HP9836C colour alignment (grey scale tracking)

[Tony Duell]

[...]

> read/write to the first byte (I think) of the 256 byte sector buffer on 
> the card. As soon as I spotted this was implemented with 2114s [1]s, I 

I forgeot to expand that footnote, so here it is : 

[1] The PCB has overlapping 16 and 18 pin IC positiones, for either 2112 
(256*4) or 2114 (1K*4) SRAMs. In the latter case 2 of the address lines 
are grounded, so only 256 locations are used. Presumably HP fitted 
whichever chips where cheaper/more avaialble at the time (I've seen 
boards with 2112s, I've seen them with 2114s. I've never seen one with 
one RAM of each type, but there's no logical reason why this won't work). 
Incidentally, this PCB is unusual (for HP) in that the silkscreen has, 
next to each IC positionm the last 4 digits of the HP part number _and_ 
the standard number (like 'LS164').

[...]

> The monitor on the bench, with the computer's motherboard on top of it. 
> Monitor cable plugged into the back of the motherboard. Video boards (and 
> brightness control cabled up in the normal way.

I also didn't mention that I had problems with the brightness control. 
It's quite simple, just a 500 Ohm (IIRC) pot with a thumbwheel glued on
the spindle, which can be turned through a slot in the bottom of the case. 
The pot is mounted on a little metal bracket screwed to the metalwork 
around the vidro PCBs. The 3 connections of the pot are wired to a socket 
which fits onto a header plug on the graphics PCB -- why it connects 
there is , I assume, to make it easy to get to the plug, since all the 
grpahics PCB does with those 3 signals is route them to pins on a 
connector ot the text PCB under it, where they go into the DAC circuitry.

Anyway, when I got the machine, turning that thumbwheel did nothing. It 
turned out the potentiomenter's spindle was very stiff to turn, and the 
glue holding the wheel in place had failed. So turning the wheel simply 
caused it to slip on the spindle. There was also minor corrosion damage 
to casing screws in that area.

So I pulled the wheel off and remove the pot from its bracket. It's a 
physically small thing, a cylinder about 1/2" in diameter and 1/2" long, 
with tags on the face opposite to the spindle. The 'top' -- the face 
where the spindle came out, was held down by bent metal tabs, covered 
with some kinder of selant. Most of that was scraped off, the remainder 
removed by heating it with a soldering iron, and the tabs bent up. 

After removing the top, the 'wiper' was seen to consist of a carbon bock 
backed with a spring wire fitted into a plastic moulding on the end of 
the spinde. These parts were removed and the spindle slid out from the 
top. And yes, there was minor surace corrosion here. Cleaned up all the 
parts and put it back together. The pot tested correctly electrically, 
and turned a lot more freely. 

I was not happy about glue the thunbwheel back on, though. I don't trust 
glue. The thumbwheel consists of a disk with a tube moulded onto it, the 
potentiometer spindle going into this tube. The outside of the tube was 
slightly tapepred, so the first thing I did was machine it straight. I 
then made a brass collar to fit over this part of the wheel, with a 
radial hole tapped for a grub screw. A clearance hole for said screw was 
drilled i nthe tube and finally a flat milled on the spindle. That's not 
going to slip!.

IMHO that's how it should have been made in the first place. WHy HP cut a 
little corner here in otherwise well-designed and well-made machine I 
will never know.

-tony