• Please review our updated Terms and Rules here

CBM PET 3032 STRANGE BOOT

I have been over the issue of damaged sockets many times.

You cannot tell the condition of the socket, in most cases, just by looking. It requires an IC pin on a small wire handle to feel the tension of individual socket claws.

The worst types of sockets are single wipe and the TI sockets that grab the pin from side to side.

The type of socket you have there are normally pretty good, being a dual wipe type. These have an interesting feature, it is possible to replace just one claw at a time, because of the way they were manufactured, the claw comes out of the socket top. So if you are careful you can solder a wire to the claw on the top. The heat the pin on the solder side of the pcb, making sure it is fully melted first, and withdraw just the one claw assembly and replace it with one from a donor socket.

But I would not consider replacing a socket, unless one or more of its claws failed the test for spring tension with the test pin on the wire handle.

The other thing is, never put anything into an IC socket that is thicker than a standard IC pin which is only about 0.25mm to 0.3mm, thick and just under 0.5mm wide. For this reason the test pin must also be inserted on the correct axis.
 

Attachments

  • testpin.jpg
    testpin.jpg
    199.9 KB · Views: 4
Last edited:
Hugo Holden said:
I have been over the issue of damaged sockets many times.

You cannot tell the condition of the socket, in most cases, just by looking. It requires an IC pin on a small wire handle to feel the tension of individual socket claws.

The worst types of sockets are single wipe and the TI sockets that grab the pin from side to side.

The type of socket you have there are normally pretty good, being a dual wipe type. These have an interesting feature, it is possible to replace just one claw at a time, because of the way they were manufactured, the pin comes out of the socket top. So if you are careful you can solder a wire to the claw on the top. The heat the pin on the solder side of the pcb, making sure it is fully melted first, and withdraw just the one claw assembly and replace it with one from a donor socket.

But I would not consider replacing a socket, unless one or more of its claws failed the test for spring tension with the test pin on the wire handle.
Click to expand...
Now i already changed this socket but thanks! I'll keep the advice in mind for next time!
 
Put your PETTESTER ROM back in the new UD8 socket (very carefully...) and set Nivag's ROM links all to OFF (both E and F ROMs if they are not already OFF) and see if that works now with PETTESTER...

If not, we are going to have to check for continuity between all of the pins of the ROM sockets...

Note: I have also covered previously in a repair thread potential display issues (related to the floating GRAPHICS) signal if the VIA is removed.

All these 'pearls of wisdom' should be written down in a notebook and referred to in the future...

Dave
 
daver2 said:
Put your PETTESTER ROM back in the new UD8 socket (very carefully...) and set Nivag's ROM links all to OFF (both E and F ROMs if they are not already OFF) and see if that works now with PETTESTER...
Click to expand...
Yes Pettester for now is working...
 
Give it a thorough test (say 20 DRAM passes again).

Are all the chips (e.g. PIAs, VIA etc.) back in the machine?

Obviously the PETTESTER EPROM is in UD8 and not the 'correct' EDIT ROM at this stage.

Dave
 
Ok,

So switch Nivag's 'E' ROM link to ON (to run the PETTESTER code from Nivag's board) and repeat the PETTESTER test, but for a longer period (say 30 passes).

Dave
 
daver2 said:
Ok,

So switch Nivag's 'E' ROM link to ON (to run the PETTESTER code from Nivag's board) and repeat the PETTESTER test, but for a longer period (say 30 passes).

Dave
Click to expand...
Ok thanks Dave, i ll do this tomorrow! Goodnight Sir!
 
Goodnight.

We need to 'flush out' whether this is a ROM issue or a RAM issue. Running Nivag's PETTESTER code should find a fault if it is a RAM issue...

Dave
 
If @Desperado has the technology to make ROMs then we could always burn PETTESTER into a ROM using the version that works at F800; that would eliminate the editor socket for a while?

PS
I should really add that to my board if I ever make any more.

There is an alternative, more primitive, tester on the board but it is more designed for PET 2001 8K... Bank OFF OFF OFF ON; https://github.com/vossi1/2001test
 

Attachments

  • PETTESTE2KF04.zip
    1.2 KB · Views: 3
Last edited:
daver2 said:
Ok,

So switch Nivag's 'E' ROM link to ON (to run the PETTESTER code from Nivag's board) and repeat the PETTESTER test, but for a longer period (say 30 passes).

Dave
Click to expand...
I am desperate! We are at dram step number 30 but the computer does not crash.... what causes the crash then? :(
 
This probably indicates that it is not the DRAM that is at fault but the PETTESTER EPROM or the IC socket that the PETTESTER is sitting in or (more likely) some PCB tracking connected to the UD8 socket.

What ROM sockets have been changed on this board (apart from UD8)?

Dave
 
daver2 said:
This probably indicates that it is not the DRAM that is at fault but the PETTESTER EPROM or the IC socket that the PETTESTER is sitting in or (more likely) some PCB tracking connected to the UD8 socket.

What ROM sockets have been changed on this board (apart from UD8)?

Dave
Click to expand...
UD9,8,7 and 6
 
Because for safety I wanted to replace the old white sockets, at least these main ones..
 
If you put the EDIT ROM back into UD8 and it 'works' what does that actually tell you?

If it is the EPROM that you burnt the PETTESTER code into - then all well and good. However, if it is something else - all that will happen will be that the PET may be unreliable in operation and crashes.

Dave
 
You must log in or register to reply here.
Back
Top