EDO RAM on ErgoPro s450

[KachiWachi]

Recommended by Wim's BIOS...

[KachiWachi]

From the Wim's AWARD page -

Opti Viper-M (Pentium Motherboard)
82C556M
82C557M
82C558M

Also M519/529, Amptron PM7400. From the manual, these can use EDO.

Opti Viper (?)
82C556
82C557
82C558

From - http://users.erols.com/chare/chipsets.htm -

Viper is Parity, can use FPM, EDO, and SDRAM. It mentions N+ (556M/557M), but not an 558M...also Parity.

Hey, thanks guys! Very promising pieces of information!!

The OPTi Viper chips on this s450 do not have M appended in their codes. So they must belong to older generation.

But if they really support EDO, then a BIOS change might well resolve my problem. Worth trying!

Now I'm not quite aware how s450 actually boots. I have understood that it has a traditional ROM device (in the '95 handbook it is called the Remote Boot ROM) and a Flash device. I am wondering if the ROM device actually contained a known, working, factory installed BIOS which could be used to overwrite the Flash BIOS in case of emergency? Or is it used for some other purpose?

Anyway, only the (Remote Boot) ROM is in a socket, the Flash device is not. So if that ROM device is used for something else than a backup for the Flash, then hotflashing does not work for me.

Any other ways to protect against irrecoverable crash due to incompatible BIOS in ErgoPro s450?

[edwin]

Remote Boot ROM

for booting over the network I think.

'95 handbook says that when 'Network' is selected as the Boot medium, an expansion board with an ICL BIOS will handle the boot procedure. I doubt if there is any documentation on how this kind of boot can be implemented!? At least it could be a way to get a proper BIOS file on the s450 hard disk. But is there a way to flash the BIOS in runtime?

I have thought of using the RSU (Resident Setup Utility) menu 'BIOS reload' to do the trick. But again, if the new BIOS fails so that RSU cannot start, then I cannot patch the BIOS!? I guess it depends on the BIOS SW structure. If that RSU part is separate and not changed if the BIOS is rewritten, then I am safe!?

[KachiWachi]

I have an old Intel EtherExpress PRO-10+ NIC with a BootROM on it. When I enable that feature on the NIC, the PC will start, but then switches over to the NIC ROM and executes from that. The manual says this is used for "dumb terminals" where the actual boot runs from the server itself. I don't recall if the machine completes POST or not first though (I think it does ).

Your (Remote Boot) ROM is on the motherboard itself??

Yes it is! Unlike the '95 handbook says. And additionally, as said, it sits on a socket, but the Flash not. I didn't really see any network boot era back then...

I checked that 'BIOS reload' function a bit closer. It comes in two forms: 'BIOS modules' and 'BIOS and Loader'. Can I assume that the most critical and device dependent part of BIOS resides in code that is executed later than the initial start, POST and possible Setup? If so, then it should be safest to select 'BIOS modules'.

I have download the s450/133n BIOS #53, "b53_e.zip". I guess it includes both 'the loader' and 'the modules'!? At least both the loader version and BIOS version were changed when I updated the s450/100 BIOS with a newer version earlier. Can the RSU function 'BIOS reload'-> 'BIOS modules' load only module part of the BIOS file, "B53_e.ldb" and thus, as I understand it, minimize the risk of invalidating the PC with a possibly non-compatible BIOS?

Ok, I managed to get the PC boot with jumpers set so that the boot block of the BIOS should be protected against inadvertent writes. So now the jumper block 1 has jumpers on positions 3, 5, 6 and 9. J9 should protect the boot block. I still cannot say why it didn't boot in the first place. Probably there was a bad contact in the cpu socket!?

I think there's now a chance to try out a BIOS version which is actually not for this motherboard (AC41644), but its successor AC41655, which has been documented to accept EDO DRAM. This experiment may well fail, but at least it has been tested...

Oh no! It wasn't too easy.

First I tried to make it with a bootable diskette. After noticing that I needed to override any mismatches the flashing program finds, I managed to start the flash operation. However, apparently since I have the boot block protection jumper on, the program ended with a flash erasure error! Now if I had removed the protection, both the BIOS and the boot block would have been overwritten, I believe. That's not what I wanted.

Then I turned to the RSU, selected the 'BIOS reload' function and under that, the 'BIOS modules' operation. After exiting and selecting configuration update, the screen turned black and something started to happen. I heard some beeps and after a while the PC rebooted. But, as expected either it had also chosen the compatible (actually the same) BIOS file or it was also unable to rewrite the flash. The net result anyway was that nothing had changed!

So an expert is needed again...

I took the risk and tried the BIOS #53 of ErgoPro s450/133n (motherboard AC41655) in an ErgoPro s450/100 (motherboard AC41644)! It didn't work

I was able to erase and reprogram the Flash. The first sign indicating that something was wrong was found on the Info page of the BIOS Setup program (RSU). It showed a CPU speed of 180MHz although a 133MHz Intel Pentium was installed.

Then I let the boot proceed and before the POST had been completed the machine froze. I waited a few minutes after the first error notification appeared: BIOS(61): Hard disk controller error, status = 20H and a little later: BIOS(23): Hard disk init failed. Still later on: Press <F1> to continue or <Ins> for setup mode. I continued with a boot diskette and soon I had a DOS prompt. It worked so I shut down the machine and replaced the FPM DRAM with EDO one. With EDO DRAM inside the machine didn't start up anymore. It didn't even enable the display. The display showed that there was no incoming video signal! So the symptom was the same as earlier with the original and updated correct BIOS version with EDO sticks inside.

This must mean that there is no way to enable the use of EDO DRAM in s450/100

Luckily I was able to reprogram the Flash with the #50 BIOS file. So the pc is still in order!

Thanks to all of you who helped me to learn and understand the problem, especially to you Nick!

-vp-

[KachiWachi]

I knew this chipset looked familiar...

A friend at work found a Compaq Presario 7222ES in the trash recently (yes it works just fine).

This board has the OPTi Viper 82C566/7/8M chipset in it. The installed memory is 2 8MB SIMM EDO/7NS. The board is stamped "EDO ONLY".

I can't tell if it is parity or not, since the SIMMS are not easily removable at the moment. If I get some free time I will pull one and see which type it is.

It has an SMC Super I/O (FDC37C931).

[KachiWachi]

OK...installed RAM is made by Micron.

Module : MT16D232M-7 8MB SIMM EDO/7NS
Components : 16 - MT4C4007JDJ-6 (1Mx4 EDO, 1K refresh)