SSTV2 wrote on 2021-05-13, 01:18:

Regarding wire patches - there are three main differences between our boards, your MB lacks a wire patch between U56 pin 76 and RP27 pin 6 and a patch that links U26 pin 2 with pin 10, also, your board has a link between U19 pin 9 and U39 pin 6, which my board lacks. Perhaps link U56->RP27 is formed on the bottom side of your board?

Sorry for the delay -- it seems I don't get any notification about activity here.

I'm reluctant to pull the board out of this system since it seems the little plastic clips are brittle and I've already broken one. It looks like you were able to get everything up and running; do you need anything else from this board?

-Hinoserm

Hinoserm wrote on 2021-05-25, 20:46:

I'm reluctant to pull the board out of this system since it seems the little plastic clips are brittle and I've already broken one.

Ouch, if you haven't fixed that clip yet, I'd recommend using a two part epoxy glue for the repair job + you could drill small holes on both ends of the plastic pieces and insert short springy metal rods into them so that clips could be bent again multiple times w/o breaking.

Hinoserm wrote on 2021-05-25, 20:46:

It looks like you were able to get everything up and running; do you need anything else from this board?

All the major issues have already been fixed, thanks for asking 😀, though there are still some unanswered questions.

In the TH99 documentation for this board, it's mentioned that J35 connector is used for IDE interface LED, is that true? I can see that the front panel header has three 220 Ohm resistors connected to it (typical current limiting resistor value for LEDs), which I presume are for POWER, "TURBO" and IDE activity then what J35 is actually for?

Could you run some benchmarks on your Premmia for comparison (1 speedsys bench if possible)? I would like to know whether your board suffers from the same memory throughput limitation or if it was fixed with the slightly different wire patching.

I recently got my hands on a huge lot of older motherboards. Among those boards are 3x AST Premmia EISA 486 boards like the one OP has. I noticed that instead of having the two 6 pin AT power supply connectors next to each other these boards have the power connectors stacked on top of each other. Hence, I can not apply the black to black rule in order to connect a power supply to any of these boards. I wanted to ask if someone can chime in with the pinouts of the power connectors for these boards. I haven't been able to find any information anywhere and I don't want to take a gamble and possibly damage these boards. They already have seen enough abuse. i.e. On one of the boards the TI chip closest to the EISA slots is partially unsoldered. Thanks and sorry to revive an older thread.

Nikola99 wrote on 2021-06-27, 07:47:

I recently got my hands on a huge lot of older motherboards. Among those boards are 3x AST Premmia EISA 486 boards like the one OP has. I noticed that instead of having the two 6 pin AT power supply connectors next to each other these boards have the power connectors stacked on top of each other. Hence, I can not apply the black to black rule in order to connect a power supply to any of these boards. I wanted to ask if someone can chime in with the pinouts of the power connectors for these boards. I haven't been able to find any information anywhere and I don't want to take a gamble and possibly damage these boards. They already have seen enough abuse. i.e. On one of the boards the TI chip closest to the EISA slots is partially unsoldered. Thanks and sorry to revive an older thread.

If you have a multimeter handy you could just measure from the psu connector pins to the (E)ISA power pins. the only things you won't easily measure like this are the power good signal and i think that's it though th PG signal tends not to be used a lot of the time.

Sorry for reviving such an old thread but I guess that goes hand in hand with old hardware and I may also answer many of the questions listed above, in context.

I just got an AST Premmia 4/66D that I managed to revive. Initial problems were due to acid damage on one of the smaller IC's at the top right corner (it was probably stored vertically). Thankfully, the acid dripping missed the CPU socket. Some vinegar and ample amounts of Norwegian tap water as well as electronic cleaner and it’s now posting. Bios version is 2.03

Some info:

This PC was in mint condition apart from the acid damaged IC. If anyone needs board pictures or help with the original configuration, just let me know. There are a few flimsy botch wires on my revision too.

Configuration:

-66 MHz 486, socketed, with the “Pentium Overdrive Ready” wording on the socket. The CPU surface mount option was clean/empty.
-ATI Mach 32 with fully populated memory slots. Shows 2 MB in the BIOS.
-256k cache module
-An EISA SCSI adapter, amply named EISA SCSI HA. One internal and one external port. Onboard Motorola 68000, which off course made me smile 😊. Meager results from google, and it seems this board with built to order by an unknown AST sub-supplier.
-Quantum SCSI, unknown size. Not tested but spins up and seems to initialize.
-64 MB of ram, over two dims.
-ISA Soundblaster (OPL3 included) (forgot the model number)
-ISA I/O card.

This PC must have cost a fortune back in the day!

Next steps:

Swapping the evil RAYOVAC 4.5V battery with a generic 3 x AA/AAA battery pack from some old LED Christmas lights or something. As far as I can tell, there is no need for a diode, and the black/red wires should be negative/positive. Any alkaline 4.5V combination should work according to various youtube videos.

Now, the biggest challenge is figuring out the EISA Config thing.

If anyone can shed some light on this, or even better, point to a source for the original EISA CONFIG UTILITY DISK, then I’d be very grateful. Archive.org has not produced anything yet.

I was fooling around with Amigas back when EISA was a thing, and completely missed out on this rather strange bus design. So, you need the disk, but do you also need files/codes or something for the cards, including the normal ISA cards?

Hello and welcome to Vogons!

No need to apologize for posting in this thread, it will remain relevant for as long as this forum exists, simply because of the somewhat under-discussed, rare and unusual Premmia 4/XX series of computers.

After I revived and tested the performance of the Premmia MB, I did not continue to assemble the full system. This was somewhat influenced by the very poor performance of the memory bandwidth and the non-standard dimensions of the board. For these reasons, I cannot answer the question of whether this system even requires the EISA config util, but as far as I know, the regular "Premmia" line of computers simply did not have a dedicated EISA config util. I know that the "AST Premium SE/TE" and "AST Premmia MTE/SE" line had their own dedicated EISA config utils. Also, in theory, the "Premium SE/TE" EISA config util should not be compatible with a regular "Premmia" because "Premium SE/TE" used an ASIC chipset designed by AST, while the "Premmia MTE/SE" EISA config util might work, since it may also be using the TI TACT84500 chipset (can't confirm). Try EISACU 2.84 on this webpage, it's meant for "Premmia MTE/SE", EISA config files are also there. You can also check these EISACU and CFG files uploaded by Horun.

By the way, congratulations on your new acquisition of old hardware 😉 The system seems to be nearly maxed out in terms of upgradeability, it would be interesting to see how well it performs.

SSTV2 wrote on 2023-09-12, 22:09:

Hello and welcome to Vogons! […]
Show full quote

Hello and welcome to Vogons!

No need to apologize for posting in this thread, it will remain relevant for as long as this forum exists, simply because of the somewhat under-discussed, rare and unusual Premmia 4/XX series of computers.

After I revived and tested the performance of the Premmia MB, I did not continue to assemble the full system. This was somewhat influenced by the very poor performance of the memory bandwidth and the non-standard dimensions of the board. For these reasons, I cannot answer the question of whether this system even requires the EISA config util, but as far as I know, the regular "Premmia" line of computers simply did not have a dedicated EISA config util. I know that the "AST Premium SE/TE" and "AST Premmia MTE/SE" line had their own dedicated EISA config utils. Also, in theory, the "Premium SE/TE" EISA config util should not be compatible with a regular "Premmia" because "Premium SE/TE" used an ASIC chipset designed by AST, while the "Premmia MTE/SE" EISA config util might work, since it may also be using the TI TACT84500 chipset (can't confirm). Try EISACU 2.84 on this webpage, it's meant for "Premmia MTE/SE", EISA config files are also there. You can also check these EISACU and CFG files uploaded by Horun.

By the way, congratulations on your new acquisition of old hardware 😉 The system seems to be nearly maxed out in terms of upgradeability, it would be interesting to see how well it performs.

Thanks for the reply.

I have put the project on hold for now.

Managed to clean up the acid damage and found a triple AA battery pack that serves as the CMOS battery. I also found an eisa config utility as well as a couple of disks with cfg files that boots and seem to work.

My trouble is mainly centered on the PM2012B/90 SCSI controller and the original Quantum 1GB SCSI disk. After the card is configured the computer no longer boots from floppy, let alone the SCSI disk. It just pauses after the SCSI ROM display has listed the disk (seemingly with proper naming and config). Keyboard works so it’s not fully frozen. Tried basic fault finding like swapping RAM sticks, changing EISA slots, and reseating stuff, but no joy. The disk must have been working when the computer was put on storage, but despite this it wasn’t configured to terminate. I found the jumper and fixed it, but the symptoms didn’t change.

Some times the bios will pause after counting the RAM, but gives two beeps and continues if I press F1. It doesn’t prompt for F1 though. Just a trick I found during googling. It may then try to initialize disks, of which there are none (IDE) before the SCSI card starts up.

Since I don't have similar equipment laying around it’s extremely time consuming trying to isolate ehat can possibly be multiple faults or config errors, or both, in between intermittent instability caused by age, acid damage and whatnot.

The BIOS is extremely simple with just basic stuff like date, keyboard, ram test and floppy settings.

A few questions though:

My EISA config “worked”, but could it be that other EISA utilities and newer config files would configure things “better”, solving the issue?

Are there any keyboard shortcuts to access settings during the SCSI ROM post?

Should the floppy be connected to the SCSI card after it’s configured?

Didn’t try using only one stick of RAM, and not sure if the Motherboard works with only one.

Both with the disk connected and not, the situation reminds me of when bad IDE/SATA disks just makes the computer act funny/unstable, even when everything seems to work fine.

If you suspect that the effects of the battery corrosion may still be affecting the operation of the EISA bus, I would advise you to pull the mainboard out and check the pins on the larger QFP chips with a needle, they may be loose. If everything checks out, then there might be some resource conflict due to bad EISA config. According to this page, your SCSI controller does not support FDD, so it should not affect onboard FDD controller. You might want to check this archive, it contains a .CFG file, some utils and an installation instruction for your SCSI adapter.

Hi
I would also like to update my ast premmia 4/66d. I have created the diskette with the flash.ast file. what is the next step? where can I find these bios update utilities?

many thanks

ah found 😀

i have the bios v1.0 as a bin file. can i convert to ast to possibly flash back ?

tt0ny wrote on 2023-11-30, 12:27:

ah found 😀

i have the bios v1.0 as a bin file. can i convert to ast to possibly flash back ?

Hi,

it should be possible, but you'd have to convert it manually using a hex editor. That *.AST file has the main BIOS part inside (120KB) and it can be manually extracted and in theory - put back in.