PDA

View Full Version : just got a nice Compaq Portable (first model) today



Mike Chambers
May 10th, 2010, 11:10 PM
i bought it from duked on my IRC server (Lineman Duke on here) and it came in today. it's missing the hard drive cover, but thats just aesthetics. i'll take care of it in due time.

(warning: humongous pics in the links)

http://rubbermallet.org/compaq_portable_1.jpg

unfortunately, after a few hours something bad happened to the video RAM. now it looks like this:

http://rubbermallet.org/compaqwtf.jpg

it is just the card for sure though. it works fine with a VGA card or my herc CGA clone, but of course neither of those have connectors for the internal monitor. i just ordered a new one w/ warranty for $20, so i'll have it back up and running right soon with any luck.

it's a really really cool machine. a far cry from my osborne 1 :p

Minerva10210
May 11th, 2010, 01:33 AM
Hey, that's cool!

I must say that the green monitor sort of "fits" better for the era than my IBM 5155's amber monitor. It just sort of looks better :) ;)

I found an old ST225 over the weekend at my parents place, so I'll have a machine similar to yours then :D

per
May 11th, 2010, 01:52 AM
Regarding the video memory, it looks like some sort of addressing error. Since all characters are displayed in pairs, Address line 0 must fail at some point. In addition, based on all the other repetitions, it seems like every 8 characters repeat two times, every 32 characters repeat four times, and every 512 characters repeat at least four times. This means that all of address lines 3, 5, 6, 8 and 9 must be failing.

So to organize, every read from vram appears to have the following address (lower 10 bits, MSB first. X indicates correct bit, 1 or 0 indicate that this bit is always in said state):
00X10X0XX0

I got the failing bit values by examining the actual correct screen for the first 128 characters and their byte positions:

16d bytes of data read from offset 00d/32d/64d/96d: "__lleeII__lleeII" derives from data at actual offset 64d "__leetIRC_v1.1__"
16d bytes of data read from offset 16d/48d/60d/112d: "__uurr____uurr__" derives from data at actual offset 60d "_Query_focus:_$w"
and so on...

I aren't sure about the internal monitor connector pinout for the Compaq portable, but with the IBM 5155 you can use any card with NTSC composite output. As said, this may not be the case with the Compaq, and this may have to be investigated further.

Mike Chambers
May 11th, 2010, 02:01 PM
Regarding the video memory, it looks like some sort of addressing error. Since all characters are displayed in pairs, Address line 0 must fail at some point. In addition, based on all the other repetitions, it seems like every 8 characters repeat two times, every 32 characters repeat four times, and every 512 characters repeat at least four times. This means that all of address lines 3, 5, 6, 8 and 9 must be failing.

So to organize, every read from vram appears to have the following address (lower 10 bits, MSB first. X indicates correct bit, 1 or 0 indicate that this bit is always in said state):
00X10X0XX0

I got the failing bit values by examining the actual correct screen for the first 128 characters and their byte positions:

16d bytes of data read from offset 00d/32d/64d/96d: "__lleeII__lleeII" derives from data at actual offset 64d "__leetIRC_v1.1__"
16d bytes of data read from offset 16d/48d/60d/112d: "__uurr____uurr__" derives from data at actual offset 60d "_Query_focus:_$w"
and so on...

I aren't sure about the internal monitor connector pinout for the Compaq portable, but with the IBM 5155 you can use any card with NTSC composite output. As said, this may not be the case with the Compaq, and this may have to be investigated further.

sharp eye, per. looks like you're right. i thought it may have been some sort of addressing issue as well because of repeating patterns. speaking of the monitor pinout, i read that the monitor in there takes standard TTL input like any CGA card puts out so i think if i find or figure out the pinout for the internal connector i can just stick my hercules CGA clone card in there and wire it in. at least i could do that until i get a new compaq CGA card.

the internal connector going to the monitor in there has 9 wires on it, just like a CGA output as 9-pins. this is the standard CGA pinout:


Pin Name Description
1 GND Ground
2 GND Ground
3 R Red
4 G Green
5 B Blue
6 I Intensity
7 RES Reserved
8 HSYNC Horizontal Sync
9 VSYNC Vertical Sync

http://pinouts.ru/connectors/dsub9f.gif

considering the functions of each wire, do you think doing a little trial and error with the herc CGA card on that monitor could damage the monitor/card? i actually already ordered a new compaq card last night, but it'll be a good week until i get it. i'm very impatient. btw, i do know for sure which is the ground write on the connector so that one won't need any trial and error. i just wonder if one of the sync lines going into the color/intensity or vice-versa could cause damage.

Floppies_only
May 11th, 2010, 11:14 PM
considering the functions of each wire, do you think doing a little trial and error with the herc CGA card on that monitor could damage the monitor/card? i actually already ordered a new compaq card last night, but it'll be a good week until i get it. i'm very impatient.

Without documentation from Compaq you can't know which wires leading to the monitor are which. I'd wait a week.

BTW, it seemed to me that these machines often had failures of the motherboard. It might be a good idea to treat it gently and not use it when the temperature is high.

You might be able to fill the hole left by the missing second drive with paper. Get a piece of "mactac" from the copy store and cut it so it will stick to the inside of the case surrounding the hole. Then take a piece of black paper from the art store and cut it to the size of the hole. Put the black paper onto the sticky side of the mactac and the mactac inside the case. It won't stop radio frequency interference from leaking out of the case, but it will maintain the internal airflow the way it's supposed to go.

I wonder if the power supply creates enough current for a hard drive?

Sean