Hitachi, R.I.P.

The Fate of All Computers
January 19 2004


Sunday evening, January 18, 2004, as I was replying to a posting on color printers on the Web Consultant's Association Mailing list, my Hitachi 4140 VisionBook computer emitted a "Pzing..." sound and the screen blinked off for a second or three. Just as it sank into me that the machine might be dead, the screen came back with a letter I was typing still intact. A moment later, I resumed typing, trying to finish the letter up before checking further.

As I was typing, the Hitachi started giving intermittent disk drive timeout and /dev/hda reset failure errors, eventually locking up waiting for the disk drive. I pulled the drive and re-inserted it several times trying to clear the time-out errors. Actually, I had noticed problems before; the hard drive would sometimes not start up after the computer was "awakened" from low-power "sleep" mode which I used to cut power draw from 1.45 amps to 150ma overnight. I'd pull the drive out, and push it back in, restarting it. Did not work this time.

I powered up the HP notebook, and got back on line, reflecting that my Canon printer had also died this week, leaving me somewhat inconvenienced up here.

Problems come in threes, they say. What's next? Well, my favorite keyboard is beginning to have some problems detecting the control key... But I suspect it is the USB converter getting bogged by several messages regarding the caps lock key and LEDs. (We had this problem at Qume in 1988 with the design of the QVT-PCT keyboard. It had doomed several other keyboard designs as well, including one my boss had worked on. I was the first one to understand the nature of the problem, characterize it, prove it, and provide a keyboard-level protocol change which substantially eliminated it on all subsequent keyboards.)

Late that evening, I removed the hard drive from my Hitachi and replaced it with the original drive it had come with, a drive containing Windows 95. It still would not boot. I also tried the ROM BIOS utilities and a Linuxcare Rescue CD-ROM, but neither could find the hard drive. It looks like the hard drive controller had failed, not the disk drive itself.

I put the apparently failing drive on another computer (Oscar's old Hitachi which I had repaired some time back), and that booted ok, except for incompatible drivers for just about everything. Unfortunately, that old machine lacks a CD-ROM drive, ethernet, and a USB port, making it pretty much useless and preventing me from loading the drivers it needs to do anything.

I have the disk drive and data; but nothing to access it via. And the backup, on a USB drive, doesn't seem to work with the HP and SuSE Linux 8.1

Conclusion: My faithful Hitachi is dead. At least till I can set up a decent repair table with good lighting up here. But realistically, with the way notebook computers are built using surface mount devices and custom ASIC chips, there is almost no hope for it short of finding an identical model with a broken screen. And maybe not even then, as sometimes they change components without changing the model number. I really should have bought that $80 spare!


Laptop
Year
Features
Toshiba
1987
8088, 2fdd
(Loaner)
Mammut HP
1989
286 12mhz,
20megHD
NBS-16
1992?
386 16mhz
40mHD
GRiD 386
$1 surplus
386 12mhz
20mHD Plasma
Canon
1995?
486, 33mhz
340megHD
Color! Win95
Hitachi
1998
P133Mhz
2GigHD
HP Pavilion
2003
P1.33Ghz
18GigHD

Memories

The Hitachi lasted 5 years... I hauled a lot of data in that machine! ISI, IBM, Silicon Graphics, Financial Engines, IPIX, Mall-Net, SitePsych... some smaller projects; and almost two years out here in the dust, heat, and cold of the mountain. It was a really good solid machine!

I remember back in 1998, I was looking to move up from the 33MHz 486 based Canon notebook I'd used on site at Cisco, HP, ISI, Lockheed, and other places. It and another Hitachi were sitting dead on the demo counter at Fry's in Palo Alto, offered at half price. I asked the salesman if the price would be the same if I got one of them working. He said yes. I swapped a few parts (memory chips, CD-ROM drive, and such) between the two machines and that was the one that started working there on the demo counter. $750 or so, and off we went! (I'd done similar with the Canon, which had been missing a power supply and ball of the built-in trackball.)

When I had the kidney problem a few weeks later, this was the machine that kept me connected to the world. The AC Adapter kept overheating (possibly because it was meant for a different model Hitachi), so I set a Ziploc sandwich bag full of water on it, later building a little octagonal tower out of Kynex sticks to hold the bag in place. Eventually, I found a thick, heavy, machined steel puck, once an arm-stop weight from an old, coffee table sized SUN Micro hard drive; and a five inch long aluminum heat sink to keep the AC adapter cool. I remember at a Linux Installfest at the Computer Literacy Bookstore in San Jose, one guy looked at the shiny machined metal blocks on the power supply and commented to another fellow, asking what kind of high tech thing that might be. The other fellow replied "A heat sink". He'd asked me the same question five or ten minutes before and been much disappointed by that answer.

In Half Moon Bay and on the ranch, I used the Hitachi from my bedroom to run the cluster of machines that ran keyword reports and did other things. On the ranch, I reviewed the day's security photos on it. It was slow with all the images swapping on the hard drive; but it worked well enough to do the job.

I'd sometimes drive with it hot on the seat by me as I went looking for places to stay, driving instructions and notes handy, ready to take more notes. The really nice part was that it ran on 12 volts without fancy adapters. It was a sweet reliable machine!

You work with a machine for five years, you get to know all it's little quirks. And then it's gone...

Mammut...

I think the only portable machine that I liked as much, was Mammut, the HP/Heath SuperSport I chanced upon in 1988 (1989?). A 12mhz, 286 with a 20 meg hard drive. It was really something at the time! ...at the time... The only part of it that beats the current machines, is the solid keyboard it had. The case was pretty solid, too!

I use to sneak Mammut into IBM in a knapsack before PC's became common there. That's why I named it Mammut, the name German General Romnel gave the captured British APC he used as a staff car. It was already operating under false colors; the case said it was an HP. It was really a Heath-Zenith, and some months after I got it, I installed the Heath Zenith upgrade ROMs so I could use an external double density 5.25 inch floppy drive with it. For a while, it was our unofficial database machine at IBM, keeping track of test cases before we submitted them, reports and the like. It was the machine I put together The Baltic Printout on, and the machine I sent that fateful fax to CNN on, telling them to monitor the growing situation in Lithuania. That was really some machine in it's time! Quite the contrast for the Hitachi, which was already obsolete when I bought it.

Mammut also died of an electrical failure, a break in the wiring on the electroluminescent backlight nearly set it on fire. Then I broke the screen trying to fix it. I kept the carcass around for another three to five years before giving up trying to find a replacement screen and throwing it out. Not that I'd go back to it for major use today... but yes, I guess I liked it a lot more than the Hitachi because in it's day, it was far more of a high tech machine. Notebook computers hadn't quite taken off the way they have today. And because of the solid, full travel keyboard; something you never see today, not even on a desktop machine.

That was one great machine! And with Word Star and TabTalk, would STILL be a darned good machine for typing notes. Yes, it is one machine I'd have kept forever.

"Real" Machines

The only contender for Mammut would be the GRiD units I picked up for $1 each at Confer Metals. But the Grid keyboard has an awful layout, and they can't run but ten minutes without AC power. Still, the chocolate brown magnesium case and red Plasma displays give these machine a distinct impression, Mil-Tech, more like the impact of driving an original Hummer or even an army tank. Indeed, every time I look at one of them I think of Phil's army tank. Alas, they were obsolete when I got them.

But the Real computer I owned, one of the three pesonal computers that started the home / personal computer revolution, was the IMSAI 8080.

(The three, Apple II, MITS Altair, and IMSAI 8080 were all open specification, bus oriented machines spawning many board-level component vendors. That open bus was the key feature IBM copied when they built the IBM PC. And to further strengthen the machine, they endorsed only one operating system. Arguably the wrong one.)

(I chose the IMSAI over the MITS simply because the front panel switches were broad and flat. After toggling in the bootstrap loader on the small chrome rod-like switches of Don Mimlitch's MITS Altair, my fingers hurt!)

Nevertheless, much as I miss the IMSAI front panel and the LED Memory Access Dispay I built for it, you couldn't really do anything useful on it in 1976. Hardly any way to store anything on it till the floppy disk drive came out in 1978, and CPM gave us a common data format and operating system to exchange data on, soon followed by fairly standardized programming languages. That made personal computers actually useful, if only for writing, at least at first. My own attempt to create a writing system called Script-80 not withstanding, Word Star and SuperCalc (and their Apple II predecessors) made personal computers useful in business. The early-on real contender, the Apple II, started without a "real" cross-application operating system, and the resulting chaos of non-interchangeable tape and disk and data formats kept it from taking over. Little things... little things... Apple had the best chance of them all. THE first appliance type personal computer, and later many other innovations, and they blew it just enough from becoming number one.

The text oriented IBM PC and graphics oriented Apple Lisa (with ideas stolen from major investor Xerox and written by a custom programming company called Microsoft) that came out in 1982, though expensive, were something. Once again, Apple lost, this time due to price. It wasn't till the Local Area Network evolved, that their utility started challenging the IBM Mainframes with terminals in the corporate world. Even at IBM itself, the PC did not catch on till 1989, if only as a fancier terminal into the mainframe and IBM's corporate world wide communications network.

Elsewhere, dial up bulletin boards started early, even before real personal computers. The Source, The Well, Delphi, Compuserve, and AOL were early dial-in communities. Much of this was predicted, and in some cases based upon DoD's Project Delphi report, which I read with fascination in 1975. Then came Portal and Netcom, giving common people access to The Internet; and with that, the world changed again. Small dial-up provider communities met the international USENET News communities, formerly limited to the cognosi in universities and high tech industry. Much was lost in that flood, and the wave of USENET spam that followed! And then came the browser, Netscape, blowing open the doors to the masses who could not spell, touch type, and in many cases could not even form coherent written paragraphs.

In retrospect, it looks so obvious that it should happen this way. But so few understood it. And I was just a little too early, looking too far or not far enough ahead with my networking ideas, text processor Script 80, my VED editor, the "Community 100" typewriter proposal to SCM, and the TabTalk editor/browser. Too far, and just a little off to the side, with no one well enough versed in technical matters to discuss and evolve the seeds of ideas with. (Irene wasn't even close!) Still, I played in the Olympics of the mind, and that is no small thing!

The new thing is wearable computers. The Palm Pilot is old hat. I wrote about them in 1982, I think. The big thing holding it back is an overlay "heads up" style display. Till someone really does something about that... But I never made enough money nor the right people to pursue that research. Someone else will make that fortune as the PC, Browser, Camcorder and Cell Phone merge in a wearable appliance and evolve new communities. The Borg will not be; but there will be many transient little faintly Borgistic communities. If I but had a few million for hardware research, I would easily ride that wave, for I see it coming, see what it needs, etc. Or... think I see it. It is as inevitable as the internet!

Diagnosis

This Hitachi notebook machine only failed me twice before, once when a memory chip died here on the mountain, and back in Half Moon Bay, when a flexible PC board connector on the drive tray failed. Oh yes, and the night of the Croy Fire; but that was just a loose connector.

There was another problem, though. One of the hinges had jammed some time ago and the case broke there while I was closing it. With the leverage and small size of the hinge, I didn't even notice the shattered plastic till much later. The etched flat cable to the LCD screen passed through this broken hinge. Sooner or later, the cable would tear at this point. Which is why I had stopped taking the Hitachi anywhere.

Since this disk drive failure seemed to involve power failure to multiple systems, I think this is more serious than a broken wire. I think something in the drive controller path failed, shorted, the short burned out, and after the power regulators cooled off a second or two later, they reset. As power came back, the screen came back up and I heard the drive spin back up. The damage was probably complete at that moment; but Linux buffers disk access heavily. As Linux filled various buffers and tried to write to the drive, and as other tasks started trying to load various utility routines from disk, the Linux operating system core started issuing more and more error messages till the pending queue filled and it locked up. I'd seen this mode of failure before on other machines. Amazingly, if the disk drive comes back up, Linux resumes without crashing. Not the case here; access to the hard drive was burned out.

I took the Hitachi apart. I've fixed a lot of stuff in my time; I sometimes enjoy the puzzle of fixing things. The thing is glued together, plastic wrapped around this part and that board to keep them from touching the metal flashed case, etc. It was a chore to take apart! Too late, I found the keyboard has to be forcibly pried loose first, revealing additional screws. But I already knew the answer by then.

The drive controller is integral to the motherboard. That was the first sign that it was unlikely to be repairable. Worse, the drive controller electronics sit beneath the drive bay, which is soldered down to the motherboard for strength. The whole thing is myriads of itsy bitsy surface mount parts with a few large ASIC (Application Specific Integrated Circuit) chips. No visible burn marks, nothing to go on. And even if I saw an exploded chip, unless it was a resistor, capacitor, or diode, something common; nothing could be done. They don't make the same chips anymore. This isn't a desktop machine where you can pull out this board or that and replace it.

It's really hard to admit this is Non-Repairable. I loved that machine. But the keyboard was almost gone, the screen resolution and color depth makes the display obsolete and unusable for good color photo and current graphic work, the processor speed is obsolete, and it's... just Dead. Even fixed and working, it would not be worth more than $100, $150 tops working at Foothill, our techno-community Electronic Flea market. Heck, I remember turning down one just like it for $80 a year or two ago. Non-working? $10? $20? Cost $15 for a space to sell from. Last time I went to sell, I lost money on the deal.

(Al keeps saying I ought to sell something, just so I'd meet interesting people down there. He's probably right. But what? What clever things can I put on a CD-ROM that would sell to folks who might hire me, and not think badly of me for seeing me sell it? Has to be clever!)

With the motherboard partly blown and the keyboard bad, the only thing of any use, is the obsolete screen, and only to repair the same exact obsolete model. So reluctantly, I put it in the trash. What else?

Discarding Old Tech...

It reminds me of all the old TV sets we had in our cellar so long ago. Dead, almost dead, just needing a few minor parts. Till the next generation of TVs came in, and they were all worthless. One spring day, we set 13 TV sets out for the town's spring cleaning pickup. Phil had the same thing. We cleaned his basement one night, hauling I think it was around twenty old, once very valuable obsolete TV sets to the dump, taking two or three trips to do it, probably in 1970. Mostly console models and others in wooden cases, they had been stacked two high and three deep. We freed up a LOT of space!

It never feels right throwing such once-valuable things out; not then, not now, especially remembering how much those things meant to us during their lives. Maybe that's why I still have about four... five... maybe six quite obsolete computers scattered between here and the warehouse. It almost makes one cry for the memories. I can close my eyes, and see the old IMSAI sitting there, blinking its lights. But today, a computer is the data that is stored in it. The rest is just a rapidly evolving replaceable shell; little more than a work truck one buys and eventually discards when it wears out. Still does not help the memories though.

In the end, I pulled out the CPU chip, a tiny fan, and an oddly shaped heat sink as mementos. And of course, the memory chips we bought in 2002, and all the screws. The screws have the greatest chance of reuse.

Well, maybe it wasn't the end. A day later I pulled it all back out of the trash. I already have three other semi-dead notebooks; one more won't take up that much space. It started going together quite nicely till that double sided flexible ribbon, a myriad of fine etched lines with a ground plane on the other side, the one going through the sharp broken hinge to the LCD screen, tore at the hinge. (Picture Below) No way to fix that! No way to fix the drive controller either. Enough dead TVs already! Back to the trash barrel!

But no. It's still sitting there on the table... spares... Hey, I think it should still boot off the CD ROM... and run on a CRT... with an external keyboard... Knopix???

Or should I have pounded a stake through its silicon heart while I had the chance?

Power

On the mountain, the Hitachi ran on 12 to 14.5 volts day in and day out without complaining despite the official rating of 19 volts. Rated at 2.37 amps, it drew 1.3 amps most of the time, with rare excursions to 2 to 3 amps.

In contrast, this HP draws 2.45 amps at idle without graphics, and often up around four amps when active. Peak draw is over 9 amps, sustained for many minutes, enough to blow a fuse on my truck, and exceeding the 7 amp max full sunlight rating of the solar panel. Looks like I need another solar panel and a few more lead-acid storage batteries to run the HP much at night. Or to get back to civilization. (Which wouldn't be such a bad idea... except for the pollution.)

While the HP is ten times faster and has a larger, much nicer screen; I will long miss the Hitachi for it's reliability and low power consumption. It's that low power consumption that let me get away with using only one solar panel up here. That's going to be more of a problem now...

Rest in peace, Hitachi, rest in peace...
You carried me a long distance... As I carried you.

Meanwhile, life goes on. I have the data safe on disk, and the new HP machine is running with most of it.

-JVV-

Copyright (C) 2004, JVV



The Final Blow -- a torn high density video cable


The Hitachi 4140 is on the right.