Mamba wrote on 2024-11-06, 13:30:

It is an Intel N82802AB, probably, as the writing are not super clear. So it should be a 512Kb 64x8, but there are also voltage […]
Show full quote

It is an Intel N82802AB, probably, as the writing are not super clear.
So it should be a 512Kb 64x8, but there are also voltage parameters that I am not super certain of.
Thing is that I do not want to fry the chip and lose time, can you confirm that this one will be ok? (AT27LV512A)

https://www.mouser.it/ProductDetail/Microchip … v2TLXL8GA%3D%3D

Thanks

The AT27LV512A-90JU is a "one time programmable" or OTP and is not laid out in 64k blocks, it will not work properly as a replacement.

Horun wrote on 2024-11-06, 18:56:

Mamba wrote on 2024-11-06, 13:30:

It is an Intel N82802AB, probably, as the writing are not super clear. So it should be a 512Kb 64x8, but there are also voltage […]
Show full quote

It is an Intel N82802AB, probably, as the writing are not super clear.
So it should be a 512Kb 64x8, but there are also voltage parameters that I am not super certain of.
Thing is that I do not want to fry the chip and lose time, can you confirm that this one will be ok? (AT27LV512A)

https://www.mouser.it/ProductDetail/Microchip … v2TLXL8GA%3D%3D

Thanks

The AT27LV512A-90JU is a "one time programmable" or OTP and is not laid out in 64k blocks, it will not work properly as a replacement.

That is why I ask first….
Any cheap alternative similar to the original one?

I found nothing as a replacement.

As far as I can tell, the board's BIOS chip doesn't need any other requirement than being a standard 3v3 part. PMC's 49FL002T would be one.

Horun wrote on 2024-11-06, 20:05:

I found nothing as a replacement.

This troubles me…

Mamba wrote on 2024-11-06, 20:53:

Horun wrote on 2024-11-06, 20:05:

I found nothing as a replacement.

This troubles me…

It is a specialized serial eeprom that only Intel made (or had made afaik) and is 4mbit (512kx8). I see them on ebay and aliexpress.
datasheet: https://pdf.datasheetcatalog.com/datasheet_pd … nd_N82802AC.pdf
The Intel N82802AB8 appear to be an upgraded version but no good pinouts found for it.

Horun wrote on 2024-11-06, 20:58:

It is a specialized serial eeprom that only Intel made (or had made afaik) and is 4mbit (512kx8). I see them on ebay and aliexpr […]
Show full quote

Mamba wrote on 2024-11-06, 20:53:

Horun wrote on 2024-11-06, 20:05:

I found nothing as a replacement.

This troubles me…

It is a specialized serial eeprom that only Intel made (or had made afaik) and is 4mbit (512kx8). I see them on ebay and aliexpress.
datasheet: https://pdf.datasheetcatalog.com/datasheet_pd … nd_N82802AC.pdf
The Intel N82802AB8 appear to be an upgraded version but no good pinouts found for it.

I ordered this
PM49FL004T-33JCE

From what I understand should have the same spec. No?

Mamba wrote on 2024-11-06, 21:25:

I ordered this
PM49FL004T-33JCE

From what I understand should have the same spec. No?

That has same general pinout and says "Read compatible to Intel 82802 Firmware" so yes it could or should work ok once programmed.
Only thing missing is the random number generator but not sure what function it has for your board..
Nice find !!! It did not show up on a eeprom cross reference chart...

I ordered from China the Intel one as well.
But will take forever to reach me.

Horun wrote on 2024-11-06, 21:37:

That has same general pinout and says "Read compatible to Intel 82802 Firmware" so yes it could or should work ok once programme […]
Show full quote

Mamba wrote on 2024-11-06, 21:25:

I ordered this
PM49FL004T-33JCE

From what I understand should have the same spec. No?

That has same general pinout and says "Read compatible to Intel 82802 Firmware" so yes it could or should work ok once programmed.
Only thing missing is the random number generator but not sure what function it has for your board..
Nice find !!! It did not show up on a eeprom cross reference chart...

Ok here the update.
I flashed the compatible Eeprom and booted the motherboard.
Black screen at first but my post card was scrolling numbers so I said to myself that the motherboard was alive.
I tried different sticks of ram and gpu.
I figured out that with a gpu on the agp slot the board won’t post.
So I put a pci gpu and it boots.

I receive the message “checking nvram…. Update Failed”
But I can enter XP.
So I assumed that flashing with an external programmer is not enough.but using the flash program bundled with the bios I have the message “NO FLASH ROM PRESENT” ????

What can I do?

When I put an AGP card, there is no post and I obtain this code:

d3 14

All LEDs are on.

The Reset LED is on ? D3 is "start memory sizing" and 14 is "timer test" from the 1995-97 AMI code pdf.
Maybe something wrong with the AGP slot. How much ram is installed and what AGP card ?
Maybe a memory or resource conflict...

Horun wrote on 2024-11-17, 21:32:

The Reset LED is on ? D3 is "start memory sizing" and 14 is "timer test" from the 1995-97 AMI code pdf.
Maybe something wrong with the AGP slot. How much ram is installed and what AGP card ?
Maybe a memory or resource conflict...

There is no reset led on.
The system boots fine with a pci card.
I tried different gpus and I obtain the same message.
Still those messages make me think.
My feeling is that the external flash process (or using Uniflash also) does not flash the chip in the proper way.

I need to find an Intel Eeprom or a way to fool the modified amiflash program.

Open to suggestions

Check the board for a jumper. as I recall Supermicro still used a jumper to lock the bios and thus prevent anything from flashing a chip. It might also be in the bios settings, oddly enough. SM made good documentation, the manual should have detailed step-by-step instructions.

"Universal AGP" is woefully NOT VERY universal. Many AGP cards will fit and not work correctly. Also, when a board will boot from PCI but not AGP. The board is either physically defective. AND/OR the caps that feed the AGP slot need replaced. Once I confirm a working agp card, and its still horked, i start looking at caps. thats been my experence.

There are old posts on SM support group sites which point to this board having similar issues with various AGP cards (Voodoo Banshee, Matrox G400 & possibly TNT2 were noted as having worked ok). Like you, PCI cards worked OK.

Seems these SDRAM 840-based boards (DME, DM3 & DM6) weren't the most stable offerings from SM, and were quickly hooked in favour of their RDRAM-based successors (DRE & DR3) - apparently they also suffered SCSI corruption issues which required some component rework.

luckybob wrote on 2024-11-17, 22:59:

Check the board for a jumper. as I recall Supermicro still used a jumper to lock the bios and thus prevent anything from flashing a chip. It might also be in the bios settings, oddly enough. SM made good documentation, the manual should have detailed step-by-step instructions.

"Universal AGP" is woefully NOT VERY universal. Many AGP cards will fit and not work correctly. Also, when a board will boot from PCI but not AGP. The board is either physically defective. AND/OR the caps that feed the AGP slot need replaced. Once I confirm a working agp card, and its still horked, i start looking at caps. thats been my experence.

Thanks but I do not think that a jumper on the motherboard can prevent me to flash the bios on an external programmer.
Plus, I also used Uniflash and the process was successful.
Still I got the message checking nvram update failed.
Will look for this jumper anyway.

Edit: the manual talks about a JP6 to be disabled before bios recovery but I am not able to find it and the manual itself in jumper description jumps from 5 to 7…. No 6.

Tried the bios recovery method as well, it loads the bios and then I get 7 short beeps (repeating), that I guess is an error.

I am strongly convinced that the bios is not correctly loaded and this somehow prevents the AGP to properly work.

Mamba wrote on 2024-11-18, 05:31:

Thanks but I do not think that a jumper on the motherboard can prevent me to flash the bios on an external programmer. Plus, I a […]
Show full quote

luckybob wrote on 2024-11-17, 22:59:

Check the board for a jumper. as I recall Supermicro still used a jumper to lock the bios and thus prevent anything from flashing a chip. It might also be in the bios settings, oddly enough. SM made good documentation, the manual should have detailed step-by-step instructions.

"Universal AGP" is woefully NOT VERY universal. Many AGP cards will fit and not work correctly. Also, when a board will boot from PCI but not AGP. The board is either physically defective. AND/OR the caps that feed the AGP slot need replaced. Once I confirm a working agp card, and its still horked, i start looking at caps. thats been my experence.

Thanks but I do not think that a jumper on the motherboard can prevent me to flash the bios on an external programmer.
Plus, I also used Uniflash and the process was successful.
Still I got the message checking nvram update failed.
Will look for this jumper anyway.

Edit: the manual talks about a JP6 to be disabled before bios recovery but I am not able to find it and the manual itself in jumper description jumps from 5 to 7…. No 6.

Tried the bios recovery method as well, it loads the bios and then I get 7 short beeps (repeating), that I guess is an error.

I am strongly convinced that the bios is not correctly loaded and this somehow prevents the AGP to properly work.

According to an older manual version, JP6 & JPWAKE are the same (see pic & table)

"EMERGENCY BIOS RECOVERY PROCEDURE

1. Turn your system off and place the floppy disk with the save BIOS image file (see above procedure, step 5) in drive A.
2. On your keyboard press and hold "CTRL" and "Home" low-level BIOS flash hot keys. (You need to disable "keyboard wake-up" function via jumper JPWAKE.)
3. Turn on power with the hot keys depressed until your floppy drive is accessed. Your screen will remain blank until the BIOS program block is flashed.
4. If your system either reboots or displays a message to reboot system then the procedure was successful.
5. If the recovery procedure was not successful, please contact our RMA department for replacement."

PC Hoarder Patrol wrote on 2024-11-18, 07:24:

According to an older manual version, JP6 & JPWAKE are the same (see pic & table) […]
Show full quote

Mamba wrote on 2024-11-18, 05:31:

Thanks but I do not think that a jumper on the motherboard can prevent me to flash the bios on an external programmer. Plus, I a […]
Show full quote

luckybob wrote on 2024-11-17, 22:59:

Check the board for a jumper. as I recall Supermicro still used a jumper to lock the bios and thus prevent anything from flashing a chip. It might also be in the bios settings, oddly enough. SM made good documentation, the manual should have detailed step-by-step instructions.

"Universal AGP" is woefully NOT VERY universal. Many AGP cards will fit and not work correctly. Also, when a board will boot from PCI but not AGP. The board is either physically defective. AND/OR the caps that feed the AGP slot need replaced. Once I confirm a working agp card, and its still horked, i start looking at caps. thats been my experence.

Thanks but I do not think that a jumper on the motherboard can prevent me to flash the bios on an external programmer.
Plus, I also used Uniflash and the process was successful.
Still I got the message checking nvram update failed.
Will look for this jumper anyway.

Edit: the manual talks about a JP6 to be disabled before bios recovery but I am not able to find it and the manual itself in jumper description jumps from 5 to 7…. No 6.

Tried the bios recovery method as well, it loads the bios and then I get 7 short beeps (repeating), that I guess is an error.

I am strongly convinced that the bios is not correctly loaded and this somehow prevents the AGP to properly work.

According to an older manual version, JP6 & JPWAKE are the same (see pic & table)

"EMERGENCY BIOS RECOVERY PROCEDURE

1. Turn your system off and place the floppy disk with the save BIOS image file (see above procedure, step 5) in drive A.
2. On your keyboard press and hold "CTRL" and "Home" low-level BIOS flash hot keys. (You need to disable "keyboard wake-up" function via jumper JPWAKE.)
3. Turn on power with the hot keys depressed until your floppy drive is accessed. Your screen will remain blank until the BIOS program block is flashed.
4. If your system either reboots or displays a message to reboot system then the procedure was successful.
5. If the recovery procedure was not successful, please contact our RMA department for replacement."

Great,
Will try this

Edit: There is no jpwake either….
Edit-2: I locaterd the Jp6 jumper on the motherboard, it basically allows to let me start the bios recovery function, if I move it, the process won't start. So at the end of the day the problem remains because I hear 7 short beeps and code F4 (the AMIBOOT.ROM is not the correct size).

Certainly not their clearest bit of technical documentation!...

...JP6 = Keyboard Wakeup = JPWAKE