

If the supply is supposed to output 5V, but does in fact output 8V, it is either broken, or the monitor is mistreating it in some way. As you found out, your supply outputs two voltages, around 8V and 18V. If it is a switch-mode supply, the output voltages are in sync. Just suppose for the moment that your supply is meant to provide +24V and +5V, and the +24V output is overloaded because something in the monitor is shorted. This would cause the power supply to pump a lot of energy into the monitor trying to get the +24V output up to the desired voltage. This excessive energy pumped into the supply will for some part end up on the +5V output which is not overloaded and bring this output too high. The strategy for regulating such a multi-voltage supply can be implemented in different ways: You could try to just regulate one of the outputs, and the other will be "good enough" unless the supply is overloaded, or you might regulate the average output voltage. There are a lot of possibilities, making a lot of theories plausible, but the only thing that can help diagnose your supply is knowing how it is supposed to work, so again a schematic would be very helpful.

Yeah that's what I thought about the 8V being wrong, thanks for confirming it. There is 12 chips on the video input board alone 😒 There is also some chips on the mainboard but that is a lesser amount. I think I should check the voltage again, this time over the chip and not to ground. If it still is 8V-ish then I need to order chips.
Is it problematic to change the chips to a modern equivalent? 74s22 => NTE7413 74s05=>SN74HC14N Or are they equivalent?
Thanks for the tip about the power supply, is it also possible that the 18v should be 12? I dont know if a dead capacitor or voltage regulator could throw the whole supply off like that but i guess its possible. I should probably check all the diodes, transistors and caps also.

Reply 19 of 22, by mkarcher

User metadata
Rank l33t
Rank
l33t
SVD wrote on 2021-11-30, 11:30:

Yeah that's what I thought about the 8V being wrong, thanks for confirming it. There is 12 chips on the video input board alone 😒 There is also some chips on the mainboard but that is a lesser amount. I think I should check the voltage again, this time over the chip and not to ground. If it still is 8V-ish then I need to order chips.
Is it problematic to change the chips to a modern equivalent? 74s22 => NTE7413 74s05=>SN74HC14N Or are they equivalent?
Thanks for the tip about the power supply, is it also possible that the 18v should be 12? I dont know if a dead capacitor or voltage regulator could throw the whole supply off like that but i guess its possible. I should probably check all the diodes, transistors and caps also.

It usually is OK to substitute 74-series chips, but you need to know what you are doing. There are two different conventions about how you define high and low. The "old" convention is called TTL convention. You need to pull down an input to below 0.8V to make it low, and you might need to pull quite strong. On the other hand, inputs rise back to high on their own, but if you need it quicker, you can push a little current into it. As soon as the input is higher than 2.0V, it is guaranteed to be read as high. The "new" convention is called "CMOS" convention. This one is symmetrical. Everything below 2V is guaranteed low, everything above 3V is guaranteed high. You don't need to pull strongly to keep an input low. You should avoid mixing TTL and CMOS logic, although connecting CMOS outputs to TTL inputs will work. CMOS outputs can pull strong enough either way. TTL outputs might not reliably go high enough to make CMOS inputs high.

So you need to be careful when substituting chips: 74xxx, 74Sxxx, 74LSxxx use TTL levels for input and output. 74HCxxx use CMOS levels for input and output. If you want to replace a 74LSxxx chip and can't get a suitable LS replacement, you can also use a 74HCTxxx chip. The "T" in the type number tells you that the inputs are TTL compatible, although the chip itself is built using HC (high-speed CMOS) technology. A 74S22 ("dual 4-input NAND gate with open collector output") can not be replaced by a NTE7413 ("dual 4-input NAND fate with schmitt trigger inputs"), because you may not replace open-collector output gates by standard (totem-pole) output gates. You also should avoid replacing 74S series chips ("schottky") by 74LS series chips ("low-power schottky"), because the low-power chips are slower and have less strong outputs. I failed to quickly find a substitute for the 74S22 that will work in any circuit.

The 74S05 again is a chip with open collector outputs, so you need to replace it with a chip that also has open collector (or open drain for CMOS chips) output. The 74HC14 is thus wrong for two reasons: It doesn't have open-collector outputs, and it's a HC chip that may be incompatible with TTL inputs. You may substitute a 74S05 by a 74S06 or a 74S16, though. Again, I fail to find easily available modern equivalents.

The 18V output might be supposed to have 12V, but it might also be supposed to 16V, 18V or 19.5V. Without specifications for that supply, we don't know. In my experience, there is no "set of standard voltages" in classic CRT monitors. Power supplies like the one in your monitor typically use an opto-isolator that is used to communicate "we have enough voltage on the output, push less energy, please" to the primary side. Overvoltages might be caused by the optoisolator being weak or defective. That's something you don't find easily when you use standard component test procedures.