Service Forum

[W. Irving]

I needed to update the BIOS of my IBM 600 (2645-910). According to the flash utility, the rewriting succeed, so I restarted as instructed and all it does now is return the error code "175". Apparently this means "Bad EEPROM CRC #1".
Shall I assume the EEPROM in question is the BIOS chip?
What can I do? Is the EEPROM soldered? Even if it is I reckon I could remove it. But what should I flash it with?

Thank you for any help you can provide!

[victor]

The eeprom with the problem is not the one with the BIOS code,
the BIOS is fine, otherwise the laptop would be dead by now.

What exactly is the model, 600? or 600x, 600e?
_________________
Victor Voinea
ALLservice HQ, Romania.

[W. Irving]

Thanx for your reply victor!

It's a regular 600.
So it's a problem with the infamous 24RF08 (or whatever it's called on the 600, I think I read somewhere it's a different chip)?
There were no passwords set at all.

How do I fix this problem? I'm an above average solderer and I don't fear having to build an EEPROM writer.

[victor]

Then is a 24C01.
You have to open the laptop, read the 24C01 and send us the eeprom dump to fix the sums. To read the eeprom you can use R24C01 that comes with the R24RF08 kit or PonyProg or whatever else.
Futhermore, R24C01 is made to work for Thinkpad kind of chips so I suggest you to use this one.
The eeprom legs layout is the same as 24RF08, the procedure is the same only that you need the other software.
Oh, use the driven circuit and do not unsolder the chip, is not necessary.

CRC repair will be free, as this model is quite obsolete.

Good luck!
_________________
Victor Voinea
ALLservice HQ, Romania.

[W. Irving]

Thank you victor!

I'll get to it next week, when I've bought the MAX232. Seems very simple! I will use r24c01 and the driven circuit, as you recommended.
I only have one question until then. In the instructions in the readme.pdf, it says to have the Thinkpad running while reading and writing. Is this correct?

[bob]

Hello,

I wouldn't see other way to power the eeprom, unless you remove it from the board.
Just do like in the tutorial, but imagine that you read the 24C01 instead of 24RF08, using the r24C01. The dump should have 128 bytes of data.

[W. Irving]

It works!
I have the dump of the 24C01. 128 bytes. Where do I email this, Victor?

I had a few problems though. For future reference, please note:

1. Make sure the serial cable is not a null modem cable (often distinguishable by the fact that besides the 9 pin connectors it also has a 25pin connector in each end, and they are all females).

2. The baud rate (probably) should not exceed 116kbit/s, which is the max baud rate for the MAX232. These settings worked for me:
Baud rate: 115200
Data bits: 8
Parity: none
Stop bits: 1
Flow control: none
Although I suspect the circuit, being so simple, isn't too picky!
Change in Device Manager -> Ports category -> double click the COM port you are using -> Port Settings tab.

3. Is the mainboard is actually running? Sure way of finding out: if you can detect ~5V across GND (pin 4) and VCC (pin 8 ) of the 24C01, then you're good. The mainboard itself, stripped down, only seems to draw around 4mA when running, and 2mA when in standby. But do not rely on these figures. Also, there is an amber LED on the bottom of the mainboard (same side as the chip, for a TP 600) that lights up for a brief moment when you switch it on.

4. Use itester.exe to find out if your circuit is good. SDA and SCL, when active (the program alternates between high/low), should provide a ~5V voltage. Although this method will give you a positive reading even if the baud rate is too high. Have not tried if you need to have the chip connected to measure this - I think not.

Once again - thank you Victor!

[victor]

I just saw your post, send the file to support-at-allservice.ro

HAve you read it with R24C01, right?
_________________
Victor Voinea
ALLservice HQ, Romania.

[W. Irving]

I did indeed.
The command I used, complete with required switches:

[victor]

Ok, I will take a look in the morning. To write the fixed dump you'll have to use PonyProg, I don't have a "W24C01" for the moment
_________________
Victor Voinea
ALLservice HQ, Romania.

[W. Irving]

Flashing went well. Verified the write - succeeded. Did it manually too by running r24c01.exe to retrieve the new contents, and compared this to the file I received from you Victor. They matched.

Assembled the laptop, and now it gives me an 00188 error instead.
Although this time it doesn't just print the error code in white and black in the top left corner - it detects 32MB of RAM (some sort of onboard RAM I assume, I have 192MB installed but it seems to disregard it), displays the IBM logo, enters the system setup GUI, displays the 00188 code, and then stops.

According to Lenovo, 188 means "EEPROM CRC #1 error.". Though I found another manual claiming 188 means "Bad EEPROM CRC 2."..
Googled a bit and found a page suggesting the serial number is incorrect (which strikes me as a bit odd since the serial number in the newly dumped file is correct!), which can be fixed by running the IBM Hardware Maintenance Disk (allegedly, it is the only disk the computer will attempt to boot). Downloaded this (v 1.72), put the contents on a floppy and booted. No difference.

What do I do now?

[victor]

0175 is the first error prompted, but there could me more than one error in the dump. In this case is a serial sum error, is also possible to have a power problem.
I will send you a new file.
_________________
Victor Voinea
ALLservice HQ, Romania.

[W. Irving]

Thank you Victor!

I forgot to mention that it does prompt for a password, and it also accepts the password you gave me.

[W. Irving]

Perfect! The new binary works brilliantly! The laptop has an incorrect serial number and UUID, but that's not important.

Thank you Mariana and Victor, you've been a great help and I'd buy you both dinner had you not been so (relatively) far away!

Problem solved.