Compaq Portable Ii Restoration Project, They don't make them like they used to

So I got an old Compaq Portable II on eBay. The twenty, count them, twenty megabyte hard drive seems functional (spins up and self-tests, doesn't sound weird, doesn't reset its heads), but it has no system on it, and the floppy is disconnected. The keyboard is also slightly broken. Powering it up reveals that everything else seems to work.

The floppy drive is an old 360kb unit that can be connected to a more modern system using a side-connector floppy cable (which I happen to have); this would allow me to put the floppy images from the Compaq website (yes, they still keep this super-old stuff in their archive) on disks and install the OS to the hard drive... if I had any 5.25" floppies. Which I don't, and which I doubt I'll be able to locate for a non-insane price and in decent conditions.

A thread in a retrocomputing forum reveals that the system is compatible with 1.44MB 3.5" floppy drives, of which I have, I dunno, a dozen or so salvaged from old computers. So a plan was hatched to open up the old beast and put a proper floppy drive in it, from which to install the least old DOS the machine can run (it was sold with 3.1, but since it's a 386 it should be able to run newer versions). The keyboard has to be repaired too, but it's probably just a few springs.

This was two years ago. Today I finally found the willpower to start tearing the thing apart.

The internal construction pleases me greatly: everything is clearly labeled, compartmentalized and built to be easy to service, even going so far as to use oblong screw holes. Why? So you can remove the metal panels without having to completely remove the screws - just unscrew them slightly and slide the panel sideways and up. No running after screws on the floor.

The outer construction, on the other hand, doesn't please me at all. The plastic case (which will require thorough cleaning, as someone spilled sticky sweet stuff on it ages ago and now it reeks something disgusting) is made of four panels: two side ones, and two top and bottom. They are mounted in such a way that removing the top and bottom panels requires you to remove the side panels first, but you can't properly remove the side panels until you've removed the top and bottom ones.
You're supposed to, I think, push on some tabs connecting the side panels in order to unlatch them, but no amount of pushing I did seemed to convince them to let go. In the end I just stuck a large screwdriver in there and used it as a lever; I broke a tab, but the others finally yielded (with sickening CRACKs as they did so).
I'd like to find the people who thought it was a good idea to mount the plastic panels like that instead of with screws, and kill them. Well, unless they're already dead of old age, that is...
Edit: well, I'm stupid. You're supposed to unscrew the case and lift it up entirely, taking apart nothing. D'oh...

The computer seems to use a weird backplane system, where the motherboard only links the different components together and does nothing at all by itself. The CPU is on a card fitted in what looks like a shorter ISA slot; the card housing the CPU is about as long as the computer itself. Other components I still have to figure out.
Edit: incorrect. CPU is on board, not on card.

And now, fun fact time!

Dead spiders found so far: eight
Dead spiderlings found so far: a few dozen, all strung up in their parents' nets
Live spiders found so far: zero. Hoping very hard it stays that way; I'm a friend to spiders, but not when they are inside equipment I'm putting my hands in
Spider eggs (undoubtedly frozen by generations and waiting for me to shine some light on them to hatch and spawn horrible beasts that'll eat me in my sleep) found so far: zero. Hoping this stays that way too
Times I've gone "eargh" while opening what can only be described as a mass grave of spiders: more than I can possibly remember

I wonder how it came to be that all the spiders died. It seems they all died together, as all the spiderlings are of the same size; if it was natural causes there'd be less of them, and they wouldn't all be so tiny. The most likely explanation I can come up with is that someone put the computer in a closed box; air could still get in, but prey couldn't, and the spiders all starved to death.

The original project was just to find the floppy connection socket, stick in a cable and connect a 3.5" drive; however, since I'd feel rather uncomfortable running a machine with so much dead stuff in it, I think I'll disassemble it completely and clean everything properly. I'll vacuum all the spidery stuff out (yes, I do know I must not use vacuum attachment with metallic ends), then start the process and document it with pictures.

Now if I could just stop this crawling feeling I have on my skin...

It may even still have the operating system on the harddisk.

These old things usually needs you to tell them how big the harddisk is and what type it is, so go into the BIOS (if you can) and specify the 20 mb harddisk, try different types until it boots, lot's of the REALLY OLD portables I've had - works that way!

given the topic subtitle, i thought this was appropriate.. ..(missing a 'D', but still mildly amusing..)

Joongle, I didn't think of that... the computer does warn me that the bios battery is dead, so it stands to reason that the hard drive settings were lost. I'll replace the battery when I take the thing apart (hoping against hope it's a primary and not one of those awful packs of coin-sized NiCads) and input the drive settings again.

Edit: the hard drive is a MFM unit, of the "dumb" kind that required a separate controlled card. However, under it there's a card that, if I'm not mistaken, translates MFM to IDE/ATA, as the cable that comes out of it looks a lot like a standard 40-wire IDE one. That then plugs into the main board.
If that's really IDE, it should be possible to interface this drive to a modern system (a P133 would probably see it; a Q6600 probably wouldn't), or more importantly to interface a suitably old IDE drive with the Portable.

Edit2: yup. From here: "From what I've read, the HD interface on the MB is IDE but many systems were shipped with MFM drives attached to MFM-IDE converters/adapters. "

Aaaand to top that of, the bios of the thing, most likely were customised to fit that interface converter thing. So Still...try to find the original harddisk settings, and it'll most likely boot.

Can I wash electronic boards with demineralized water?

I've performed some preliminary cleaning, but dust is so deeply ingrained that just vacuuming isn't enough.

To answer that, you might find the following disturbing:

Back in the days, I was a certified Commodore Technician, repairing Commodore 64, Amiga computers.

In order to clean certain customers computers - usually used by children and heavy smokers, we cleaned ALL the computers under the watery showers.

I used hot water + brown soap (natural soap, no additives), and they got a thorough rising every time.

After that, they where placed in a dry cabinet (a cabinet with a HOT FAN) placed there until the boards made creaking sounds from the extreme heat, this is necessary in order to get rid of the water still stuck under the microchips & sockets & random mechanical parts.

WARNING: Do NOT soak the disk-drive & harddisk!


The keyboard can be soaked, provided you dismantle it entirely, yes ,that also means the two wafer-layers were the contact points are.

I'd rather not get such old equipment to such fiery temperatures. It'll stay dusty; it'll be fine as long as it's de-spidered.

blow it out with an air hose ... good enough unless you have to go to the next level of cleaning .
use a long bristled brush to apply a little "physical assist" .
yes , evict the dead engineers .

ps ... de-ionized water is not the same thing as distilled .

Ions are what make water conductive. What problem would non-ionic trace impurities cause?

mineral deposits .

The deionizing process removes the dissolved minerals. That's the whole point.

Minerals can be dissolved in water in huge amounts as ionized salts. Non-ionizing elemental minerals or mineral compounds have very low solubility and can only be suspended in water, temporarily, in any reasonable concentration. They quickly settle and are easily filtered out.

Deionized water can contain parasites, bacteria, and viruses among other non-ionic contaminants. But I don't think the pcb will mind those.

System is de-spidered, fairly clean and ready to be reassembled. Replaced old dead cmos battery (a Tadiran 3.6V primary lithium AA) with two alkaline AAs in a holder. According to other sources, the CMOS works with even one alkaline, though the clock goes slow; with two, it should be fine. I put some paper and tape around the holder so if the alkies decide to leak they won't damage the mainboard.

I retract what I previously said about the backplane system; the 286 CPU is in fact socketed on the mainboard.

Don't worry about the spiders in old equipment, you will likely never find a live one. They all die of starvation, as there's no insects inside of electronic equipment as a rule. The spiders are too stupid to know this...they are attracted to cracks & crevices, especially those with any airflow through them...spiders like to spin webs in windy areas.

I've been working on the thing all night long, and I finally have some success to report.

The CMOS battery is emulated by the two AAAs to a sufficient degree; the computer no longer complains its battery is low.
The new 3.5" drive works fine; I still have to test it with 1.44MB disks, but an old DOS 3.2 720KB disk I have is read just fine.
Two keys on the keyboard seemed to have the spring broken, as they were stuck down; fortunately, some wiggling after opening the keyboard restored them to functionality. Sadly, the enter key seems to not register keypresses (every other key does). For this reason I wasn't able to do anything with the system other than sticking a DOS boot disk in the floppy drive and making it boot.
The computer says system properties are not set (this was expected). However, instead of inviting me to press a function key like all civilized machines do (and always have done), this one wants a special diagnostic disk to enter setup. Thanks a bunch, Compaq! Fortunately, I was able to download a diagnostic disk image from a site I found linked somewhere, so as soon as I repair the enter key I should be able to get it to recognize the hard drive and work properly.
Finally, the green-phosphor CRT seems in excellent shape. It has super-long ghosting, but I think this is to be expected.

The results are promising. I'll post back when I have more. In the meantime, I'll see if I have an old ISA ethernet card sitting around somewhere... I'd love to be able to get this thing online there are a few old BBSes left online (accessible via telnet), and I'd love to surf them using the Portable II.

Telnet still works on multi-GHz machines too, you know.

I know, and I've connected to a BBS with my quad-core main computer, but it just ain't the same thing. If I can get this thing online I want to experience first-hand how people used to do it way back in the eighties.

I connected a floppy drive to the aforementioned main computer (which must hate me for having given it such a low-tech device) and I'm ready to put stuff on the portable. The only thing left to do is to fix the enter key.

You're going to need a 300 baud acoustic coupled modem to get the real feeling

Yeah, I'm not that old-school.

heh heh heh ...

Restoring the CMOS is proving a horrible nightmare.
The compaq setup disk (which required me to install VirtualBox and create a DOS virtual machine) doesn't work: the computer says "load failure" and stops there. I've managed to find an image of a supposedly working setup floppy from a german site, but it's in an obscure format (.dsk) that needs archaic software CopyQM to work. CopyQM is DOS, so it doesn't work with Win7 x64. I'm trying to run it in the emulator as I did with the Compaq files, and it recognizes the image file, but it keeps refusing to write to the floppy.

I'll probably need to drag out my old P133 and install DOS on it just to be able to write the goddamn setup disk image.

It's running, finally. It refused to boot on any microsoft DOS above MS-DOS 3.2 (including 3.21), so I finally gave in and tried DR-DOS 3.41. That worked.
Obviously, I only found a proper MS-DOS 3.2 720k image after I did that... *sigh*

Anyway, I put the compaq utilities on the DR-DOS disk and they worked. The hard disk started up, and booted... MS-DOS 6.00. How can it be that it refused to boot floppies of anything above 3.2, but boots 6.00 from hard disk fine, I don't know, and frankly I'm way too tired of tinkering with the damn thing to figure it out. I've been trying to get it to boot for the past two days, and I'm mentally exhausted.

The system is fully working now; all I have to do is find an ISA ethernet card. I hope I have one somewhere...

Are you using the SYS command to prepare the diskette? My 286s had 5.25" floppies, but I have an old Tandy with a 720k floppy (no HD) that will boot and run DOS 6.2 (lol, if it still runs at all).

My first 286 (16MHz, 40MB HD, 3 Mb RAM) had DR-DOS 6-something (not sure), and it was a pretty good OS. Some of the file utilities were better than MS-DOS, and I used them until I went to FAT32.

After further testing, it's clear that anything above MS-DOS 3.2 and below 6.0 causes it to hang, but 6.00 boots. 6.22 makes it hang again.
I've no idea why it does this, but at least I know if the HD dies I can put 6.00 back on it.

I've found an old version of Word that runs on DOS. It's weird to do wordprocessing on a green phosphor screen
Oh, and Alley Cat works perfectly!

Maybe you can get the changelog somewhere to see what new features make it hang?

I have to deal with nano and vi on a weekly basis. Not a green phosphor screen, but terminal based none the less...