HP9836C colour alignment (grey scale tracking)

Tony Duell ard at p850ug1.demon.co.uk
Tue Sep 30 17:24:10 CDT 2008

> > But what worries me is a comment in the HP service manual. It basically
> > warns you not to twiddle these adjustments (the same sort of message is
> > printed on the metal cover over the monitor chassis).
> I lost a great deal of trust in their stuff some time ago when I ordered 
> a "service manual" for some H-P product and received a very small bundle of 
> paper that was shrink-wrapped,  and on opening it I found "The monitor is 
> normally not repairable,  but is replaced as a unit..."  I sent it back and 

Alas, yes....

Very few HP classic desktop computer service manuals contain scheamtics. 
You might, if you are lucky, get PSU schemaitcs, but not for the 
processor, etc, sections. I am, of course, trying to remedy this...

As an aside, I feel it should be illegal to sell a 'service manual' for 
an electronic device that doesn't contain full scheamtics. Those are 
essential for servicing/repair. Of courwe I'd have no objection to 
companies selling 'boardswapper guides' provided they don't misrepresent 
them by calling them 'service manuals'.

Anyway, the HP9836 'service manual' is a boardswapper guide, and contains 
very little information that's not obvious from 10 minutes of looking at 
the machine. There's a large section on running the diagnostics, with 
what to do for each error, and alas the latter is 'repalce the <foo> 
board'. And very often those error messages give a lot more information 
once you've learnt to interpret them

The only really useful section is the pinouyts of all the motherboard 
connectors (this series of machines has amotherboard in the bottom of the 
case which contains the keyboard interface and HPIB circuitry. The 
keyboard connects to an edge connector on the side of the motherboard, 
the CPU, floppy controller and expansion backplane plug into edge 
connector sockets on the motherboard. The video system consists of 2 PCBs 
-- in the 9826 the text board plugs into the motherboard, the graphics 
board plugs onto the top edge of the text board (and extends forwards 
over the internal monitor), in 9836s the 2 video boards connect to ribbon 
cables on the front of the motherboard). Anyway, know the location of 
things like the address bus was a great help!

Be warned there are some errors in those pinouts, though. The main one 
that I've found relates to the 'video coax' in the 9836C. In all 9836s 
(monochome and colour) there's a special ribbon cable consisting of 4 (75 
ohm?) coax cables bonded together. It's terminated by 8 pin sockets at 
each end. One end goes onto a header plug at the rear of the motberboard 
(which is connected to the adjacent DA15 connector for the monitor), the 
other end plugs into the text video board under the disk drives. On the 
colour model, one of the cables isn't used, the other 3 are R, G, B. 
Anyway, it's clear that one row of 4 pins is the screens, the other row 
is the centre conductors, and it's equally clear that the latter carry 
the video signals. But the pinout in the manual shows the colour signals 
on 3 pins at one end of the connector, which would put one of them on a 
screen. It's obvious how this has happeend, it's the difference between 
numbering the pins 

1 2
3 4
5 6
7 8


1 5
2 6
3 7
4 8

The dismantling procedures in that manual are a little odd too. Anyone 
who can follow the procedure for removing and replacing the internal CRT 
of the 9826 and not end up dropping it or using choice language is a lot 
cleverer than I am. You're supposed to hold the CRT against the front 
bezel (from the inside) And fit 4 screws (2 of which are almost 
impossible to get to as components on the monitor PCB get in the way, as 
does some chassis metalwork) with a founding spring between 2 of them. 
Not thanks!. My method may take a little longer, but it's not stressful. 
Here it is (basically)

Remov top cover, board hold-down strip, PSU PCB, CPU PCB, floppy
controller and expansion backplane. Remove graphics board (on top of
monitor cover -- 3 screws) and text video board. Remove the internal
floppy drive (6 screws + LED and power cable). Remove metal plate under
floppy drive. Remove keyboard (5 screws on underside of machine, unplug
ribbon cable and unscrew earth wire at keyboard end. Unplug ribbon cable
and motherboard end and remove from machine. Unscrew power switch 
assembly from case, unclip and open cover, then release actuator from 
switch and remove the actuator/cover parts. Remove monitor cover (4 
quarter-turn fasteners), remove monitor to motehrboard ribbon cable 
jumper, discharge final anode to CRT ground (faston tab on monitor PCB), 
then disconnect anode cap. Unplug yoke, CRT base wiring and earth wire 
from monitor PCB. Undo  2 screw holding bezel to the bottom of the case, 
carefully tilt it backwards (mind that CRT!) and pull it fowards to free 
the 3 tabs.  Place the bezel on the bench fase down, remove the CRT 
socket, undo the 4 screws (note how the earthing spring is fitted) and 
take off the CRT moutning frame. Lift out the CRT -- NOT by the neck.

There's one curious thing about that servic manual. It contains no 
schematics at all. Not even of the mains side of the PSU, which is 
trivially field-repairable. Juat about the only electronic tests you are 
asked to make are the PSU outputs at labelled testpoints on the PSU 
board. It doesn't, for example, describe the use of the testpoints [1] and 
LEDs on the disk controller board. And yet it contains this complex 
procedure for setting up the grey scale tracking involving an instrument 
-- the photometer -- that few people are likely to own (many fewer than 
would on the equipment to do proper electrical tests.

[1] I think some of those are mentioned in the CE manual, which is 
similar to the service mnaul but contains a couple more useful bits of info.

> told 'em I wasn't going to pay that particular invoice as that was of 
> absolutely no use to me whatsoever.  :-)
> > And that misadjustment can cause various problems :
> > Visible flyback lines
> Yup,  if you crank it way up.

I _think_ the only misadjustment that will cause this is the A1 ('screen 
grid') adjustment on the flyback transformer. Maybe setting the clamps 
way too low will do it.

> > Missing colours
> If you don't have it turned up high enough.

Or if you have the clamps turned up too high, the electron cguns will be 
cut off for low, but non-zero input to the DACs.

> > Incorrect colours (compared to another 9836C)
> Either of the above.
> > Significantly decreased reliability
> Maybe if you crank them _all_ up?
> > Now, the first 2 are very ovbious, and I'll not have problems with that.
> > The third doesn't bother me at all, since I only have one HP9836C. But
> > the last does worry me.. Particularly if the part that fails is the
> > flyback transformer. The CRT seems to be a standard one (I can find no
> > data on it, but some CRT testers/boosters list it as one of the CRTs they
> > can handle). I would have a chance of finding a replacement. The
> > transistors on the video board are all standard, so I could find
> > replacements. But HP never supported component-level repair on these
> > machines, and the flyback transformer was only avaialbe as the complete
> > deflection PCB. And that's long unobtainable.
> I wonder if it would be possible to get from some aftermarket suppliers?  I 

Als I doubt it. IT was not a common machine, so I doubt anyone bothered 
to make a nrw flyback for it. Who would buy one? If you follow the HP 
manual, all you'll do is replace the complete deflection PCB, there are 
no tests to do to check votlages from the flyback, etc. So unless you're 
clueful you won;'t even know you need a new flyback transformer.


> > I don't if any other HP monitor used the same transformer -- the horizontal
> > scan rate is unusual -- about 25kHz -- but other HPs used the same sort of
> > rates. 
> The scan rate supported (with the flyback time being the proper portion of a 
> scan line and such) is not only determined by the transformer but also by 
> what all else is hooked up to it,  capacitors in particular.  And HV 

Sure. There is the obvious resonating capacitor connected to it.

> regulation ties into this too.

This monitor is a little unconventional in that there are 2 output stages.

The horizontal sync output from the computer goes through a couple of 
monostables (one of which is set by the horizotnal centring control), one 
of these monostables is held reset if the monitor's internal SMPSU 
outputs are incorrect, thus shutting down the EHT side. 

The output of that circuit drives the first power transistor. This has 
the horizontal yoke as its load (but not the flyback transformer). The 
flyback pulse from the yoke is detected and used to indicate that the 
yoke is doing something. This signal, along with a similar feedback 
signal from the vertical defleciton circuit are used to (a) turn on a 
green LED indicating that the deflection circuit is working and (b) 
enable the drive from the output of the monostable section to the second 
power transistor, which drives the flyback transformer. 

Now the power to that stage comes from a TIP122 transistor which is 
contolled by an op-amp circuit. One of the inputs to that comes from a 
potential divider connected to the EHT (25kV) supply, another comes from 
a voltqge reference circuit. This ,of course, is the EHT regulator 
circuit. A divider block on the side of the flyback transformer provides 
the A1 and focus supplies, other windings (brought out on the PCB) 
provide a -ve bias for the CRT control grid and a supply for the CRT heater.

Of course this information is not in the HP manual...

> I can't imagine H-P deciding that they were going to make their own flyback 
> for this one monitor as opposed to using something else that already existed.  
> Not impossible,  but it just doesn't seem too likely to me.

I think it's very liekly they did this (or rather got some company who 
made flyback transoformers to make one to their specification). My 
experience with working on monitors, terminals, etc, is that the flyback 
transformer is almost always specifci to that model. I say 'almost 
always' because Philips sold a few more general-purpose ones and those do 
turn up, e.g. in KME monitors (as used on PERQs, Whitechapel MG1s and so on).


More information about the cctech mailing list