• Please review our updated Terms and Rules here

Help for Commodore 8296

emilen72

Member
Joined
Mar 26, 2013
Messages
17
Location
Terni, Italy
Hi, I have got a PET 8296 in very bad state (recovered from storage near sea) and attempt to fix it.

I have put socket on all chip and installed all good ones.

Original falty chips are a lot, some TTL, CPU, CRTC, both PAL and Kernal\Basic rom

The only good was the Edit and Char rom

I burn PAL substitute on Eprom and adapter, K/B rom and even the Edit and Char so do not use the original ones...

At boot I got the crisp sound but after the PET is stuck on a pattern screen... (I have build on breadboard a CVBS adapter, the result is a bit left shifted but is still readable)

So I burn the PETTESTER v.4 adapted to PAL 50Hz and it loop at the first VDU test.

I then got on screen an alternated patterns with half good test result in upped half screen and shifted second half good in the lower half of screen... see the image for more details :-(

I retested all chip and seem to be all good at the T48 programmer TTL test.

Any idea how to look forward? maybe the PAL UE5 and UE6 are not good (I have tested both re-reading these mounted on "pal to eprom adapter" with T48 as 27C512 and seem to be ok)

Tnx in advance to all for any suggest

BR

Emiliano
 

Attachments

  • IMG_7386.jpeg
    IMG_7386.jpeg
    2.5 MB · Views: 12
  • IMG_7387.jpeg
    IMG_7387.jpeg
    1.6 MB · Views: 14
This looks like the first PETTESTER screen (for the video RAM) that is not writing and/or reading correctly.

On a conventional PET this is separate static RAM. On the 8296 it is DRAM.

How have you tested the DRAM devices?

It may not be the DRAM of course, it could be the DRAM address multiplexers or a DRAM refresh problem.

Just brainstorming the possible areas.

Dave
 
Thank Dave for your reply….
I tested the dram chip on my xeltek superpro programmer…

I will check on the suggested areas

I wish to know the reason why the pattern are alternate…

Tnx
Emiliano
 
They are all in blocks of 8. This implies something with addressing to me because 2 ^ 3 = 8.

The characters that are there look to be correct (ignoring the shift of the image to the left).

The inverse '@' characters are wrong, and appear in a consistent block of 8 consecutive characters.

An inverse '@' character is the hexadecimal value $80.

More than that I can't tell.

Dave
 
Back
Top