compucolor

[Jim Battle]

Although many people on this list won't be particularly interested in 
the auction part of this post, scan down a bit to read some more 
traditional cctalk fare.

I want to bring notice to the sale of a compucolor machine on ebay. 
They are rare enough that it seemed worth mentioning.  Also, it is filed 
in a category that would make it easy to miss (Computers, IT & Office > 
Vintage).

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=1247&item=5192237918&rd=1&ssPageName=WDVW

It is located in Australia.  There is about a day left.

disclaimer: I'm not the seller, but have bought disks from him before.

Since we are on the subject, the purchase of the said disks motivated me 
to get my Compucolor working.  I've had it for a couple of years but 
between moving, having other old computers to tend to, and the fact that 
it was reported to have produced smoke the last time it was powered up, 
I never got around to it.

The first step was to check for any damaged components or traces, but 
none were apparent, so I just gave it a go.  Before the variac police 
come and get me, it was powered up just before I received it. 
Miraculously, it simply just worked.  I imagine the smoke that the 
original owner experience was due to dust on the CRT burning off.  I 
have no other explanation.

Anyway, the disk drive wouldn't read anything.  I tried to INItialize a 
blank disk but that failed too.  Lacking a strobe or even fluorescent 
lights, I couldn't tell from the tach disk on the drive if the speed was 
off or not, but it seemed like a plausible cause for all the disks to be 
  dead (I should mention I scoped the read logic and it was detecting 
transitions).  I finally just twiddled with the speed pot to find that 
the speed was way off.  Although incandescents have a lot of glow during 
the power line's zero crossing, near the right speed I could see the 
strobe pattern well enough.  The real reason I couldn't see the pattern 
was that the speed had been so far off.

After adjusting the speed I was able to read most of the disks, although 
sometimes with retries.  Fortunately, all the disks that ISC put out for 
the compucolor recorded all the programs on both sides (it was a single 
sided drive) so even with hard sector failures I was able to get everything.

Now the main problem is pincushioning and color convergence.  I did some 
simple adjustments to improve color convergence, but without doing 
something much more involved, it isn't possible to get all regions to 
converge at the same time.  For now I'll just live with the problem.

One of the disks that I have is the disk formatter program.  ISC sold 
preformatted disks at $10 for two, and they didn't supply the software 
so that one could format the disks at home.  Apparently late in the game 
they relented and sold the formatter program, which eventually made its 
way to me.  Now I can mint more formatted disks (I hope -- I haven't 
actually tried running the program yet).

Disks couldn't be formatted using other computers because of the hokey 
(although dirt cheap) disk interface.  The Compucolor has a TMS 5501 
multifunction interface chip which contains, in part, a serial port 
controller.  This serial port controller is used for the RS-232 serial 
port of the machine, and it is good up to 9600 baud (but there is no 
hardware flow control).  Anyway, ISC took advantage of an undocumented 
test mode of the chip to drive the serial port at 8x speed.  The disk 
looks like a high speed serial channel and the data is simply 
conditioned and drives the r/w head of the disk drive.  I haven't looked 
into it yet, but they must be encoding each real data byte into two 
transmitted bytes in order to ensure sufficient transition density and 
no accumulated DC bias.  The net effect is that the disk holds only a 
bit over 50 KB.