Pet 2001-8 monitor - Any checks I should do before I try powering it?

[daver2]

Excellent. Well done.

What I was going to say is that just because A1 is counting doesn't mean that it is 100% working - only 99% working!

When a NEXT pulse is received (on A1 pin 2) the outputs from the counter should all be reset to 0.

If A1 is just 'freewheeling' and not resetting, then it is possible that the counter starts off at a count of something other than 0. This means that the first line of the displayed character will be part way through the character (exactly as you are observing). Since each line of characters are 8 video lines high, and the counter is 3 bits (a count of 8) then everything will be nicely synchronised, but not exactly what we want.

The output from the counter (B1 pin 8) will operate when the counter is 7 and the display has either just turned on (of off) - DIS.ON. This signal is used to operate E6 (in readiness for the next line of characters).

I suspect that A1 is not being reset when commanded to by pin 2.

I was going to coach you in displaying the critical signals on your oscilloscope - but there appears to be no need now.

I will be interested in your investigation.

Dave

 

[arcaic]

@daver2 You got it bang on.

As per your suggestion I started to look at A1 in more detail. In Hugo's PETREPAIR1.PDF on page 22 there is a scope image showing RA9 which on the PET 2001-8 is A1/11 so I set this up in the FALSTAD simulator using |Video ON as a trigger and then compared them to my PCB.

Like a fool I don't have an exact capture of that exact scenario but the following is the same but at different points....


Hugo's scope image...

Here is the simulator...


In both Hugo's scope capture and the simulator you can see that when triggered by NOT V ON RA9 has two high pulses of equal length followed by a very narrow one but on my board that very narrow one wasn't happening.

I looked at all the inputs to A1 (where RA9 is generated) and the reset on PIN2 seemed fine and the other reset on PIN3 is pulled high all the time yet the counter did not seem to reset.

At that point I thought given an 74LS93 is only a quid or two I'd swap it and see what happened and it fixed it.

I'll see how it behaves with the reset link off now and prob just check all the sync signals and timing but given how the NEXT signal seems to pretty much control EVERYTHING and that's partly driven by RA9 and feeds back into itself I'm hopeful they are all behaving now.

Thanks again everyone.

 

[daver2]

No problem.

An intetesting fault though...

It is these type of strange faults that keep us interested in these repairs. It enhances our knowledge. Hopefully you have gained a few new skills in the process?

Dave