I tried to research this and on the first page of results is the Amibay thread where you bought the card lol.

My guess here is that it simply supports VGA-like video modes (on a multisync monitor) but not as the default text mode, since EGA text is already so close to VGA text that you wouldn't gain much by supporting it through the DIP switches.

If you had some software that specifically supports this card, it could probably put it into the "VGA" modes.

There's also this: https://forum.vcfed.org/index.php?threads/qua … ut-there.60175/

Nice little card 😀 Shame there doesn't seem to be much info about it online but crossing fingers the manual might pop up someday! Just a thought, have you tried installing it into a different system? Maybe the modes where it doesn't POST would actually POST in another machine? (e.g. 286+)

dionb wrote on 2025-04-25, 23:41:

Even aside from not getting anything like VGA, this confuses me completely - can't make head or tail of what's happening here. Whatever it is, it's all TTL though.

The DIP switches are most likely the standard EGA DIP switch settings, see http://www.techhelpmanual.com/65-ega_switch_settings.html for example. I compared your table with that table, and for all the 12 combinations listed on the page, they match up. If there is a different combination to explicitly set "multisync" or "VGA-like" monitor, you would expect to get at least EGA high-res on boot, which you don't: Two of the extra combinations crash the video BIOS during initialisation, and the other two result in MDA and CGA timings.

Furthermore, the card has no hardware to output analog video. It definitely can not be connected to a VGA monitor. Nevertheless, it is quite likely that the name EGAVGA somehow alludes to some VGA-inspired capabilities. In this case, I suggest that this card is just another late Super-EGA card that supports 640x480 on digital TTL multisync monitors, possibly at VGA frequencies, but usually those cards used lower dot clocks and less blanking than the quite generous VGA timing. Try setting up the card as "primary EGA with enhanced display" (OCCO from your table) and then blindly set video mode 12h (640 x 480, 16 colors) as you would do on a VGA card. I wouldn't be surprised if this is the way to get that card into 640 x 480 mode.

dionb wrote on 2025-04-25, 23:41:

I then tried to change the jumper (J4) from 1-2 (? -not marked, but 'up' when mounted) to 2-3 ('down'). This seemed to make green look a bit more washed-out in EGA hi-res mode, but no other changes.

Typically, you expect a jumper on EGA cards to choose what pin 2 on the EGA monitor connector is used for. The original CGA monitor connector specification had ground on pins 1 and 2, whereas EGA re-purposed pin 1 to "secondary red". The "secondary" color pins for "secondary red" and "secondary blue" are only used in EGA high-res mode (matches your observation that the jumper only has an effect in hi-res mode). Light green is supposed to be "secondary red + full green + secondary blue", with "secondary" meaning "33%", "primary" meaning "66%" and "full" meaning "secondary + primary = 100%". If you set the pin-2 jumper to "always ground", as required for some CGA monitors that require ground on pin 2, you instead get "no red + full green + secondary blue", which is a slightly blueish, but more satured green than the intended "light green". I'd guess the "more washed out green" is the "acutal EGA light green", and the "less washed-out green" is "EGA light green without red". You can easily buzz out whether I am correct by checking that the central pin of JP4 is connected to pin 2 of the monitor jack and one of the outer pins is connected to ground.

I think same. Mode 12h and 11h (mono) are worth a try.

This is just one of gazillion EGA cards/clones made back then. Most of them had 256K VRAM too.

http://bitsavers.org/components/chipsAndTech/ … _436_198808.pdf

Hi, thanks! It's an Super EGA card, essentially. It has both 640x480 and 800x600 support (reqs. 38 MHz oscillator).
Hercules is being mentioned, too. That's good, I think.
With the right BIOS support it should be able to do Standard VGA (mode 12h) just fine.
Given that no custom palettes are being used, I suppose.
Normal DOS applications should run fine. Games with mode 13h ("MCGA") in mind are a different story.

mkarcher wrote on 2025-04-26, 09:30:

[...]

The DIP switches are most likely the standard EGA DIP switch settings, see http://www.techhelpmanual.com/65-ega_switch_settings.html for example. I compared your table with that table, and for all the 12 combinations listed on the page, they match up. If there is a different combination to explicitly set "multisync" or "VGA-like" monitor, you would expect to get at least EGA high-res on boot, which you don't: Two of the extra combinations crash the video BIOS during initialisation, and the other two result in MDA and CGA timings.

OK tnx (and should have thought of looking at default EGA myself...)

mkarcher wrote on 2025-04-26, 09:30:

[...]
Furthermore, the card has no hardware to output analog video. It definitely can not be connected to a VGA monitor. Nevertheless, it is quite likely that the name EGAVGA somehow alludes to some VGA-inspired capabilities. In this case, I suggest that this card is just another late Super-EGA card that supports 640x480 on digital TTL multisync monitors, possibly at VGA frequencies, but usually those cards used lower dot clocks and less blanking than the quite generous VGA timing. Try setting up the card as "primary EGA with enhanced display" (OCCO from your table) and then blindly set video mode 12h (640 x 480, 16 colors) as you would do on a VGA card. I wouldn't be surprised if this is the way to get that card into 640 x 480 mode.

Jo22 wrote on 2025-04-26, 15:26:

Hercules is being mentioned, too. That's good, I think. With the right BIOS support it should be able to do Standard VGA (mode 1 […]
Show full quote

Hercules is being mentioned, too. That's good, I think.
With the right BIOS support it should be able to do Standard VGA (mode 12h) just fine.
Given that no custom palettes are being used, I suppose.
Normal DOS applications should run fine. Games with mode 13h ("MCGA") in mind are a different story.

Quick check: no, it can't display Doom (mode Y), but yes, it can display 12h 😀

Except for the palettes...

(my testbench is a mess and I can't find either a NIC or one of my CF adapters+cards - this was the only thing on the default HDD that would output 12h)

mkarcher wrote on 2025-04-26, 09:30:

[...]
Typically, you expect a jumper on EGA cards to choose what pin 2 on the EGA monitor connector is used for. The original CGA monitor connector specification had ground on pins 1 and 2, whereas EGA re-purposed pin 1 to "secondary red". The "secondary" color pins for "secondary red" and "secondary blue" are only used in EGA high-res mode (matches your observation that the jumper only has an effect in hi-res mode). Light green is supposed to be "secondary red + full green + secondary blue", with "secondary" meaning "33%", "primary" meaning "66%" and "full" meaning "secondary + primary = 100%". If you set the pin-2 jumper to "always ground", as required for some CGA monitors that require ground on pin 2, you instead get "no red + full green + secondary blue", which is a slightly blueish, but more satured green than the intended "light green". I'd guess the "more washed out green" is the "acutal EGA light green", and the "less washed-out green" is "EGA light green without red". You can easily buzz out whether I am correct by checking that the central pin of JP4 is connected to pin 2 of the monitor jack and one of the outer pins is connected to ground.

*beep*

Spot on. Central pin is connected to DE9 pin 2, J4 pin 3 (well, bottom one, at least) is connected to GND. So J4 2-3 connects DE9 pin 2 to GND - and that was the jumper setting that looked 'washed out'

georgel wrote on 2025-04-26, 12:55:

This is just one of gazillion EGA cards/clones made back then. Most of them had 256K VRAM too.

http://bitsavers.org/components/chipsAndTech/ … _436_198808.pdf

Hmmph. If Quadram were still around and I had bought this card new, I'd have a mind to throw the Trade Descriptions Act at them for the "VGA" bit in EGAVGA. Caveat emptor at this point though 😉

Nice relatively high-end EGA card though.

Jo22 wrote on 2025-04-26, 15:26:

Hi, thanks! It's an Super EGA card, essentially. It has both 640x480 and 800x600 support (reqs. 38 MHz oscillator).

No such luck, it's a 16.257MHz one.

There is an empty pad V12 for a second oscillator though...

dionb wrote on 2025-04-26, 22:32:

Jo22 wrote on 2025-04-26, 15:26:

Hi, thanks! It's an Super EGA card, essentially. It has both 640x480 and 800x600 support (reqs. 38 MHz oscillator).

No such luck, it's a 16.257MHz one.

There is an empty pad V12 for a second oscillator though...

Most likely, the oscillator on the card is a dual-frequency oscillator: 16.257MHz for high-res EGA and MDA-compatible modes, and around 24MHz for 640x480. Only the 800x600 mode requires the separate 38MHz oscillator that's not equipped on your card. 24MHz also is sufficient to generate a 132-column text mode, which this card might also implement.

mkarcher wrote on 2025-04-27, 10:31:

[...]

Most likely, the oscillator on the card is a dual-frequency oscillator: 16.257MHz for high-res EGA and MDA-compatible modes, and around 24MHz for 640x480. Only the 800x600 mode requires the separate 38MHz oscillator that's not equipped on your card. 24MHz also is sufficient to generate a 132-column text mode, which this card might also implement.

The oscillator is a Kyocera KXO-01, which sems to be a fixed-frequency device.

In the specs of the P82C435 you linked to, it says:

CLKSELO and CLKSEL 1 (bits 2 and 3 of the Miscellaneous Output Register) determine which frequency is used: 14.317 MHz, 16.25 MH […]
Show full quote

CLKSELO and CLKSEL 1 (bits
2 and 3 of the Miscellaneous Output Register) determine
which frequency is used: 14.317 MHz, 16.25 MHz (up to
25MHz) external crystal or external oscillator (up to 25MHz).

Given the "(up to 25MHz) [...] or external oscillator" it sounds like it's able to do 25MHz based on a 16.25MHz input.

dionb wrote on 2025-04-27, 11:57:

The oscillator is a Kyocera KXO-01, which sems to be a fixed-frequency device.

Oh, that's interesting then. Can you determine the video frequencies Mode 12h runs at?

dionb wrote on 2025-04-27, 11:57:

In the specs of the P82C435 you linked to, it says:

P82C435 datasheet wrote:

CLKSELO and CLKSEL 1 (bits 2 and 3 of the Miscellaneous Output Register) determine which frequency is used: 14.317 MHz, 16.25 MHz (up to 25MHz) external crystal or external oscillator (up to 25MHz).

Given the "(up to 25MHz) [...] or external oscillator" it sounds like it's able to do 25MHz based on a 16.25MHz input.

No, it's not. You snipped the sentence before, which says "CLKIN is the clock input from an external multiplexer (LS153)". The 74LS153 chip is a (dual) 4-to-1 multiplexer, U16 on your card. This chip will forward one of four clock signals to CLKIN, depending on CLKSEL0 and CLKSEL1. Furthermore, this chip will also forward one of the four DIP switches to the pin SWITCH of the 82C436 bus interface chip. This enables the BIOS to read the four switches with only requiring one data line to be connected to the switch interface circuit.

The four clock inputs of the 74LS153 are supposed to be connected to the 14.318MHz clock from the ISA bus, the 16.257MHz clock from the local oscillator, the "external clock" pin on the feature connector and (on the original EGA) one undefined input. This undefined input is typically used for higher resolutions on Super-EGA cards. The datasheet allows clocks up to 25MHz. It's interesting to the 25MHz limit at that place, as there are two variations of the P82c435: The standard variant (up to 25MHz) and the 38MHz variant (up to 38MHz, as the name implies). If your card has no 25MHz clock source, mode 12h is likely running on the 16MHz oscillator as well, resulting in something like 21.6kHz / 45Hz.

georgel wrote on 2025-04-26, 12:55:

This is just one of gazillion EGA cards/clones made back then. Most of them had 256K VRAM too.

http://bitsavers.org/components/chipsAndTech/ … _436_198808.pdf

That's good to know, actually, because quite a few EGA games expect 256KB.
Example: https://www.mobygames.com/game/22616/chip-n-d … s-castle/specs/

That's one of the reasons why I like VGA cards. They have full EGA memory expansion as default, so things "just work".
A minor downside is the loss of original 60Hz refresh rate, though.
EGA games running on VGA cards run at ~70-72 Hz, I think.

Jo22 wrote on 2025-04-27, 19:13:

EGA games running on VGA cards run at ~70-72 Hz, I think.

At 70Hz (instead of 60Hz). Except if they are designed to run at half the frame rate. In that case, they run at 35Hz instead of 30Hz. With a patched video mode parameter table, you can get the VGA card to add that much vertical blanking to EGA modes that the the refresh rate drops down to 60Hz, and a lot of VGA monitors will still display the image perfectly.

mkarcher wrote on 2025-04-27, 13:41:

dionb wrote on 2025-04-27, 11:57:

The oscillator is a Kyocera KXO-01, which sems to be a fixed-frequency device.

Oh, that's interesting then. Can you determine the video frequencies Mode 12h runs at?

Not at present, which is irritating the hell out of me - so I've bitten the bullet and bought an oscilloscope. It should arrive next week, at which point I should be able to answer this.

dionb wrote on 2025-05-02, 23:04:

mkarcher wrote on 2025-04-27, 13:41:

Oh, that's interesting then. Can you determine the video frequencies Mode 12h runs at?

Not at present, which is irritating the hell out of me - so I've bitten the bullet and bought an oscilloscope. It should arrive next week, at which point I should be able to answer this.

There is software to measure the frequency as well. I think vidspeed does a H/V frequency measurement as well as (video) memory speed measurement.

mkarcher wrote on 2025-05-03, 06:22:

[...]

There is software to measure the frequency as well. I think vidspeed does a H/V frequency measurement as well as (video) memory speed measurement.

Might try that this weekend - but the oscilloscope arrived yesterday and today I had time to play around with it. Took me a while to figure out the pinout of the feature connector (only place I could get the horizontal frequency using my probe), as there was no pin indication on the card and the manual of the IBM EGA has it in a different position to how it is on the card. In the end it's exactly the same as on the IBM EGA.

But I have it - and it's interesting...

In DOS text mode: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - really clean
In "mode 12" high-res: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - with a lot of noise
In low-res EGA: 15.7kHz horizontal, 59.95Hz horizontal (positive polarity)

dionb wrote on 2025-05-08, 17:43:

In DOS text mode: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - really clean
In "mode 12" high-res: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - with a lot of noise

So you actually get a 640*350 output in "mode 12h". If by "noise" you mean the 59.74Hz vertical has some jitter on it, it might be that the 640*480 mode is a letterboxed 640*700 interlaced mode.

mkarcher wrote on 2025-05-08, 21:24:

dionb wrote on 2025-05-08, 17:43:

In DOS text mode: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - really clean
In "mode 12" high-res: 21.85kHz horizontal, 59.74Hz vertical (negative polarity) - with a lot of noise

So you actually get a 640*350 output in "mode 12h". If by "noise" you mean the 59.74Hz vertical has some jitter on it, it might be that the 640*480 mode is a letterboxed 640*700 interlaced mode.

These are the waveforms-

DOS text:

Hires "mode 12":

dionb wrote on 2025-05-09, 15:42:

Hires "mode 12":

The attachment Quadram_EGAVGA_hires_hsync.png is no longer available

What you see here is likely cross-talk from the video signal, while the text mode is probably taken with a mostly black screen. The EGA hi-res mode is known for squeezing the horizontal timing as much as possible, i.e. display up to the latest microsecond before the beam starts retracing. What you see in this picture is that obviously the video signal continues for a short moment into the retrace pulse. This is allowable, as the monitor takes some time to start the retrace process in response to the HSYNC line going high, so up to the time the horizontal sweep electronics requires to recognize and respond to the pulse, the beam continues to advance normally and can display some pixels. During the retrace period, there is no "noise", it starts only after the horizontal deflection re-settled the beam to scan the next line.

If you want to know whether the video is interlaced, if you have two channels of that scope, take a graph of both VSYNC and HSYNC at the same time. Trigger on the (falling) edge of VSYNC, and check whether you get a stable relation between VSYNC and HSYNC (progressive video), or you get two different variants of HSYNC overlapping each other (interlaced video).

Little update:
Today I unexplectedly got my hands on a proper TTL EGA monitor (without having to sell a kidney), an AOC CM-312, so I used the EGAVGA to put it through its paces. It's beautiful, after fine-tuning the phase controls it displays 15.75kHz low-res and 21.85kHz high-res perfectly with nice subtle scanlines.

But...

Having tried a pile of EGA games and things with VGA mode, it's pretty clear: this is an EGA card pure and simple and it can't display regular VGA mode 12. In Ultris for example it just gives a black screen when you select VGA. Feed it other VGA modes and the PC even reboots in some cases (eg Bloodwych). Not sure exactly what's going on with that Colonization install screen that did display (albeit without correct palette), but I suspect it's doing mode 10 640x350 and is the one thing that the EGAVGA can display.