Perq 2T1 display fault

From: Tony Duell <>
Date: Sat Feb 21 16:38:41 2004

> On Sat, 2004-02-21 at 03:48, Tony Duell wrote:
> > > So, the 2T1 has a standard KME as you say - it looks to have correct
> > > horizontal and vertical drive, with the exception that it skips every
> > > few lines, leaving a blank horizontal gap. I can see what look to be
> >
> > That does _not_ sound like a monitor problem. You're getting missing
> > video every few scan lines, right? In which case, either you've got some
> > Analogue signal that's blanking the display _and_ which is locked to the
> > vertcial sync signal, or you've got a problem with the memory board
> > (middle board in the cardcage) in the PERQ.
> Ahh, fair call - I was wondering along the lines of bad cap or otherwise
> in the vertical drive, such that it was the display itself which was
> producing uneven deflection (because I can only see vague character

I'd love to know what fault could essentially put stair-steps in the
vertical deflection waveform. That waveform is generated in an analogue
manner, by chaeging a cap connected to the TDA1170 chip, BTW...

Anyway, what I'd do at this point is try to get a grey raster from the
monitor. Unplug the video BNC cable, leave the other cables connected (so
the monitor gets the right scan frequencies, etc. Turn up the brightness
(and if necessary, tweak the sub-brightness pot on the monitor PCB -- I
can find out which one that is if it's not obvious). Look at the screen.
You're aiming for an even, grey, rectangle. If you get that, the
deflection sside is fine. If you get black lines across it, then I am
going to have to think very carefully...

> shapes I can't tell if I'm actually missing data, or if there are blank
> lines effectively inserted into the display by the monitor itself)
> I am not a display expect by any means! :)
> > How far does it get through the self-tests? What's the final number on
> > the DDS display (front of the CPU box) when it tries to boot?
> It gets to 999, which I assumed was normal - but maybe not! Maybe the

That's right for POS (which is what I left on the machine when I last
fiddled with it). 255 is right for PNX, 400 for Accent (IIRC). Unless
there's a really strange problem, I think we can assume the PERQ is
booting properly.

OK, that means the PERQ's CPU can access memory, and that the memory has
passed the self-tests. So there's not a RAM problem. But there could be a
video controller problem.

Once we know whether the monitor can give a plain raster, I will dig out
the appropriate schematics (memory board (that's where all the video
stuff is ) or monitor and see what I can deduce

Received on Sat Feb 21 2004 - 16:38:41 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:36:43 BST