Broken IC on old Pentium 2 motherboard

[Per Hansson]

I have already linked you a BIOS for your board in post #88.

This is based on the BIOS string you have posted in your images.
Rasz showed you in post #70 that the manufacturer is: FYI (Full Yes Industrial) based on your BIOS string.

[james_s]

I use a TL866 now in place of the old Willem I had, I'm pretty sure it will handle most bios chips.

Looks like someone already found you the bios, lacking that I would experiment with others from similar boards and see if any of them work.

[spilihps]

I have already linked you a BIOS for your board in post #88.

This is based on the BIOS string you have posted in your images.
Rasz showed you in post #70 that the manufacturer is: FYI (Full Yes Industrial) based on your BIOS string.

Oh sorry, i forgot that. The thread is getting quite long!
Thank you very much for your input.

I use a TL866 now in place of the old Willem I had, I'm pretty sure it will handle most bios chips.

Looks like someone already found you the bios, lacking that I would experiment with others from similar boards and see if any of them work.

Yeah that is the one im looking to get!

[PKTKS]

(..)
Reflash through a external programmer then? If i could get hold of a image then that would be possible, but that is a long shot in this case i think.

Yes outside the board w/proper BIOS checking if the
chipset still holds the new data.

TL866 or RT809 both will do the trick
https://www.aliexpress.com/item/33000308958.html
https://www.aliexpress.com/item/33003747546.html

Paul

[PKTKS]

For the record...

If you ask me which one (today) I would rather prefer...

Just a few years ago you would be cursing that closed source
press button pile of garbage from last decades...

No longer the case ... MiniPro works absolutely fine
under WINE with just the minor glitches of MS poorly
inter-operational toolkit UI (just UI fonts.)

Otherwise.. go plain console from SOURCE which
in any case works perfect

https://gitlab.com/DavidGriffith/minipro/

Paul

[spilihps]

(..)
Reflash through a external programmer then? If i could get hold of a image then that would be possible, but that is a long shot in this case i think.

Yes outside the board w/proper BIOS checking if the
chipset still holds the new data.

TL866 or RT809 both will do the trick
https://www.aliexpress.com/item/33000308958.html

For the record...

If you ask me which one (today) I would rather prefer...

Just a few years ago you would be cursing that closed source
press button pile of garbage from last decades...

No longer the case ... MiniPro works absolutely fine
under WINE with just the minor glitches of MS poorly
inter-operational toolkit UI (just UI fonts.)

Otherwise.. go plain console from SOURCE which
in any case works perfect

https://gitlab.com/DavidGriffith/minipro/

Paul

https://www.aliexpress.com/item/33003747546.html

Paul

Thank for your help and input!

Its also great to see how far Wine have come.. Remembering using it way back in 2005-2006 maybe. It could be a hassle sometimes

The TL866 unit seems great, there are many that give it good feedback.

[james_s]

Wine can still be a hassle, but for a lot of things it really works pretty well. For years I tried Linux every couple of years, marveled at how much it had improved but then soon hit some kind of show stopper and set it aside, then finally a few years ago I found it had reached a point where aided by the regression in polish and quality that Windows suffered, it was finally looking really good.

I do wish the MiniPro GUI software was open source in the first place though, then at least we could fix the chinglish. The software doesn't live up to the quality of the device and drags down the overall experience but it's still quite good and a bargain for the price.

[PKTKS]

Wine can still be a hassle, but for a lot of things it really works pretty well. For years I tried Linux every couple of years, marveled at how much it had improved but then soon hit some kind of show stopper and set it aside, then finally a few years ago I found it had reached a point where aided by the regression in polish and quality that Windows suffered, it was finally looking really good.

I do wish the MiniPro GUI software was open source in the first place though, then at least we could fix the chinglish. The software doesn't live up to the quality of the device and drags down the overall experience but it's still quite good and a bargain for the price.

Pretty much WINE "just works" for reasonable written applets.

For those ones bundled with underlying "licenses" and "rights"
your mileage will surely vary.

For TL866 this folk made a decent job (ditching the original MiniPro for good)
by using a nice QT interface to handle the device firmware itself

Check https://github.com/radiomanV/TL866

[spilihps]

So i have pulled the old flash chip,
28F1000PPC-12C4, i was thinking that it might be a good idea to have a replacement chip, since this died at the same time as the other components.

Is there compatible replacements available from a good source like Mouser?

In my other motherboard thread (Seems to be in a mobo rescue streak right now ), i needed the SST39F020, that has more memory capacity than this one. (They do have the 010 variant also)

Would it by any chance be compatible?

Its not so easy to compare these memory chips

[james_s]

It ought to be easy, the datasheet is readily available.

https://www.alldatasheet.com/view.jsp?Searchword=28F1000PPC&sField=1

Looks at a glance like a 128k x8 flash  eeprom. I'm on my phone at the moment so I'm not going to try to study the datasheet but it ought to tell you everything you need. A larger part may work but in that case you'd probably want to concatenate multiple copies of the firmware to fill it.

[spilihps]

It ought to be easy, the datasheet is readily available.

https://www.alldatasheet.com/view.jsp?Searchword=28F1000PPC&sField=1

Looks at a glance like a 128k x8 flash  eeprom. I'm on my phone at the moment so I'm not going to try to study the datasheet but it ought to tell you everything you need. A larger part may work but in that case you'd probably want to concatenate multiple copies of the firmware to fill it.

Yeah the 010 variant is the same size at least, and pin compatible.. but I don't know if there are other pit falls?

[Bud]

Yeah, i actually had a Willem programmer for some Z80 experiements. But it was very unreliable so i discarded it

Two mods make Willem perform like a tank:
-replace the parallel port input buffers to Schmitt input buffers
-increase programming power supply current

mine was ok when used with the desktop but flaky when used with a laptop. It also did not burn properly big ROMs. It is now flwaless from the laptop through a 15 feet parallel cable and burns everything i throw at it from 27c16 to 27c1000 (the biggest UV eraseable  ROM i have) and just couple days ago programmed a 512kB flash.

[james_s]

With the wrong chip you might have trouble trying to update the bios through the motherboard but as long as the pinout and voltage are the same it should work.

[spilihps]

With the wrong chip you might have trouble trying to update the bios through the motherboard but as long as the pinout and voltage are the same it should work.

Thanks James for your input.
I can live without not being able to program from the board. Rather that then gamble on all the fake chips on ebay..

[Rasz]

eprom chips from ebay wont be fake, at worst you will get pulls from ewaste, nothing wrong with those

[james_s]

That may be true of windowed eproms but the sort you'd find in a square 32 pin package used in motherboards of this era are forms of eeprom and will not have a quartz window, I've been burned by fake eeproms multiple times that were either mask or OTP ROMs with data already on them or something else entirely. Any part can be fake, the quartz window of traditional eproms just makes it a lot easier to tell what you're actually getting.

[spilihps]

Also, the ones on eBay are actually quite expensive compared to the new ones from mouser..
So I figured if there is a chance for them to work as a replacement I will give them a shot..

eBay is around 10€ each and mouser is about 1,2€ each

[spilihps]

So at last an update to this never ending story!

I received my new EEPROM actually found the same part 28F1000, and burned the new firmware on to one of them, bought three just in case.

But a new problem arise! Of course!

Now we only get to code 31 (Wich is check RAM) The interesting thing here, is that it does not begin counting the ram, neither does it even show the text line, it just sits there and blinks the cursor on the line after "PENTIUM II CPU at 350MHz"

The only "Life" after that is that after maybe 2 seconds the "IRDY" LED on the post-card goes out.

Now, i tried to dump the old ROM and burn that one. And then we get the same exact behavior as we did with the old rom (Just did this to check the programmer) it counts the ram and such.
So now im not sure if the .BIN file from that site is botched, or we got something else wrong with the card? Cannot find any other bios file though.


******************UPDATE 2:

So guys, you wont believe this.

A while back i bought a new P2 350mhz (SL2S6) to try the board out with. When i plugged that in, nothing happened, so i thought it was maybe a dud, since the seller was kind of sketchy.

And just for the fun of it, i plugged it in with the new ROM, guess what, the board boots in to Win95, but the cpu only register as 300mhz..

I cannot believe whats going on here?

Maybe wrong case on the CPU, and that i have both a dead CPU and original ROM?


Also, i get "Unknown Flash type" before "Starting Windows 95".

****************UPDATE 3:

So yes, the case was replaced on the CPU, after lifting it i found the SL2QV chip, seem to be only for the 300mhz.. GEES.

So i guess i had a _partly_ broken CPU and _partly_ broken EEPROM?

[Rasz]

Yay! 
cpu was suspect from the start due to fried regulator

[james_s]

That wouldn't surprise me at all. If the regulator failed such that you got excessive voltage on the CPU it could have easily damaged all sorts of stuff. Sounds like you lucked out, thanks for the followup and success story, you ought to really know your way around early 2000's PC motherboards at this point, for whatever that's worth.

[Per Hansson]

Great perseverance!
Strange that the board did not POST at all with the new eBay CPU with the original ROM.
But that it works fine now with the new ROM.
Well, I guess it could have been unsupported with the original ROM?

As for the CPU speed you set that with jumpers on the board.
If you don't have the manual I found the attached quick reference.
The number you posted is not a sSPEC but a part number for the CPU's TagRAM.
I would be surprised if someone made a counterfeit P2 CPU in 2020, let alone to cheat 50Mhz

[Rasz]

Great perseverance!
Strange that the board did not POST at all with the new eBay CPU with the original ROM.
But that it works fine now with the new ROM.
Well, I guess it could have been unsupported with the original ROM?

I bet that was coincidence, it should totally work

As for the CPU speed you set that with jumpers on the board.

locked multipliers on that cpu gen, fsb auto detected with B21 pin

The number you posted is not a sSPEC but a part number for the CPU's TagRAM.

>TagRAM chip mounted on the back of Pentium II Slot 1 processor (300 MHz)

I would be surprised if someone made a counterfeit P2 CPU in 2020, let alone to cheat 50Mhz

more like someone dug it out of garbage in china and slapped whatever cooling was around

[Per Hansson]

Okay cool, yea it was many years ago that the P2 was the king of the hill so maybe my memory does not serve me

[spilihps]

Yay! 
cpu was suspect from the start due to fried regulator

Yeah it was! I did however have hopes for it since it counted all the ram, and worked fine scrolling through the bios settings. I in my naive mind thought that it either works or dont when it comes to a CPU

So this resulted in two casualties CPU and the GPU. Wonder if they are repairable? 

All due to a temporary overvoltage from the PSU i guess then.

Thank you very much for your help and input, and keeping up with me during all the confusion haha!

That wouldn't surprise me at all. If the regulator failed such that you got excessive voltage on the CPU it could have easily damaged all sorts of stuff. Sounds like you lucked out, thanks for the followup and success story, you ought to really know your way around early 2000's PC motherboards at this point, for whatever that's worth.

Yeah, i think everything started out with a temporary overvoltage from the PSU.
Also, thank you for your help and input!

At least i know my way around this particular board now. Probably not worth much when it comes to knowledge haha!

Great perseverance!
Strange that the board did not POST at all with the new eBay CPU with the original ROM.
But that it works fine now with the new ROM.
Well, I guess it could have been unsupported with the original ROM?

As for the CPU speed you set that with jumpers on the board.
If you don't have the manual I found the attached quick reference.
The number you posted is not a sSPEC but a part number for the CPU's TagRAM.
I would be surprised if someone made a counterfeit P2 CPU in 2020, let alone to cheat 50Mhz

Thanks Per for all your help with this also, everyone have been great. I did so much testing and tinkering i lost my self at times 

And also the other CPU, i have no idea. And i still dont understand why the original CPU counts the ram OK with original roms, but cannot start the boot process. And that it dont even get to the memory count with the new ROM file.

[james_s]

And also the other CPU, i have no idea. And i still dont understand why the original CPU counts the ram OK with original roms, but cannot start the boot process. And that it dont even get to the memory count with the new ROM file.

It all depends on just what is wrong with it, if you really wanted to go down that rabbit hole then I would suggest looking at each of the address and data lines with a scope or logic analyzer. If you have for example one address line that is stuck then it will cause some addresses to appear at the right place while others could be accessing something completely wrong. The difference between the BIOS ROMs could be as simple as some instruction pointing to memory in a slightly different location. I suspect something like this is what you have happening, one damaged address pin which is not involved in reading the code used by the memory test.