• Please review our updated Terms and Rules here

8032-SK Restoration

It might be UD4 of course but looking at pins 1,2 & 13, they all have activity but they dont ever seem to be high at the same time at any point which means that the output on pin 12 will always be high, which it is.

Noteable that pin 13 appears to be high only when pin 1 is held low in the blanking phase.

Need more time to study the signals to make sure I'm not getting it arse about face but won't get back on it until Tuesday.
 
Noteable that pin 13 appears to be high only when pin 1 is held low in the blanking phase.
Gary,
That does not sound right. Pin 13 is related to to Display Enable and should be high most the 50 Hz vertical display time EXCEPT for the blanking periods.

Wait a minute, are you still running the NOP Generator? If so, the CRTC will not be initialized. You need to PETTest or your ROMs. Check for proper 50 Hz Vertical and 20 KHz Horizontal timing.
 
Grrrr

I decided to buy a NOP as it was only £4 and has LED's for NMI, IRQ,RST but !!!!!!!!!!!!!!!

They forgot to cut off the data pins, so while the resistors were connected, the data pins were also connected to the motherboard !!!!!!!!!!!!!!!!

Pins cut and now I have nice wave forms

Laser Business systems ROP generator more like :)
To be fair, I had the same issue but they emailed me to tell me. Did they not email you?

Colin.
 
Gary,
That does not sound right. Pin 13 is related to to Display Enable and should be high most the 50 Hz vertical display time EXCEPT for the blanking periods.

Wait a minute, are you still running the NOP Generator? If so, the CRTC will not be initialized. You need to PETTest or your ROMs. Check for proper 50 Hz Vertical and 20 KHz Horizontal timing.

I am away from the machine for now and might, as I said, have it arse about face but it did seem as if it was wrong on initial look but I wanted to leave it for a few days and have a look at the drawing before I go back and have a look.

And the NOP isn't installed, just the standard ROM. Will probably try PETTEST next time to give a more repeatable video signal.
 
Last edited:
looking at pins 1,2 & 13, they all have activity but they dont ever seem to be high at the same time at any point which means that the output on pin 12 will always be high, which it is.
I took a quick look on my 8032 and saw the same thing as you did. I think we are being tricked because, with the welcome screen, there is only a tad of real data at the beginning of the 50/60 Hz vertical rate and I was syncing the scope at the 20 KHz display enable rate.

For this test, we may need a full screen of repeatable data with a lot of pixels on.
 
Cheers for checking.

I will have a look at cobbling something, but PETTEST would give an initial full screen so on Tuesday (bloody work !!! four 12hr night shifts) I will give that a go.

I also might just whip out that IC and test it out of circuit just to confirm if its a dud.
 
I have one of them but due to a faulty socket, I had already removed and replaced the 6502 socket with a turned pin one anyway. I have still added an extra socket on top in an attempt to protect the turned pin one I soldered in because I don't fancy ever having to get it out.....
Colin.

I agree

This is the issue I have in mind when replacing IC sockets in vintage computers.

Probably the better quality socket is the machine pin type, especially the original Augat ones, the new clones are almost as good. And I use these in the case that it is likely the IC I put in them might only be once or perhaps removed 1 or two times ever.

But if I think there could be a situation in the future where I might want to take the IC in and out multiple times, for example RAM sockets since these have a penchant for failure, I use a dual wipe socket instead and I also apply lubrication to it (Inox mx-3) to reduce wear on the socket.

The main reason is that if something happens to the dual wipe socket, they are very easy to remove as the pins can either be removed one by one from the top, and very easy to free with the solder sucker to remove the whole IC socket in one go.

The "problem" with the machine pin socket is that the round cross section of their pin uses up a greater area of the plated through hole and they are not as flexible, so after solder sucking, sometime the side of the pin can still remain stuck to part of the perimeter of the plated through hole, risking the hole when the socket is extracted. Also because of the machined step on the pin, there can be cases where the pin remains still soldered to the top of the pad on the boards top surface in places after solder sucking, risking lifting that. And also because the socket plastic is very brittle, it puts too much force on pads and tracks attempting to cut the plastic for individual pin removal.

So the bottom line, is it is better not to have to remove a machine pin socket. This is the reason I often use the dual wipe, but not always.
 
End of the day, I'm still not getting video :)

Next time I think I will pull the UD4 and test it, should only take 10 minutes as the PET with its nice big holes makes sucking easy (!)
 
Cheers for checking.

I will have a look at cobbling something, but PETTEST would give an initial full screen so on Tuesday (bloody work !!! four 12hr night shifts) I will give that a go.

I also might just whip out that IC and test it out of circuit just to confirm if its a dud.

I was tricked. I tried it with 2000 CHR$(255) characters on the screen, and got good video data on UD4-pin12. If you don't, you have a dud IC.
Yes, PETTEST should give a good screenful of pixels or at least lots of pulsing on the video data signal.
 
Last edited:
So I looked again and I can see video on pin 1, no row on 2 but pin 13 is always low when video is present. Put pettest in and got this !

Video
 
At least the screen works and is nice and bright. With the standard rom, I do get the startup bleeps, any ideas ?
 
At a very quick glance - that definitely looks like a hardware or CRTC fault.

I will have a look at the video on a better machine later. I can't view it from the Company laptop (!) and the phone is useless...

It does look reminiscent of a recent (similar) fault I have observed though...

Dave
 
Tried a differed 6845 CRTC and got the same result. Both are from a BBC and are the Htachi version
 
interestingly, not seeing any activity on the video bus ESD0, there should at least be something attempting to write to bit 0 ?

need to check the buffer.
 
If you are using my PETTESTER then yes, there should be plenty of activity on ESD0...

I think you are on the right track. Check that the signals that should be constant are present and stable (e.g. SA0-9, VIDEO LATCH, CLK1B, /CLK1 etc.).

From the video it looks like that you have a stable picture (that is the H and V drives are OK - meaning that most of the timing logic is OK as well) - and that the problem is related to the VIDEO signal.

I remember what the last issue was now - the data bus buffers between the CPU and the VIDEO RAM were enabled - so that activity on the data bus was 'bleeding' into the video circuitry.

So, yes, ESD0 might be your way to crack this.

CORRECTION: There will be NO activity on ESD0! Let me explain. PETTESTER writes $00, $01, $02, ..., $FE, $FF into consecutive video locations. With the ODD/EVEN RAM scenario this means that $00, $02, $04, ... gets written into EVEN RAM with $01, $03, $05, ... written into ODD RAM...

Dave
 
Back
Top