BIOS Savior -- How to write the Internal PROM ??

[GadgetBuilder]

I got the Bios Savior RD1-8X for my Chaintech VNF4 but can't figure out how to program its internal PROM. The VNF4 has a SST49LF004A/3.3V and the BIOS Savior has an i82802AB.

I tried Winflash v1.75 (and 1.51) by switching to the Bios Savior's internal flash after booting, then running Winflash to write this internal flash with the same V4 BIOS I wrote to the original flash some time ago using WinFlash.

Winflash gives: "BIOS ID error Are you sure to program" I said Yes. Then flashed it. It whirs and clanks for 3-4 minutes and then says "Flash Block verify error". It won't boot with the new flash but works when I flip the switch back to the original BIOS PROM.

So, I booted a floppy with AWDflash v8.32, flipped the switch to the Bios Savior's internal flash and used the following command:
Awdflash.exe VN210427.bin /cks9d20 /cc /cd /cp /r /WB /py

This fails with "Unknown Flash Type".
I tried adding the checksum with the idea that this might over-ride the ID error but it didn't work.

The problem seems to be that the BIOS has some type of ID built in and this is checked against the ID in the BIOS file. Since the blank BIOS chip I'm trying to program doesn't have the ID already it doesn't let me program it to have that ID -- kind of a Catch 22.

Any ideas how to get around this?

The BIOS savior seems like a good idea but it is defeated by the built-in software safeguards which don't have an obvious way to over-ride them... makes me wonder how hot-flashing is done since it seems like the same situation.

The VNF4's BIOS settings are at defaults: Flash BIOS Protection - disabled, System BIOS Cacheable - disabled.


Tried Uniflash; this produced a clearer error message: "Unknown Flash Chip". I wrote to the support email and quickly got an answer (nice people who clearly care about their program) but unfortunately:
"nForce 2, 3, 4(...) chipsets are not supported by UniFlash."

Any help or ideas would be appreciated.

[Qortado]

Have you tried awdflash /f /qi options ?

[GadgetBuilder]

Thanks for the response Qortado. Based on your suggestion I tried the /f and then the /f /qi switches appended to my original command line. The result was the same: "Unknown Flash Type"

----------------
More info and thoughts:

The VNF4 uses the SST 49LF004A/3.3V. The chip on the RD1-8X is an i82802AB. I believe both of these are FWH type and should be compatible, as shown in the IOSS compatability list.

The problem seems to be the flash programs used to update the PROM (WinFlash, AMDflash) will not write the RD1-8X's chip. WinFlash attempts to write it and seems to write the area from 0% to 60%, skips to 96% and writes to 100%. I believe this corresponds to the BIOS followed by the Boot Block. It then fails to verify, so the PROM was not written correctly.

I noted that starting WinFlash while the RD1-8X PROM is selected erroneously identifies the PROM as 49LF004A. I believe this is why the write fails -- there is probably some slight difference in the required write procedure between the chips. Just a guess based on watching what happens. I believe the original PROM and the RD1-8X PROM are compatible for reading but not quite compatible for writing which causes this problem.

AMDflash doesn't indicate what chip it detected, just "Unknown Flash Type". Possibly the chip ID info is destroyed (don't know whether this is possible with the Intel chip).

I believe what is needed is a flash program which directly reads the chip type from the PROM on startup rather than using information stored at bootup.

Winflash/AMDflash don't allow designating the chip type (as Uniflash does) to handle situations like this.

[Qortado]

OK, very strange indeed.

I suggest you download an Intel-based awdflash and see if it helps.

I see awdflash 8.57 supports "INTEL E82802AB /3.3V(4Mb)"

Btw, what awdflash versions have you tried so far ?

[GadgetBuilder]

The latest AWDflash I have found (and am using) is v8.32, found at:
http://support.octek.com.au/FAQ/faq_0002.htm

I poked around inside v8.32 with Notepad and found: "INTEL E82802AB /3.3V(4Mb)" so it looks like it should handle the chip. It looks like the chip ID may be gone, although I'm unsure if this is possible?

I wrote to IOSS, who make BIOS Savior, and they think I may need their later version, RD1-PMC4, because my mobo was produced after January 2004. This seems to conflict with their compatability list:
http://www.ioss.com.tw/web/English/RD1B ... Sheet.html

Where both the RD1-8X and the RD1-PMC4 are listed as compatible with the SST 49LF004A/3.3V

I bought the unit from mwave last week and didn't see the PMC4 listed so I purchased the 8X: http://www.mwave.com/mwave/DeepSearch.h ... ALL=y&TP=2

[Qortado]

So what are you going to do ?

Looks like you'll have to return this 8X and get a PMC4 right ?

[GadgetBuilder]

So what are you going to do ?

Looks like you'll have to return this 8X and get a PMC4 right ?

I think I've been had -- mwave doesn't list the RD1-PMC4 plus their return policy is such that it wouldn't pay to return it anyway.

So, I'm going to drop the idea of the BIOS Savior. There is no guarantee that the PMC4 would work any better than the 8X with my mobo and I'd feel like a real sucker if I ended up with two expensive doorstops

I wrote to IOSS again and asked that they correct their compatability list so others don't make the same mistake I did.

[Qortado]

I think Intel doesn't work since it's listed as FWH only, PMC Pm49FL004/3.3V in the other hand is listed as a LPC/FWH, would it make any difference ?

[GadgetBuilder]

According to IOSS's site, LPC and FWH don't matter, they are both supposed to be compatible with the 8X and PMC4; see the blurb at the top:
http://www.ioss.com.tw/web/English/RD1B ... Sheet.html

I think that LPC = Low Pin Count. Probably a marketing-hype way of indicating that some of the pins are multiplexed or something like that.

After all, FWH = FirmWare Hub. This apparently means that Intel incorporated a Random Number Generator (RNG) onto the flash memory chip.

It has an "Alice in Wonderland" quality, where words mean exactly what they want them to mean, not what they mean in common usage

[Qortado]

I think it's more of a motherboard BIOS compatibility problem with i82802AB rather than a problem with the FWH itself.

For the record, i82802AB is a FWH-only flash ROM just like SST 49LF004A while as PMC Pm49FL004 is a two-in-one flash ROM supporting both FWH/LPC interfaces, in my opinion that's why IOSS discontinued RD1-8X using i82802AB and chose PMC Pm49FL004 instead for compatibility concerns.

I see this rather old Intel FWH was/is usually found only on i800 series chipsets, as I came up with mainly Intel motherboards using this FWH, so I tend to think that i82802AB is less likely to be used in AMD motherboards and maybe that's why it doesn't work with your motherboard.

In my opinion, ask your motherboard vendor if they could add support for this FWH in a future BIOS, then I guess it would work just fine, till then it's not going to work. It would be nice with could test RD1-8X in other nF4 motherboards and see if works or not.

PMC Pm49FL004 might not work as well, but I've seen other nF4 motherboards using it.