• Please review our updated Terms and Rules here

PGC issues, displays only blue in CGA emulation

thisisamigaspeaking

Experienced Member
Joined
Feb 13, 2024
Messages
78
Wanted to get some feedback here before I go any further. My PGC powers and the CGA emulation does work, but it displays only blue. At first it was very dim, and I think it actually displayed only green (or maybe it was just very hard to see). It only works on one of my two LCDs that are appropriate for this. The cable's continuity checks out. The cable has the PGC's pin 5 (resolution select) connected to pin 14 of the VGA (V sync), and also has a connection from pin 4 (composite sync) on VGA's pin 13 (H or composite sync) of course, but removing that connection to 14 doesn't change anything.

pgc-problem.jpeg
I'm able to fix the sync/size/location issues with the monitor's settings so that's not a concern.

I am wondering if this is likely to be a DAC issue? Would it be reasonable to just go ahead and replace the DAC? I'm not sure if this is a failure mode of a video DAC or if it is more likely to mean the driver chips/circuits for red and green are not working. It's just a combined 4-4-4 bit DAC, not a RAMDAC afaik.

I think the PGC's CPU (or at least the CPU's general part of the boards) has to initialize for the CGA emulator to come on at all.

I have some limited electronics experience from may years ago, and know how to use the basic tools (solder, desolder, oscilloscope, etc.) but maybe I should leave this to someone with more experience? I would need to buy a scope but I have been thinking about getting a 200 MHz digital scope for a while.

I do have a logic diagram for the board.

Another thing is that DOS gives some kind of error beep when it loads and the video cuts out. I could try disabling the PGC's CGA emulation and putting another video card in it so I can run the PGC diagnostics, but the problem with VGA in general is that the usual BIOS area at C0000h (other than IBM VGA) overlaps with with the PGC's interface at C6000h.

Ideas?
 
Last edited:
It only works on one of my two LCDs that are appropriate for this.
You are trying to display IBM CGA on a VGA monitor without an IBM-CGA-to-VGA type converter unit in between (noting that per [here], some CGA-to-VGA converters are unsuitable.)

How do you know that the Brilliance 200P3 is "appropriate" for this? For example, putting aside the sync, IBM CGA uses RGBI signaling, which is {red on/off + green on/off + blue on/off + intensity on/off}. Does the Brilliance 200P3 have an input pin for 'intensity' ?
 
Another thing is that DOS gives some kind of error beep when it loads and the video cuts out. I could try disabling the PGC's CGA emulation and putting another video card in it so I can run the PGC diagnostics, but the problem with VGA in general is that the usual BIOS area at C0000h (other than IBM VGA) overlaps with with the PGC's interface at C6000h.
Actually I'm not sure that is an error beep or even related to the PGC, since I can't see on the screen. I don't remember what sounds were made. The video does cut out once DOS loads though. Waiting on some additional parts (scope, a PS/2 with VGA that should not conflict with the PGC) to do more testing.

You are trying to display IBM CGA on a VGA monitor without an IBM-CGA-to-VGA type converter unit in between (noting that per [here], some CGA-to-VGA converters are unsuitable.)

How do you know that the Brilliance 200P3 is "appropriate" for this? For example, putting aside the sync, IBM CGA uses RGBI signaling, which is {red on/off + green on/off + blue on/off + intensity on/off}. Does the Brilliance 200P3 have an input pin for 'intensity' ?
It's a PGC card. AFAIK, it's a CGA emulation and should be routed through the analog DAC and displayed at 640x400 60Hz (or something to that effect).
 
It's a PGC card. AFAIK, it's a CGA emulation and should be routed through the analog DAC and displayed at 640x400 60Hz (or something to that effect).
So the PGC card is seen from a hardware perspective as a CGA card, and the PGC card does a kind of CGA-to-PGA conversion.
So the question becomes, is the Brilliance 200P3 capable of displaying PGA.
I see at [here] that a member used a "standard 9-pin to 15-pin converter" to get a display going on a Samsung SyncMaster 500s Plus.
 
So the PGC card is seen from a hardware perspective as a CGA card, and the PGC card does a kind of CGA-to-PGA conversion.
So the question becomes, is the Brilliance 200P3 capable of displaying PGA.
I see at [here] that a member used a "standard 9-pin to 15-pin converter" to get a display going on a Samsung SyncMaster 500s Plus.
PGC/PGA should be the same as VGA (or close) aside from the composite sync, which this monitor does support. So a normal 9-pin to 15-pin RGB cable should work. I tried removing the connection from "resolution select" (pin 5) on the PGC's port to V sync on the VGA side and it didn't help. Otherwise it should all be within the capabilities of the monitor.
 
This probably *isn't* your problem, but I'm going to toss this out there:

Your POST message says you're running at 9.5Mhz; unless this board you're trying to run it on has a system for slowing down when accessing the expansion slots there's a very good chance that bus speed is going to make a PGA card unhappy. That thing came out around the same time as the 6mhz AT, and I'm fairly certain I've heard people complain about it sometimes being finicky to even get it running on an 8mhz bus.
 
So the PGC card is seen from a hardware perspective as a CGA card, and the PGC card does a kind of CGA-to-PGA conversion.
So the question becomes, is the Brilliance 200P3 capable of displaying PGA.
I see at [here] that a member used a "standard 9-pin to 15-pin converter" to get a display going on a Samsung SyncMaster 500s Plus.
The monitor is obviously capable since it's syncing...

Another thing is that DOS gives some kind of error beep when it loads and the video cuts out.
This is a red flag to me. Sounds like the system BIOS is fighting the PGC BIOS. I think you need to configure your system for EGA/VGA so only the PGC BIOS handles video.
 
The monitor is obviously capable since it's syncing...


This is a red flag to me. Sounds like the system BIOS is fighting the PGC BIOS. I think you need to configure your system for EGA/VGA so only the PGC BIOS handles video.
Seems like it should be set to CGA otherwise it is going to try to look for a EGA/VGA BIOS to try to boot instead of using the BIOS's own CGA routines, correct? The PGC will work like CGA until a PGC driver activates its advanced mode. PGC doesn't have EGA/VGA functionality (BIOS to load) afaik.

I was getting the same sound afterward when I removed the card and tried to boot again. I don't remember what sounds the system made at boot but now that I think about it, that sound may be because I removed the EMS card and the driver is still trying to load. That the card shuts off there may be a clue (doesn't like what the EMS is trying to access perhaps?).

I am waiting on some test equipment to arrive. Also am obtaining an ISA PS/2 which has built-in VGA which I believe I can use along side the PGC to help debugging (disable the PGC's CGA emulation and boot to VGA until I am ready to try to test the PGC board). I'm kind of stuck right now due to the fact that the on board VGA stopped working on my NuXT, which I also need the equipment (scope) to take a look at. I can't use the NuXT's board with the PGC anyway (even with its CGA emulation jumpered off) because the BIOS location conflicts with the PGC's interface.
 
I was under the (mistaken) impression that PGC had a BIOS extension, since VGA/PGA/EGA is a common AMIBIOS option. The system BIOS doesn't specifically "look" for an EGA or VGA BIOS; when you select this option the system BIOS just doesn't care about video and lets whatever option ROM is loaded handle it (which is why XTs that predate EGA will have this option as "none").

But you are correct, PGC has no BIOS so you need to use the CGA setting. So it must be something else. I would definitely try running it at non-turbo, and check all the low-hanging fruit like monitor cable and loose connections.
 
Just as a sanity check, did you build this conversion cable yourself? Are you sure all the grounds are properly routed and it’s not possibly doing something that might be sketchy for the PGC? (It’s probably a long shot, but maybe the PGC doesn’t like having its RGB returns tied into a common ground or something else the cable might be doing.)

As for getting another video card alongside the PGC for sanity checking what’s happening an obvious choice, if you can manage a monitor for it, would be an MDA/Hercules monochrome card. No BIOS clashes and with the motherboard switches set to mono the computer will utterly ignore the PGC and its onboard CGA emulator unless you switch over with a “MODE CO80” command.
 
So I've tried a couple things. I tried a RGB to VGA converter board and it says no signal. I've hooked a scope up to the color lines and indeed, only blue is outputting meaningful information, red and green are outputting noise.

Would it be reasonable to try to replace the DAC? It's a simple 3 channel DAC, not a RAMDAC. I found a "near" match. I assume it is the same chip but it has slightly different markings.

Original:

Code:
intech
MADE IN USA
33967
VDAC444TD
TRIPLE 4 BIT
VIDEO DAC

Replacement:

Code:
intech
MADE IN USA
33967
RGB DAC 4TD
TRIPLE 4 BIT
COLOR MAPPED VIDEO DAC

Same part number, same appearance, differs in VDAC444TD vs. RGB DAC 4TD. Sounds the same to me.

This is a valuable board so I hesitate to mess around with it. On the other hand, the scope verifies that it is indeed not working properly. I don't have experience repairing things like this, just basic circuit board experience. Opinions?
 
I would try to verify what's going into the DAC first. Display a screen full of red, or green only text, and compare it to the signals for blue only text. If there's nothing going into the DAC on the red, and green, then it's probably not the problem. A datasheet for the DAC chip would be helpful.
 
I would try to verify what's going into the DAC first. Display a screen full of red, or green only text, and compare it to the signals for blue only text. If there's nothing going into the DAC on the red, and green, then it's probably not the problem. A datasheet for the DAC chip would be helpful.
Makes sense. Don't have a datasheet but do have a pinout for it, and complete logic diagram for all 3 circuit boards. The 3 board sandwich design is going to make it a bit hard to test most things while the card is running but the pins for the DAC are exposed. The CGA emulation seems to work anyway, and that is the middle board, the two outer boards are somewhat accessible - considering it needs to be attached to a motherboard.

What made me think that it was the DAC was that it's just significant noise coming out of the video port on red and green, it's not any kind of clocked signal, compared to the blue.

TD444.png
 
Last edited:
It looks like it should be easy to solder some short wires to some of the input pins to compare signals. If the red, green, and blue inputs look similar, then it almost has to be the DAC itself.
 
It looks like it should be easy to solder some short wires to some of the input pins to compare signals. If the red, green, and blue inputs look similar, then it almost has to be the DAC itself.
Great, thanks. I didn't think of soldering wires to test things while it is in the system.
 
Not directly related, but for anyone who comes across this thread, here's someone working on a 3dfx card with a similar problem, just for an idea of the kind of probing to do.
 
Back
Top