PDA

View Full Version : 5153 Repair - Color #8 Not Showing Up



Great Hierophant
February 2nd, 2009, 08:03 PM
On my used 5153 IBM PC Color Display (the CGA monitor), color #8 (gray/dark gray) is indistinguishable from color #0 (black). The other fifteen colors display well. I can slightly distinguish gray from black if I have the brightness knob turned all the way down and the contrast knob turned all the way up, but this is hardly a practical solution.

I would like to know if anyone else has encountered this problem and if there is a solution to it. I am willing to enter the 5153 and brave electric shock to fix it.

Chuck(G)
February 2nd, 2009, 08:55 PM
You should turn the contrast down and the brightness up, not the other way around. If that doesn't work, check to see that S301 (service switch) is in the correct position.

Failing that, I have a schematic for the thing.

Great Hierophant
February 2nd, 2009, 09:10 PM
You should turn the contrast down and the brightness up, not the other way around. If that doesn't work, check to see that S301 (service switch) is in the correct position.

Failing that, I have a schematic for the thing.

I think that is what I meant to say, but I cannot comprehend the schematic as shown in the IBM PC Tech Reference. I will check the switch

Chuck(G)
February 2nd, 2009, 09:25 PM
Do you get a true brown for color no. 6?

I assume that you're aware of this site (http://www.oldskool.org/pc/cgacal/) for a calibration program.

Fallo
February 2nd, 2009, 10:39 PM
Do you get a true brown for color no. 6?

It's a rather obscure fact that CGA cards generate dark yellow, and not brown. You can see this by using a composite monitor. For whatever reason, IBM (probably wanting to mimic the color palettes of their terminals) added circuitry to the 5153 to remove the green component from the yellow, turning it into brown. Most clone RGB monitors then did likewise.

Great Hierophant
February 3rd, 2009, 06:53 AM
Yes, I see true brown for #6.


It's a rather obscure fact that CGA cards generate dark yellow, and not brown. You can see this by using a composite monitor. For whatever reason, IBM (probably wanting to mimic the color palettes of their terminals) added circuitry to the 5153 to remove the green component from the yellow, turning it into brown. Most clone RGB monitors then did likewise.

I was involved in the discussion a few years ago that definitively answered the question of whether #6 was brown or dark yellow. Logically, the digital value output by the CGA card on its digital RGBI output should give a dark yellow, but IBM (and most post-5153 clone) monitors did as you say. The color on a composite monitor is indeed dark yellow.

patscc
February 3rd, 2009, 07:42 AM
Can you post a pic of color bars, or something, preferably with all colors ?
The circuit isn't too complicated, it's just a couple of hex inverters with a voltage divider network in the front of each input.
If the problem truly is the grey/black, that's driven by the Intensity line which goes to pins 11 & 12 of Q253 (the hex inverter), and it stands to reason that this would affect the other colors as well.
That's why a picture would be helpful.
Do you have a multimeter ?
patscc

Chuck(G)
February 3rd, 2009, 09:21 AM
It's a rather obscure fact that CGA cards generate dark yellow, and not brown. You can see this by using a composite monitor. For whatever reason, IBM (probably wanting to mimic the color palettes of their terminals) added circuitry to the 5153 to remove the green component from the yellow, turning it into brown. Most clone RGB monitors then did likewise.

Not obscure at all--in fact, that was pretty evident when people began using other monitors for CGA output. What I call obscure is the circuitry in the IBM EGA monitor to synthesize brown--that many EGA monitors lack.

The reason I asked about brown #6 is that the intensity bit is tied into its decoding (as Pat has noted). If there were something wrong with the early-on intensity logic (e.g. a toasted 74LS05), a lack of brown would be an easy-to-recognize symptom. The next thing I would probably do is grab my scope and see what kind of RGB level is being presented for color #8. It might be that the jug in the 5153 is simply very old and the emission level is low and the RGB levels need to be brought up a bit to get it.

Great Hierophant
February 3rd, 2009, 10:47 AM
Can you post a pic of color bars, or something, preferably with all colors ?
The circuit isn't too complicated, it's just a couple of hex inverters with a voltage divider network in the front of each input.
If the problem truly is the grey/black, that's driven by the Intensity line which goes to pins 11 & 12 of Q253 (the hex inverter), and it stands to reason that this would affect the other colors as well.
That's why a picture would be helpful.
Do you have a multimeter ?
patscc

I don't have a multimeter. The picture here is accurate of what shows on my 5153 with a true IBM CGA card in an IBM PC 5150:
http://www.oldskool.org/pc/cgacal
Except that the dark gry bar is completely black. The text cannot be distinguished from the background. All the other seven intense colors display normally.

This suggests to me that the tube has aged so that it can no longer distinguish between the darker colors.

Where is switch S301?

patscc
February 3rd, 2009, 11:42 AM
The service switch might bring you no joy. In a lot of monitors & tv's all that it does is turn the vertical deflection off, so all you get is a thin horizontal line across the screen, which can be a bit disconcerting.
patscc

patscc
February 3rd, 2009, 11:46 AM
Could we please get one of you actual monitor, NOT one from the website of the program ?
If it seems to dark, just up the shutter time a bit. Even if you end up with a bit of jitter in the frame, for looking at the colors it shouldn't make much difference.
If possible, take the pics with the brightness & contrast set to middle position.
Do you by any chance notice the picture being smaller than it should be ? (See, lots of things to look at, hence the picture)
Thanks.
patscc

barythrin
February 3rd, 2009, 11:56 AM
This may not make much of a difference in your testing but are you just testing blank bars or text? In basic if you did color 8: print "A" do you still not notice anything?

Fallo
February 3rd, 2009, 12:56 PM
Not obscure at all--in fact, that was pretty evident when people began using other monitors for CGA output. What I call obscure is the circuitry in the IBM EGA monitor to synthesize brown--that many EGA monitors lack.

The problem was that EGA monitors can be connected to a CGA card, in which case they will be getting that dark yellow signal. IBM put the brown-generating circuitry in the 5154 for that reason, but many clone monitors overlooked this little detail.

Now, I don't know if most CGA clones (PCjr/Tandy included), generate yellow or brown. Does anyone know?

Trixter
February 3rd, 2009, 02:31 PM
They all *generate* dark yellow but the *monitors* all display dark brown.

Chuck(G)
February 3rd, 2009, 02:44 PM
Pat, take a look at the schematic, bottom of the second page, near R439 and R440 (both 2K). Does that arrangement across points "Y" and "Z" look like a master brightness adjustment (i.e. remove/add resistance to adjust across YZ)?

If so, we may have the fix for a weak jug.

patscc
February 3rd, 2009, 06:13 PM
Well, that network does supply the DC bias by means of the first grid, which is one way of adjusting brightness.
Y & Z do look like that, yes. I'd say they would run the brightness control up to max, and the add another resistor in parallel to turn it up to 11, as the expression goes, and then go elsewhere to limit the beam current so that there's a margin for error/safety if the user brightness control is set to max.

If I were twiddling with it, I'd add a 1 kOhm resistor & 1 kOhm pot in series between Y & Z and see what happens. If OP doesn't make sense of this, I can draw a diagram.

I'd still like to see some pictures. The other popular problem for loss of brightness, aside from cathode deterioration, is B+ not at the correct value. That's why pictures would be good, since you can typically tell from the screen if shrinkage is occuring.
patscc

BG101
February 3rd, 2009, 08:58 PM
I would say using a plain black background, turn the contrast to minimum then set brightness halfway, then adjust the A1 (G2?) (we probably call them different here) until the raster is just visible. That should give you a proper brightness range.

You may have to set up the greyscale though which is a bit more tricky - the check for this is a neutral background colour at any brightness level. I've no idea if monitors of that vintage have greyscale tracking (early Sony trinitrons did, amongst other TV sets so I don't see why not).


BG

patscc
February 4th, 2009, 11:54 AM
From reading BG101's post I took another closer look at the schematic, and sure enough, there's an adjustment for G2 which escaped us all earlier. It's looks to be near where the focus adjustment is located. If it's not a plastic knob, use a plastic or insulated screwdriver.
Tweaking that might fix OP's problem.

I'd worry about grayscale if the tube's really worn out (and you'll start seeing a greenish tinge). There's no tracking on the 5153 from what I can see, you'd actually have to adjust bias to each of the guns. And seeing as how the monitor can't really display grayscale anyway...
patscc

BG101
February 4th, 2009, 02:22 PM
Sorry I should have mentioned the G2 adjustment is normally on the line output (flyback) transformer next to the Focus pot. These are usually plastic ridged knobs with a screwdriver slot in them.


BG

Great Hierophant
June 16th, 2009, 08:03 PM
I have fiddled with every knob in my 5153. I was able to boost the intensity, with a combination of the R,G,B adjustments and G2 to make the color appear, although it is still a bit dark. The only way to bring it up to a more acceptable brightness level using the pots will also show very noticeable horizontal retrace lines in text mode. Surely there must be a better way. I think this monitor needs the attention of an expert. Anyone willing to try?