IBM Thinkpad 600E bios mod for processor update

[Sharedoc]

Sebi,

I suspct that the differences between PIII and PII caches are in hardware - not in setup. 600E bios does not initialize L2 because of it does not recognise CPUID. Later this causes an error situation in the bios (fires 127 error from bios exception or interrupt handler).

I'll take a look at CPUID cheating approach (4). Is stack already initialized at this phase when CPUID is studied in the code? Can I use CALL/RET? Maybe I could replace all CPUID checks by CALL into a CheatPIII - subroutine that skips CPUID and just inserts PII values in the registers.

[Sebi]

Sebi,

I suspct that the differences between PIII and PII caches are in hardware - not in setup. 600E bios does not initialize L2 because of it does not recognise CPUID. Later this causes an error situation in the bios (fires 127 error from bios exception or interrupt handler).

I'll take a look at CPUID cheating approach (4). Is stack already initialized at this phase when CPUID is studied in the code? Can I use CALL/RET? Maybe I could replace all CPUID checks by CALL into a CheatPIII - subroutine that skips CPUID and just inserts PII values in the registers.

Yes, but where is the cpuid checked? The routines I found only check the family, not the stepping or type. And familiy is the same.

The stack is NOT initialized at this point, furthermore, the system doesn´t exactly know if there is any (good) ram. You can´t use call/ret, but may circumvence this by loading the return address to sp, do a jmp to the subroutine and return from there via retn. See the bios reverse engineering article at http://www.geocities.com/mamanzip/Artic ... guide.html for details. When disassembling the bios you´ll find lots of these "indirect calls".

Sebi

[gags]

well i just bought this of the net and i think i bought the wrong cpu for my thinkpad 600-e model number 2645-8BG p2 400mhz

sSpec Number
sl4k2
Processor Frequency
750.00 MHz CPUID String
0686
Package Type
Micro-PGA2 Core Voltage
1.6V/1.35V
Bus Speed
100 MHz Thermal Guideline
24.6W
Core Stepping
PC0 Thermal Spec
100°C
L2 Cache Size
256 KB Manufacturing Technology
0.18 micron
L2 Cache Speed
750 MHz Bus/Core Ratio
7.5
Spec Update
http://developer.intel.com/design/mobil ... tation.htm
Product Order Codes

OEM Order Code
KP80526GY750256

can someone just confirm it was an impulse buy

will i be able to upgrade this in my thinkpad

thnaks

[ProtoColD]

gags, you bought the wrong CPU. The CPU you bought can be used in T21. Not the 600e/600x.

[gags]

oh well guess ill have to look again thanks for the info

[gags]

Hello.
Is there a guide on how to get into the 600E, specifically on getting to the MCC-2 module? Any noteworthy gotchas I should know about?

There´s a service manual by IBM. Go to IBM page and enter "maintenance manual 600" in the search field.

Sebi

hi,

could you post a link please? for the life of me i cannot find this guide

thanks

[Sebi]

hi,

could you post a link please? for the life of me i cannot find this guide

thanks

What? You can´t go to www.ibm.com, find the "search" field in the top right corner, and enter "maintenance manual 600e"?

I hope you can click on the right link on this page:

http://www-307.ibm.com/pc/support/site. ... WIK-3SYPX2

Sebi

[gags]

easy sebi !

thanks for the link. what came up was 100's of different things none of them being the manual.

I do know how to use my internet browser thankyou very much!

gags

[Sebi]

easy sebi !

thanks for the link. what came up was 100's of different things none of them being the manual.

Can´t believe it. Tried it again - the 2nd link says "Hardware Manitenance Manual index - ThinkPad general" and leads to a page with all available manuals. The 4th link reads "Hardware Maintenance Manual (June 1999) - ThinkPad 600". And even the 1st result has a link to the manual. That´s not 100´s of different thinks I mean...

Ok, EOT for me now

Sebi

[Sharedoc]

Sebi,

I have not yet collected enough courage to remove the 28F004 from 600X and solder to 600E. How did you remove he chip? I have been thinking to use a sharp knive and just push it below the chip and twist the pins separate from the pcb? I plan to do it without any heating. Do you think the pins will stay with the chip?

Also I have thought do the pins need twisting downwards for piggyback soldering? How did you prevent solder from jumping between the pins? Did you use alu-foil or similar?

I have a problem with my soldering iron, it has too long and big tip. Have to go and buy a new one.

How did you handle the jumper for EEPROM switch?

Is there enough room under the cooler assembly for the piggyback-28F004?

[Sebi]

Sharedoc,

I always use the same technique to unsolder chips: push a thin wire (0.5mm should be fine) under the pins along one side of the chip, and tighten it somewhere (mostly I solder it somewhere). Now pull the wire on the other end in a 40-60° angle (but as near as possible on the PCB, ofcourse) and start to unsolder the pins, gently pulling the wire a little until all pins are unsoldered. The chips can handle the little bending of the pins very well. You need to take extra care for the last 4-5 pins on the 2nd side (just fixating the chip by pressing it down should be enough).

I also bent the pins down for piggypack soldering. Just put the chip on top of the other, and bent the pins using a small screwdriver. I didn´t use any foil to solder the pins. I took almost no extra soldering, as the "used" pins should have enough. But you´ll need to put some flux on the pins. If solder jumps between two pins, "unsolder cord" (well, don´t know if that´s the right word for the german "Entlötlitze"?! know what I mean?) helps.

There´s enough room for the piggypack 28f004. For the switch I used an old switch laying arount here (with one switcher contact). I cut the /CE foil (pin 22, it´s bottom right corner, 2nd pin) and found a via to connect the middle pin of the switch. Then I connected both eeprom pins to the left / right contact of the switch. Don´t think a jumper is a good solution: remember you might have to boot with one eeprom, and switch to the other while system is running!

Sebi

[Sharedoc]

Sebi,

Thanks for the soldering tips.

About /CE: I think you need a pull-up resistor for the disconnected input, since the wire to the switch acts as an antenna switching /CE active at 50Hz?

[Sebi]

Sebi,

Thanks for the soldering tips.

About /CE: I think you need a pull-up resistor for the disconnected input, since the wire to the switch acts as an antenna switching /CE active at 50Hz?

Yes, you need a pull-up. The VCC-pins of the eeprom are on the same side as the /CE-pin, and there is a cap on each VCC pin. So I soldered a 10K smd resistor to the caps and connected each resistor to one of the /CE-pins, and from there to the switch.

Sebi

[Sharedoc]

OK, I got the 28F004 off the PCB, but it lost three pins 21, 22 and 23. I broke the epoxy to dig the remains of the leads and I got them uncovered from the epoxy mass, so I am back in business.

The wire trick didn't work for me. I used wire-wrap wire whick was too thin and is snapped twice. In the end I had to use surgeon knife.

[KachiWachi]

I don't know how much room is available (or what else may be nearby), but sometimes the girls at work use high-temperature heat guns to unsolder fine-pitch lead devices (and soften epoxy)...maybe this can help you all out here.