tony359 wrote on 2024-11-01, 09:53:

What would be the difference with the file you sent me? I don't want to sound harsh, it's a genuine question 😀

I took my file from the 86Box's romset and I am not sure how did it get there - the only thing I know about it that it boots on 86Box's emulated i430LX chipset with 82378ZB SuperIO chip.
If you do the splicing with your update files then there is a better chance that the result would match your board.

I did it myself, same result unfortunately.

I think I read somewhere that those files are "update" and don't contain the whole BIOS?
I see the BIOS I have seem to begin with BI0 - I tried swapping them around but with swapped results.

I've scoped all the pins of the BIOS. They're all active besides RP which is always high which means the system is not in deep sleep mode. I cannot see where CE goes on the board but it's got activity on the chip.

I've checked all the legs of the chipset, they're all properly soldered. I cannot believe this board has an issue. With the "right" BIOS it operates the floppy. that requires quite a few things working on the board! Also, someone on the other forum had exactly the same behaviour so...

Also in my opinion, with the correct BIOS the card will work, or at least there should be good chances, in any case using the correct BIOS, if it doesn't work, you can try another HW, for example CPU or RAM, just to be sure that the problem is not due to the MB, however to be sure that the BIOS is ok, you would have to duplicate that of an identical working MB, in this discussion (here on Vogons):

Batman's Revenge socket 4 motherboard - need manual

There are other users who should have it, maybe there is a copy or a link, maybe you could ask one of them (?)

I found this thread, the file linked is a Bios Recovery floppy creator. It creates the whole floppy this time.

bricked batman's revenge board, possibly due to SCSI bios?

I tested it but... no luck, same thing. I think my BIOS is broken beyond recovery. OR, the floppy controller has an issue as from a video I found online the floppy should seek normally during the recovery process. Mine "purrs" like a cat. The head keeps seeking back and forth. Either the BIOS routine is broken or the floppy controller is broken.

I'll see if someone can extract a copy for me. Thanks for the link.

tony359 wrote on 2024-11-01, 14:30:

I'll see if someone can extract a copy for me.

Here (attached) is an oldbios.bin file from the batrevrd.exe from the link above.
And here is the whole floppy image created by that self-extracting archive: https://kmeaw.com/batman.zip

I found that file - the recovery didn't work, the floppy does a "purring" noise (the head goes back and forth) but does not read.

I didn't think of using the BIN file manually! 😀 Thanks

I'm trying now.
in Normal mode I get 04-05 then, after a minute, some repeated long beeping. Actually, it beeps 6 times, very slowly with many seconds in between. Then after 30 seconds it starts again.
In recovery mode I get 05-06-07-08-09-0A-0b-0d (doesnt seem to stop), veeeery slowly. No beeps unless it does that after a lot of time.

Uhm.....

Here is the information on the POST codes I got from the oldbios.bin file. I might be wrong ,because I never read that much of AMI code:
04: read keyboard controller status, check "system" flag. If the flag is not set, go to "05", otherwise check if reset flag is Ctrl-Alt-Del (0x1234) - if it is, update an internal variable.
05: read northbridge revision in two different ways, ask it to disable L1 cache updates, ask the CPU to flush it and disable caching, ask the chipset to disable burst SRAMs and L2 cache, then some check over the revision number that I don't understand involving another cache flush, ask the CPU to disable shutdown on parity error; enable southbridge PCI address decoding, disable memory reads and writes between 512~1024KB, disable ISA-to-PCI forwarding between 768~896KB, disable MEMCS# generation, disable shadow memory (512~960KB), reset bit 5 at CMOS[0x10], another read of NB revision(?), disable IDE controller, set bit 5 at CMOS[0x10], hold floppy disk controller at reset and disable motor, enable ISA port access on IDE controller, load IDE registers from a table.

Thanks.

Meanwhile I found that OSC on the ISA connector is only 1.9MHz and this is coming directly from the clock IC which is - unfortunately - a factory programmable one.
I'd like to check what the floppy controller is getting when it comes to clock, that "purring" (the head moving back and forth slowly) might be a clue. The slow beeping could also be a clue that some clocks might not be right, I don't know.

The thing is that with some BIOS I can get the board to get to the recovery process pretty quickly - you see codes on the post card, it beeps ok then it starts moving the floppy in that weird way.

But another owner confirmed me that I should have 14MHz on OSC - and out of pin 9 of the clock IC. I don't have that. So that is a clue.

Thanks for looking into the codes for me, appreciated. I see IDE and Floppy are involved there, maybe that is also a clue? It's all so weird, I cannot make head and tail of that.

I don't know if you made progress with the motherboard with socket 4, for my part, I investigated the topic, and there are some things that can create problems, first of all the DALLAS chip, it must be working, if it is defective it can prevent starting, at least that's what I read on the pages I visited.

For the BIOS discussion, it seems that you have to use a bootable floppy, but the biggest problem is that not seeing the main screen, you can't see which BIOS was there (written in the chip), moreover the chip should be soldered, maybe 🤔 disconnecting it is not a big problem, but it looked like a card with BIOS easily available, instead from what I keep reading, it seems a bit complicated, in short I don'm not understanding anything anymore!

In addition, there is also the doubt about which card it really is, to tell you, mine is a Mercury, but maybe it is a Hendrix, or are they the same thing? 😨

Anyway if you want I can try to copy the BIOS of my MB, but it probably won't work, according to what I've read.

Hello Tony! I have some experience on the recovery function of old intel mainboards. I once fixed an intel endeavour socket 7 board using that method. From memory it works using the following method.
Create a bootable DOS disk first. I used windows 98SE for that, using the floppy formatting tool, copying the dos files on it.
Than copy the contents of the intel bootdisk that comes with the bios, including both bios files. (the contents of the \BIOS directory) Overwrite every file that pops up for being already on the disk.
After that set the board in recovery mode using the jumper. Make sure the floppy drive is in full working order before attempting to blind boot from it.
If the files are read correctly, a beep should sound. (it's loading the file beep.com) that is your confirmation of the process taking place. After that a lot of reading of the floppy disk is taking place. When the flashing has taken place, all activity stops. I am not sure if there is a beep from the pc speaker after finishing.

I also had to try a few times before I had succes.. It's difficult because there is no video output during recovery..

I thought I had replied to this! 😀

Thanks - the floppy IC is getting 3MHz instead of 24MHz so I need to sort out that first!

Is there anybody who could scope the clock outputs of the clock IC (U6G1, a CY2291SC on my board but I understand it's usually a different one)? It's a programmable IC so the datasheet won't help me unfortunately. I only have a good CPU output, everything else seems to be wrong but it's difficult to guess which is which.

I have sourced a ICS9159 which might be able to provide the necessary clocks, though not pin compatible so I'll have to do a spaghetti junction there 😀

Thanks!