OK, I found a SC15025CV 110MHz.
Shall I buy it?
😁

CoffeeOne wrote on 2023-12-07, 20:36:

OK, I found a SC15025CV 110MHz.
Shall I buy it?
😁

Looking at the data sheet, the command register of this DAC is again different, so it would need BIOS patches to work in Hicolor / TrueColor modes.

Have ICS RAMDACs already been considered? I think the Hercules et4000w32 cards often used AT&T or ICS. The Dynamite Power I have uses an ICS "gendac" in a 68 pin package, but I remember seeing photos of a similar card with a 44pin ICS chip. Not sure if it was "gendac" or not.

Anonymous Coward wrote on 2023-12-08, 02:16:

Have ICS RAMDACs already been considered? I think the Hercules et4000w32 cards often used AT&T or ICS. The Dynamite Power I have uses an ICS "gendac" in a 68 pin package, but I remember seeing photos of a similar card with a 44pin ICS chip. Not sure if it was "gendac" or not.

I found something that indicated that chrontel CH8391 might be compatible with AT&T, and it should be available as a 135MHz part in the 44 pin package.

https://stuff.mit.edu/afs/sipb/system/i386_nb … /html/S3-4.html

Anonymous Coward wrote on 2023-12-08, 07:28:

Anonymous Coward wrote on 2023-12-08, 02:16:

Have ICS RAMDACs already been considered? I think the Hercules et4000w32 cards often used AT&T or ICS. The Dynamite Power I have uses an ICS "gendac" in a 68 pin package, but I remember seeing photos of a similar card with a 44pin ICS chip. Not sure if it was "gendac" or not.

I found something that indicated that chrontel CH8391 might be compatible with AT&T, and it should be available as a 135MHz part in the 44 pin package.

https://stuff.mit.edu/afs/sipb/system/i386_nb … /html/S3-4.html

I cannot find data sheets of the ICS 44 pin chip nor of the Chrontel 8391. The Chrontel 8391 should be available in 110MHz and 125MHz, what is the exact part number of the 2?
The best option seems to be really the orginal AT&T 20C490-11, the 20C491-11 should be 100% compatible, too.

The ICS RAMDAC on my Cardex Challenger 9304-20 is an ICS5301-2. I believe it's a 110MHz part, but according to the pdf it was also available as a 135MHz part:

Anonymous Coward wrote on 2023-12-08, 14:46:

The ICS RAMDAC on my Cardex Challenger 9304-20 is an ICS5301-2. I believe it's a 110MHz part, but according to the pdf it was also available as a 135MHz part:

Thx for the datasheets.
The 5301 does not look promising: "Compatible to Tseng Labs" chips, so no.
But maybe the 5300 works on the Ark. 5300-3 seem to be super rare, it might be possible to get a 5300-2.

The CH8391v seems to be available. So there is only one type, which is good for 110MHz or 135MHz (?)
I am not 100% convinced that it is compatible.
Maybe @mkarcher can have a quick look?

Probably the ICS5301 isn't viable. But according to the Chrontel datasheet, the 8391 should be compatible with the AT&T 490/491/492.
My impression is that all CH8391s are capable of 135MHz VCLK.

I have already demonstrated that the ATT20C490-11 can provide a crisp image at 1280x1024x256c at 60 Hz. Since we do not have a VGA BIOS hack to attempt to force 1024x768x65k at 56/60 Hz, what is the benefit on trying to source faster RAMDACs for this card?

It was stated that 1024x768x64k @60Hz requires a 130MHz RAMDAC.
Now that a fast enough RAMDAC is potentially available, is the BIOS hack worth the effort?

I suspect the 110 MHz ATT RAMDACs will suffice for simple testing at 120 or 130 MHz, afterall, the 80 MHz RAMDACs worked for quick tests at 110 MHz. The uncertainty concerning RAMDAC compatibility may be an unwanted variable at this point. Mkarcher did mention that 120MHz may be sufficient by "reducing blanking (works on LCDs, but likely not on CRTs)".

There is a strong possibility that nobody has the time or desire to work on this BIOS hack, thus I was wondering if you guys are getting too far ahead at this point.

On the flip side, I would still be interested if any of these potentially compatible RAMDACs provide an even sharper image than the ATT490, which is already pretty clear.

feipoa wrote on 2023-12-09, 08:06:

I suspect the 110 MHz ATT RAMDACs will suffice for simple testing at 120 or 130 MHz, afterall, the 80 MHz RAMDACs worked for quick tests at 110 MHz. The uncertainty concerning RAMDAC compatibility may be an unwanted variable at this point. Mkarcher did mention that 120MHz may be sufficient by "reducing blanking (works on LCDs, but likely not on CRTs)".

There is a strong possibility that nobody has the time or desire to work on this BIOS hack, thus I was wondering if you guys are getting too far ahead at this point.

On the flip side, I would still be interested if any of these potentially compatible RAMDACs provide an even sharper image than the ATT490, which is already pretty clear.

For me it was not about to get something better, something that works would be good enough. I was not able to find something except Upsource, which I don't like.
I now found and ordered a CH1391f.

CoffeeOne wrote on 2023-12-09, 13:04:

feipoa wrote on 2023-12-09, 08:06:

I suspect the 110 MHz ATT RAMDACs will suffice for simple testing at 120 or 130 MHz, afterall, the 80 MHz RAMDACs worked for quick tests at 110 MHz. The uncertainty concerning RAMDAC compatibility may be an unwanted variable at this point. Mkarcher did mention that 120MHz may be sufficient by "reducing blanking (works on LCDs, but likely not on CRTs)".

There is a strong possibility that nobody has the time or desire to work on this BIOS hack, thus I was wondering if you guys are getting too far ahead at this point.

On the flip side, I would still be interested if any of these potentially compatible RAMDACs provide an even sharper image than the ATT490, which is already pretty clear.

For me it was not about to get something better, something that works would be good enough. I was not able to find something except Upsource, which I don't like.
I now found and ordered a CH1391f.

Shit.
I tried a CH1391f. Does not work correctly, even low resolution modes are broken. 640x480 256 colours in Windows 98 shows completely wrong colours.

As long as it didn't brick anything.

Maybe the "F" in the part number indicates some kind of change. I don't remember seeing that in the datasheet. There is only CH1391i and CH1391v.

Anonymous Coward wrote on 2023-12-23, 08:17:

Maybe the "F" in the part number indicates some kind of change. I don't remember seeing that in the datasheet. There is only CH1391i and CH1391v.

Sorry, Typo.
Yes, the card or some chips might be bricked already.
I will change back to the Ark 1491 and check if the card still works.

Is there sufficient space for a PLCC-44 socket?

feipoa wrote on 2023-12-23, 13:02:

Is there sufficient space for a PLCC-44 socket?

I guess so.
Is this the correct type?
https://www.ebay.at/itm/203060924189
?

I think so.