I've removed the BIOS chip. On the underside of the chip, I can see this:
SSTAL311
602022
3K28P134
Also, in the empty BIOS chip socket, I can now see another smaller chip! It isn't socketed. On it, I can see this:
Funny N shape then P9512SL
74AC244PC
Those numbers off the bottom of the chip aren't very informative, you really need the number off the top of the chip.
Depending on what type of flash chip it is, you may be able to use the boot ROM socket on a PCI NIC to reflash the BIOS, though you may have to do a modification to the #WE pin to disable write-protect (depends on the NIC). I think the usual limitation is that you can only use up to 1Mbit chips, but a 486 BIOS chip would probably be either 512Kbit or 1Mbit anyway.
I tried to reproduce you attempts and flashed the beta bios from the FIC FTP-site (using aflash3.exe). Flashing was succesful but after resetting "BIOS ROM checksum error" was reported. Because the Bootblock was still functional, I could boot from diskette. Reflashing the non-beta bios with aflash3 worked. I don't know what's wrong with these Beta bios files...
So, if you can't boot from diskette anymore, hotflashing is the only way to go.
Why not 'do-it-yourself' ?
OK people, I have scrapped off the sticker on the SST BIOS chip, and it reads:
PH29EE010
150-4CF
960502-E
I'm open to all suggestions as to the best way forward! 😀 There is a service in the UK costing £10.99 (about 17.50 US $ ). You get a new chip, and the file used is the one you supply. I am also happy to get and try a PCI NIC. Whichever option you guys think is the best.
Edit: I was very careful in removing the BIOS chip, but at the very last second, the chip just sprang out of the socket, and lots of pins got bent. I pushed them back OK, but some of them are probably weak now.
I did a search on ebay for flash bios, and found some crazy looking contraption called Willem. (Search for willem flash) I read one of the listings, and it mentioned support for PH29EE010, which is the writing found on my BIOS chip.
However, I'm worried that some of the BIOS chip's legs will break if I mess about with it too much. Perhaps it's easier just to get a new chip?
Edit: although it might be fun to get a "flash contraption" to flash a copy from the old chip, in to a new chip. Yushatak, do you have a similar device, like the Willem unit?
I have a Willem programmer. It's fantastic on paper but I had a bag of 10 chips to attempt to flash one BIOS onto one chip with, and it failed on all of them. Bad programmer? Bad software? Bad chips? Impossible to say. -_-
It *requires* Win9x to function - you can use a VM with proper parallel passthrough (the ironically named Parallels is the only one I'm aware of that works for things like this), but I had no luck. It has a DOS program too, but it has a reduced selection of chips that it supports through that software.
I tried both a VM and a physical machine running 98SE, and neither would flash my chips. No idea what the cause was, however.
Yushatak, I'm sorry to hear this. 🙁 I've just had a thought - it's about hot flashing. Is this possible: boot up a different 486 PC, eg an Asus 486. While this system is still switched on, carefully remove the BIOS chip. Then insert the BIOS chip from the FIC mobo, in to the now empty BIOS socket on the Asus mobo. Then run the award flash utility, and flash this BIOS chip, for use in the FIC mobo. Switch off power to the Asus mobo. Remove the BIOS chip and put it in to the FIC mobo, and switch on power to the FIC mobo.
Is this "hot flashing"? I assumed you needed the exact same type of board, but perhaps you can get away with using a different 486 mobo, for the hot flashing operation.
Rather than search for flash bios on ebay, I searched for the code on top of the SST chip, which is PH29EE010. I found some cheap Willem flash units. Here's one. I like the girl in this advert. Say I buy this thing, do you know where I can get some "blank" SST chips? Thanks a lot. 😀
wrote:Yushatak, I'm sorry to hear this. 🙁 I've just had a thought - it's about hot flashing. Is this possible: boot up a different 486 PC, eg an Asus 486. While this system is still switched on, carefully remove the BIOS chip. Then insert the BIOS chip from the FIC mobo, in to the now empty BIOS socket on the Asus mobo. Then run the award flash utility, and flash this BIOS chip, for use in the FIC mobo. Switch off power to the Asus mobo. Remove the BIOS chip and put it in to the FIC mobo, and switch on power to the FIC mobo.
Is this "hot flashing"? I assumed you needed the exact same type of board, but perhaps you can get away with using a different 486 mobo, for the hot flashing operation.
Yeah, this is hotflashing 😉
Neither you need the same type of board, nor the same type of flash ROM.
For 1 Mbit chips I usually use a Pentium mobo Asus P55T2P4. In uniflash there is a "Redetect Flash Type" function after swapping the ROM chip.
Regarding finding "blank" SST chips. Would these be OK? The advert says:
SST Flash Memory SST29EE010-150
My chip says PH29EE010. If I remove the "SST" letters from the code above, and also remove the "PH" letters from the code lettering seen on my chip, then this bit: 29EE010 is the same. Is that good enough? Thanks a lot for any advice. 😀
Thanks a lot for the info! BTW, regarding the FIC 486-PIO3 mobo - does this use a 1 MBit chip? Perhaps I should look to get a nice socket 7 classic old Pentium mobo, and it could be my "hot flash" mobo?!
Success! I just did my first hot flash! RG100 destroyer of old hardware? I think not! I did the following:
1) Found an old mobo. It just happened to be an Asus 486, with the same type of BIOS chip.
2) I loosened up the BIOS a lot, with a thin knife. This would then enable me to very easily remove the chip, while the machine is switched on.
3) I powered on the Asus mobo.
4) I ran Uniflash, to make sure everything is OK. Then, I pulled off the BIOS chip. It came away very easily, because of step (2)
5) I inserted the "bad" BIOS, which needed to be flashed. I pushed it down without too much force.
6) Inside Uniflash, I chose the Redetect BIOS option. I then flashed the chip with the 486PIO3.AWD BIOS file, provided by BastlerMike.
7) I switched off power to the Asus mobo, and easily pulled off the newly flashed BIOS chip, because I did not use much force as described in step (5)
8 ) I put the newly flashed chip in to the FIC mobo, and it works perfectly.
I'm not so much of an idiot anymore. Tomorrow, I'm going to apply for a job as a brain surgeon!
I have solved the issue I was having earlier, with the RNI 8-pin "jumper block". I found the manual for this FIC 486-PIO3 mobo. It's here. In it, there are these 2 diagrams:
Page 5
Page 6
For an AMD X5 (P75) CPU, the RNI 8-pin "jumper block" needs to be in pin holes 1-8, with pin holes 9 & 10 being empty. It was a good job I moved it, because it is now in its correct location. For a Cyrix CPU, the RNI 8-pin "jumper block" needs to be in pin holes 3-10, which is where it was originally located.
I've managed to get Quake timedemo working on this disappointingly slow VIA chipset-based mobo, at 200 MHz. I tried 1 stick of double-sided 60NS FPM. I also used the integrated IDE controller, rather than an ISA IO controller. Curiously, I set all of the BIOS timings to "max", and it worked OK.
There is no PCI divider on this mobo. There's a telltale sign that no automatic divider is being activated when the mobo's bus speed is set to 50, because I used my PCI Virge card, and it displays "trippy colours" on the screen when I run the PCI bus at 50 MHz, when Doom and Quake are running. I know this happens, because if I use a mobo which does have a PCI divider, and I set this BIOS value to "1:1", this strange graphical phenomenon occurs. I can bypass this weirdness if I use a Diamond Viper, and I'm sure there are other cards that will work OK too.
For the fullscreen Quake timedemo, I get 14.3 FPS, which isn't that great. I notice on the mobo's PCB that it says 1.1. Perhaps the 1.0 versions were faster? The bottleneck on this board appears to be how it uses the memory. I get poor "throughput", whatever that means exactly.
GR100, do you happen to know if the SiS496/497 supports the PCI divider? Or is it a UMC-only thing?
I'm sorry, I don't know if the SiS496/497 chipset supports the PCI divider. IMHO, PCI @ 50 MHz is doable. Also IMHO, I think the non UMC boards are failing some key tests @ 200 MHz for other reasons, and not the lack of a PCI divider.
wrote:I'm sorry, I don't know if the SiS496/497 chipset supports the PCI divider. IMHO, PCI @ 50 MHz is doable. Also IMHO, I think the non UMC boards are failing some key tests @ 200 MHz for other reasons, and not the lack of a PCI divider.
Hmm...it's not inconceivable that these alternative chipsets simply can't make it to 66Mhz.
