Huh, the International Atomic Energy Agency used them too but there is almost zero info on them....

Horun wrote on 2024-09-24, 03:19:

Huh, the International Atomic Energy Agency used them too but there is almost zero info on them....

I used one in the late 90s for developing EOD software, so when I saw it in a pile of junk, I knew exactly what it was. EOD guys really punish their gear 😝 we also had Panasonic Toughbooks the Japanese salesman team were horrified and then relieved when my boss (ex Army) asks whether the hardware is up to the job, the moment I say yes (like the man from Del Monte) the laptop is snatched off the desk and then thrown on the ground and run over by a Land Rover, then dropped back on the desk to see if it still worked. It did. Phew!

I thought I'd drag it out and fire it up. Amazingly, no memory errors, the screen had lots of lines and the backlight faded out, but rubbing my thumb around the display solved those (for now) the Tracker sounds good 👍

I took it apart again. This time I went all the way (without LBJ) Took the motherboard out, the card cage, the backplane and eventually the PSU. This isn't really a laptop, it feels more like a mini PC put into a keyboard case with an ISA backplane. Going through the PSU, there is a smell of leaky capacitors and indeed, two of them left a leg behind on the board. The remainder tested OK-ish, so I might reuse some of them (although Eric @ IDoCars - might well throw them across the workshop!)

and going further in...

...and on to the guts

Note when working on this board, there is a dot to mark the POSITIVE leg of the capacitor and it lines up with the orientation of the capacitor in the board (useful to pay attention here as there are lots of caps in a tight area) I'll have to order parts as I don't have some of these odd ball sizes in my parts stash. All the regulators test OK, the diodes are good and the MOSFETs are also good. There are also caps on the motherboard that will need probably need replacing, especially the SMD parts.

Got to open my eyes! It's an 8 pin mini-DIN for power and the pinout is printed on a label right next to the port. I thought it was a logo, d'oh! I'd love to get hold of a battery for this thing, it must have been epic.

Got the capacitors fitted and the MOSFETs, diodes and regulators on their heatsink refitted. This took over an hour because of the tight spacing. The designers like to put components within the footprint of other components. Yeah.

I have this exact computer, minus a hard drive. It was removed before I got ahold of it. The computer works with everything I've tried, however, I cannot find what type of hard drive and interface it uses. Seeing as you have a working hard drive and cable, would you be able to tell me what kind of drive it is and the interface?

Regards, The Engineer

The Engineer wrote on 2024-10-15, 17:57:

I have this exact computer, minus a hard drive. It was removed before I got ahold of it. The computer works with everything I've tried, however, I cannot find what type of hard drive and interface it uses. Seeing as you have a working hard drive and cable, would you be able to tell me what kind of drive it is and the interface?

Regards, The Engineer

It's a 2.5" 44 pin PATA drive. I'm replacing mine with a 64GB KingSpec SSD as the mechanical drive is getting long in the tooth now and if it goes bad, it's a lot of aggro to replace it. You might need to replace the CMOS battery as well.

EDIT:added photo of the drive, it isn't a Conner, d'oh!

I put the unit back together, quite a palaver and then discovered the SSD refused to work with the interface. I tried every BIOS setting and anything written to the drive would vanish after rebooting. This Fieldworks FW7600 has no LBA despite being 1995 dated BIOS and I suspect the SSD won't allow the old CHS regime to work. The memory errors on boot are still there, so it wasn't a PSU issue, despite the rather grotty voltage rails. Perhaps this has a CPU problem? Perhaps I need to find a BIOS update (if such a thing exists) and get LBA enabled so I can use "modern" SSDs.

I tried to use an IDE->8 GB CF Card combo that successfully worked in a couple of 386 Compaq Deskpro machines. On the Fieldworks unit, the BIOS could see the drive, but I could not write to the CF Card. I tried smaller cards, none of them worked. I then put XTBIOS onto an EPROM and put it into the 3C509 board that I have in this machine, however, it doesn't do anything. I suspect it is because the board has a PCI HDD controller.

Damn this FieldWorks machine feels dodgy, no LBA in the BIOS and weird behaviour with drives that aren't CHS/ATA. I really wish there was a BIOS update for this machine, but I can't find one. The Pentium version came out around the time this machine was new, so I doubt there was much interest in updating it. The BIOS currently installed is PhoenixBIOS A486 v1.03 1994 and an additional Fieldworks BIOS 1.1.2 11/10/95.

EDIT: oh dear, I put the original HDD back in and accidentally put the connector offset to the left by one column of pins and I got a weird beeping, turned it off, opened it up and I got the whiff of magic smoke.

The HDD no longer spins. D'oh! Putting the HDD onto the PC and I can read/write to it, so the drive is OK.

It seems the motherboard no longer functions with IDE, something let go, smell-o-vision seems to be somewhere around Q8 area. No idea what Q8 is, but looking at the tracks around it, perhaps some kind of MOSFET? I think the +5V motor pin got shorted to ground by the FUBAR with the cable.

I have never been able to find schematics for this machine, so repair is going to be tough (although not impossible) if the IDE controller has died, that is a lot more work in both repair and parts sourcing. In future, I will put a blanking pin into the cable to stop stupid things like this destroying an otherwise working system.

EDIT2: I figured out the logo was for a long gone company called "Siliconix" it was bought out by Vishay gradually from 1997 to 2005. On the upside, I know what it is, a P-channel MOSFET (data sheet attached) hopefully that is all that I killed...

EDIT3: A modern replacement appears to be available from the same manufacturer - SI9407BDY-T1-E3 - the major difference being pin #1 is now source rather than N/C and it's RoHS, yay!

EDIT4: I found a 44 pin IDE pinout and yup, putting the connector one out away from pin #1 i.e. pin #1 is now connected to pin #3 from the PC. That connects (pin #41) Logic +5V to ground (pin #43) via ground pins #24 & #26 and poooof goes the MOSFET.

I have replaced Q8 with a different, more modern device, a Fairchild FDS6679AZ https://mou.sr/4fDI0Wt and the machine is working again.

I also added a PCMCIA CF card reader to the PCMCIA slot, installed the drivers and prepared the card. That works only up to 528M bytes, but that's plenty enough to transfer files from the PC. I don't have any more time at the moment to spend on this, I'll have to find either a replacement BIOS or a more compatible SSD solution. One positive thing I have noticed is that the system runs quite a bit cooler than before and the fan doesn't run much at all - probably PSU related.

Just to help out others with this weird machine, these are all the supplied drivers specific to this system.

...and the final ZIP file (5 file limit)

I bought a new hard drive for the machine, a 1.8G Hitachi drive, entered the CHS as written on the drive and DOS detected it as 1.8G without any fuss...

I then installed DOS 6.22 and the PCMCIA ATA driver to get Windows 95B installed. Despite W95 listing the CF card plugged into the PCMCIA adapter as a "Removable Drive" it certainly isn't - well, not when the computer is powered on anyway, so I got my first BSOD...

I put a whole bunch of vintage, period correct stuff on, Office 97 Professional, VS6, VFP9, Microsoft Arcade/Return of Arcade and of course, Doom.

The machine is working well, so I decided to clean the original hard drive and reformat it. Lo-and-behold, no bad sectors after a clean and format... It doesn't sound bad either, so I've got a spare HDD. Yay!

Browsing the web is the next task to attempt, I do have the 3C509/Etherlink III card installed (minus the XTIDE EPROM) and I want to make a battery pack for this and I am definitely going to try and use it on a flight to Singapore - some of the "vintage" laptops I have used, got a pilot out to have a look, only a die hard would use a Compaq Portable 386 to write documents - one advantage of having power at my seat 😋 Everything looks so green after a couple of hours.

Playing with VB6 - oh the memories of having to write VB6 apps, so many really bad VB programmers BITD - don't blame the tool.

I got the 3com 3c509 (EtherLink III) card working, I had forgotten about configuring the EEPROM, it has been over 20 years since using four of these as a firewall/DMZ when I first got broadband. Once the correct port and IRQ were set (avoiding conflicts) I could successfully run the 3com diagnostic tests.

I tried IE4 (that came with Office 97) and it really doesn't like anything above basic HTML. My automation project, the first Raspberry Pi project I did over a decade ago, still going strong on the original hardware.

I upgraded to IE5.5 and things are much better. A real estate website I created almost 25 years ago, the server and database running on a VM (obviously not on this machine) ASP/SQL Server, displays properly without every mouse movement creating an exception - I rewrote it to target IE6 in 2002.