NotebookKiller wrote on 2021-11-17, 11:17:

Hello. I've posed such a problem with my T4400C, but managed to "fix" it. The solution is to use IDE-to-SD converter based on FC1307 chip. I've changed the device name in firmware from "SINTECHI HighSpeed SD to CF Adapter V1.0" to "TOSHIBA MK2124FC " at offset 00003F32 and correcting the checksum at 00003FFE to B249 (your's can be different). After that any SD card inserted in this adapter will be seen as Toshiba MK2124FC 2.5" 124MB disk on Toshiba laptop, but any modern PC or USB-to-IDE adapter will see it's full capacity. Create the partition with FDISK on Toshiba and enjoy!

UPD: Attached fixed firmware

This is utterly awesome. Sorry for the really late reply. I'd parked the issue as I found a donor 1850 with a broken screen. Obviously the fact remains the hard drive will eventually fail so to have the chance to replace it with something more reliable will help.

I have a couple of questions though. Did you have to desolder the flash chip to flash it? How/what did you use to calculate the checksum?

Thanks again!

I used HxD to edit firmware and calculate checksum. As for desoldering... I don't remember. SO-8 is so comfortable for soldering, that I don't even notice the process.

looks like there is a nice solution here:

https://conventionalmemories.com/wiki_cm/CF_a … shiba_portables

Great idea, but not for every laptop. And "where to buy" link doesn't work... As for IDE-SD adapter's, they're "penny for dozen" (around $10) on Aliexpress.

NotebookKiller wrote on 2021-11-17, 11:17:

Hello. I've posed such a problem with my T4400C, but managed to "fix" it. The solution is to use IDE-to-SD converter based on FC1307 chip. I've changed the device name in firmware from "SINTECHI HighSpeed SD to CF Adapter V1.0" to "TOSHIBA MK2124FC " at offset 00003F32 and correcting the checksum at 00003FFE to B249 (your's can be different). After that any SD card inserted in this adapter will be seen as Toshiba MK2124FC 2.5" 124MB disk on Toshiba laptop, but any modern PC or USB-to-IDE adapter will see it's full capacity. Create the partition with FDISK on Toshiba and enjoy!

UPD: Attached fixed firmware

Hello, I was wondering if you could help with the checksum calculation? I can't work out how you're getting B249 and I've been down a long rabbit hole researching checksum calculation and getting nowhere. I've got a GRiDCase 1520 and to use an unpatched BIOS I'd need to specify a white-listed Conner model in place of the Toshiba. Your help would be greatly appreciated.

Areal was unreliable thing back in the day. Got replaced with another 2.5" hard drive, 200MB instead.

Get any 2.5" pata even 100GB, it should show as 500MB this way. Like I do with mine using 25GB 3.5" on sata PCI card as 2GB in DOS partition on a pentium.

Cheers,

pentiumspeed wrote on 2022-02-06, 20:44:

Areal was unreliable thing back in the day. Got replaced with another 2.5" hard drive, 200MB instead.

Get any 2.5" pata even 100GB, it should show as 500MB this way. Like I do with mine using 25GB 3.5" on sata PCI card as 2GB in DOS partition on a pentium.

Cheers,

You’re missing the point. Please read the whole thread before replying. The issue is that these laptops only allow the specific models whitelisted in the bios. Ergo we need to fake this ID in the firmware of the drive connected or the bios will reject it and merely say no drive connected.

entaled wrote on 2022-02-06, 20:35:

NotebookKiller wrote on 2021-11-17, 11:17:

Hello. I've posed such a problem with my T4400C, but managed to "fix" it. The solution is to use IDE-to-SD converter based on FC1307 chip. I've changed the device name in firmware from "SINTECHI HighSpeed SD to CF Adapter V1.0" to "TOSHIBA MK2124FC " at offset 00003F32 and correcting the checksum at 00003FFE to B249 (your's can be different). After that any SD card inserted in this adapter will be seen as Toshiba MK2124FC 2.5" 124MB disk on Toshiba laptop, but any modern PC or USB-to-IDE adapter will see it's full capacity. Create the partition with FDISK on Toshiba and enjoy!

UPD: Attached fixed firmware

Hello, I was wondering if you could help with the checksum calculation? I can't work out how you're getting B249 and I've been down a long rabbit hole researching checksum calculation and getting nowhere. I've got a GRiDCase 1520 and to use an unpatched BIOS I'd need to specify a white-listed Conner model in place of the Toshiba. Your help would be greatly appreciated.

I'm greatly embarrassed, but I can't remember the algorithm and address range used for checksum calculation. Need to do re-reversing the firmware...

NotebookKiller wrote on 2022-02-07, 15:36:

entaled wrote on 2022-02-06, 20:35:

NotebookKiller wrote on 2021-11-17, 11:17:

Hello. I've posed such a problem with my T4400C, but managed to "fix" it. The solution is to use IDE-to-SD converter based on FC1307 chip. I've changed the device name in firmware from "SINTECHI HighSpeed SD to CF Adapter V1.0" to "TOSHIBA MK2124FC " at offset 00003F32 and correcting the checksum at 00003FFE to B249 (your's can be different). After that any SD card inserted in this adapter will be seen as Toshiba MK2124FC 2.5" 124MB disk on Toshiba laptop, but any modern PC or USB-to-IDE adapter will see it's full capacity. Create the partition with FDISK on Toshiba and enjoy!

UPD: Attached fixed firmware

Hello, I was wondering if you could help with the checksum calculation? I can't work out how you're getting B249 and I've been down a long rabbit hole researching checksum calculation and getting nowhere. I've got a GRiDCase 1520 and to use an unpatched BIOS I'd need to specify a white-listed Conner model in place of the Toshiba. Your help would be greatly appreciated.

I'm greatly embarrassed, but I can't remember the algorithm and address range used for checksum calculation. Need to do re-reversing the firmware...

It would be greatly appreciated if you would be able to write a little guide on it for those of us less experienced with such things please.

I have no words to describe the depth of my frustration and anger. Two days of googling various checksum calculators, trying different address ranges and algorithms in HxD - with no result. Always incorrect checksum. But... then... a flash of memory... WinHex! I've used WinHex that time! And it's Checksum (16 bit) algorithms gives correct values (use CTRL-F2)! If you select from start to 3FFF you'll have a zero checksum with correct file (original firmware or my mod). If you made some changes and want to get checksum value, select from 0 to 3FFD and press CTRL-F2, choose "Checksum (16 bit)". You'll get a number (let it be XY), that is two's complement for a checksum with bytes swapped. To get the correct value to write to 3FFE-3FFF you'll have to subtract XY from 10000h (windows' calculator in programmer mode can do this), and swap it's bytes. LM=10000h-XY, @3FFE=M, @3FFF=L. Hope you understand 😀

UPD!
In HxD to get the same checksum you need to select "custom checksum", then two radiobuttons to 16bit, little endian. Result is XY, then see above

NotebookKiller wrote on 2022-02-09, 12:00:

I have no words to describe the depth of my frustration and anger. Two days of googling various checksum calculators, trying different address ranges and algorithms in HxD - with no result. Always incorrect checksum. But... then... a flash of memory... WinHex! I've used WinHex that time! And it's Checksum (16 bit) algorithms gives correct values (use CTRL-F2)! If you select from start to 3FFF you'll have a zero checksum with correct file (original firmware or my mod). If you made some changes and want to get checksum value, select from 0 to 3FFD and press CTRL-F2, choose "Checksum (16 bit)". You'll get a number (let it be XY), that is two's complement for a checksum with bytes swapped. To get the correct value to write to 3FFE-3FFF you'll have to subtract XY from 10000h (windows' calculator in programmer mode can do this), and swap it's bytes. LM=10000h-XY, @3FFE=M, @3FFF=L. Hope you understand 😀

UPD!
In HxD to get the same checksum you need to select "custom checksum", then two radiobuttons to 16bit, little endian. Result is XY, then see above

That's amazing, thank you!

I feel your pain and anger. There's nothing worse than trying to retrace your steps online trying to find that elusive one and only page you found 2+ weeks ago and getting nothing. You know it's out there but it just doesn't want to appear again...

NotebookKiller wrote on 2022-02-09, 12:00:

I have no words to describe the depth of my frustration and anger. Two days of googling various checksum calculators, trying different address ranges and algorithms in HxD - with no result. Always incorrect checksum. But... then... a flash of memory... WinHex! I've used WinHex that time! And it's Checksum (16 bit) algorithms gives correct values (use CTRL-F2)! If you select from start to 3FFF you'll have a zero checksum with correct file (original firmware or my mod). If you made some changes and want to get checksum value, select from 0 to 3FFD and press CTRL-F2, choose "Checksum (16 bit)". You'll get a number (let it be XY), that is two's complement for a checksum with bytes swapped. To get the correct value to write to 3FFE-3FFF you'll have to subtract XY from 10000h (windows' calculator in programmer mode can do this), and swap it's bytes. LM=10000h-XY, @3FFE=M, @3FFF=L. Hope you understand 😀

UPD!
In HxD to get the same checksum you need to select "custom checksum", then two radiobuttons to 16bit, little endian. Result is XY, then see above

Thanks for tracking down the info and providing a good explanation, I'm able to duplicate the checksum now and am confident I can flash one of my Sintechi adapters. I'll try and get back with some progress soon, just finishing up a Compaq LTE/286 restoration here, then I can get around to playing with the GRiDCase again. I think flashing these cheap SD-IDE adapters will be a good solution for machines that have a whitelist. I hope it helps more people save machines from the landfill.

NotebookKiller wrote on 2022-01-17, 15:54:

I used HxD to edit firmware and calculate checksum. As for desoldering... I don't remember. SO-8 is so comfortable for soldering, that I don't even notice the process.

Could you please describe how you have flashed the chip (which tool/software)?
I used Google, but there is no data sheet available for this chip.

I used CH341A programmer from aliexpress. There's a 8 pin SO-8 flash chip on IDE-SD adapter. I've unsoldered it and flashed (or used a clip to flash it in place? really don't remember).

S1990 wrote on 2022-02-22, 15:00:

NotebookKiller wrote on 2022-01-17, 15:54:

I used HxD to edit firmware and calculate checksum. As for desoldering... I don't remember. SO-8 is so comfortable for soldering, that I don't even notice the process.

Could you please describe how you have flashed the chip (which tool/software)?
I used Google, but there is no data sheet available for this chip.

They are standard pinout SPI Flash/EEPROMs
One of mine is marked 25LC512, reports itself as Pm25LD512(C)
And another I have is marked 25V40066, reports as MX25L4005(A/C)/MX25L4006E

Both have datasheets and are supported by flashrom on linux. I imagine they work with minipro
and xgpro but I haven't tried. I use flashrom for SPI since it gives me fewer headaches.

The 25V40066 marking is wearing off, I'd guess they're all/mostly reclaimed and remarked.
Regardless, they should report correctly to any SPI programmer, I think.

I will have to desolder and program it in my TL866A programmer. It won’t program in situe. I’ve tried.

mikeyp wrote on 2022-02-22, 18:03:

I will have to desolder and program it in my TL866A programmer. It won’t program in situe. I’ve tried.

i just flashed mine without removing it from the board, to see if it would work...

1sudo flashrom --programmer ch341a_spi -w firmware.bin 
2flashrom v1.2 on Linux 5.16.10-arch1-1 (x86_64)
3flashrom is free software, get the source code at https://flashrom.org
4
5Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns).
6Found PMC flash chip "Pm25LD512(C)" (64 kB, SPI) on ch341a_spi.
7Reading old flash chip contents... done.
8Erasing and writing flash chip... Erase/write done.
9Verifying flash... VERIFIED.

This is with a CH341A programmer, but that shouldn't matter.
The chip clip is very fiddly, try holding it firmly in place.

Well, it works in the sense that the BIOS is happy to see the SD-IDE as a "Conner CP3104" now.. but I've hit a problem I always seem to run into with CF-IDE adapters. I can write but not read. I can fdisk, format /s, install DOS, etc.. I can write to the drive fine but any attempt to read from the drive, including booting, just hangs. I've gotten around it on other machines with an overlay but I was hoping that wouldn't be needed with this adapter.

Thank you all for the detailed answers. Could you provide a link for a suitable IDE-SD adapter? Which one have you used?

Is it possible to make a modification tutorial to place a cf card on the Toshiba T1800,1850? Step by step with all possible materials and details?
to be able to use it as a guide for the replacement of the original hard drive and bring these laptops back to life??