CBM PET 3032 STRANGE BOOT

[Desperado]

You should have six (6) readings. Why only four?

It doesn't matter about the state of the machine. The frequencies and voltage levels should be constant and independent of the state of the machine.

Dave

Sorry!

E3:
PIN 10-11: 500,2 KHz

E4:
PIN 10-11: 125 KHz

E5:
PIN 10-11: 31,2 KHz

 

[daver2]

They look OK.

Let's continue this tomorrow.

Dave

 

[Desperado]

They look OK.

Let's continue this tomorrow.

Dave

Ok thanks! GoodnighT!

 

[Desperado]

Now stuck with this error

 

[Desperado]

Please Daver, what can i do now? Time to logic probe?

 

[daver2]

I am still not seeing data points...

When the PETTESTER locks up in the process of testing (i.e. locks up without displaying the "mem error" message) this is of no use to us. Press the PET reset and ignore the lock up.

When the PETTESTER produces the "mem fail" message (as in the post above), note down the 4-digit hex address (01ff in the above example) and the next 2-digit hex data value (80 in the above example). Then, reset the PET again and keep doing this to get approximately ten (10) valid address / data pairs.

From that we should be able to deduce something (may be)...

There is no point in getting the logic probe out if we don't know what we are looking for - unless we are just randomly poking about in the hope of divine inspiration...

Dave

 

[Desperado]

i i still got the 01ff error but the i2 ram was already changed

 

[Desperado]

if I reset just after the error, the pettester restarts but remains stuck at the first part...if I then turn off the computer and wait a few minutes, when I turn it on again it starts again....

 

[daver2]

So, from that post (#68) I conclude that the problem is either temperature related or mechanical and is exacerbated as the temperature increases.

So, how were you using the freezer spray? Were you just liberally spraying it everywhere, or targeting an IC at a time and ensuring the silicon inside was well and truly cooled down?

Dave

 

[Desperado]

or targeting an IC at a time and ensuring the silicon inside was well and truly cooled down?

in this way! The strange is that i have always same error 01ff....

 

[daver2]

01ff is the address. Let me check what limits the memory test starts at.

That's interesting... The memory testing starts at address 0200 - so 01ff is outside of the test range.

I am about to go out for the evening - but I suggest removing the PIAs and VIA if they are in IC sockets to prevent any possibility of an accidental interrupt from occurring.

Dave

 

[Desperado]

I am about to go out for the evening - but I suggest removing the PIAs and VIA if they are in IC sockets to prevent any possibility of an accidental interrupt from occurring.

I already removed them a while ago Dave!
Have a good night Dave...cold evening in Milan today

 

[Desperado]

I am desperate!! This Pet is devil possessed

 

[daver2]

Do you observe video issues as well?

If so - I would be looking at data buffers E9/E10.

If not - I would be looking at data buffers I10/I11.

I would start off by piggy-backing each buffer in turn (one at a time).

These errors are the worst - as they are potentially glitches that you can't get a trigger on. In the old days they would have just thrown the board away and replaced it with another one!

Dave

 

[Desperado]

Do you observe video issues as well?

Yes, sometimes i see video issues on screen! ok I try first with piggy on E9/10 Thanks!

 

[daver2]

Try both the piggy-back method AND some freezer spray (but not at the same time!).

Dave

 

[Desperado]

I am desperate!!!! i have fault also with piggyback and after also with spray freezer on ue9 and ue10

 

[Desperado]

Time to remove ue9 and ue10 and testing with logic tester?

 

[daver2]

Yep.

But I would replace the buffers. A 'glitchy' suspect device will not show up on a logic tester...

Dave

 

[Desperado]

First Ue9 and after Ue10?