• Please review our updated Terms and Rules here

Commodore PET 2001-8 stuck on garbled screen

I think we have seen quite a few reasons to trust the G2 74154, particularly looking at SEL8 and nSEL8 but it cannot hurt. A good opportunity to work on dual trace skills. With a NOPPER... put one channel on A15 which is exposed on the NOPPer as START (or STOP) header and another on a G2 output. When I fixed my PET I did this... it's very satisfying seeing the traces line up and such a good test I built NOPulation into my ROM/RAM board design.

I'll fade into the background in anticipation of G4 being replaced.
 
daver2 said:
It doesn't really matter. I am more a fan of the oscilloscope though, as you can tell much more than with the logic probe. Especially with a two channel oscilloscope, because you should be able to observe the 'pulse' moving 'up' the decoder.

Dave
Click to expand...
I like the scope. It's nice to be able to see what's going on. I will do these tomorrow.
 
Nivag Swerdna said:
https://forum.vcfed.org/index.php?threads/petunias-repair-log.77296/page-6#post-944252 has my efforts from 2 years ago! :)

NOTE: Slow 20ms/div timebase

PS

"You were suppose to use pin 1 as the reference but it does look like it was fine. It is a little harder to tell since you didn't follow instructions" was the reply I got.... shows times don't change! ROFL
Click to expand...
I think I have been guilty of not following instructions. 🤔😆
Thanks for the reference scope images.
I will try the insert option with txt when I upload next lot. Much easier to follow for all especially for me with my goldfish memory.
 
Just looked at my G2 again. Is that a date code of wk 11 1993 or is that number for some thing else. I will scope it tomorrow.
DSC_1810.JPG
 
Amazing if it was still being used in 1993. Chuck added my Pet to his data base when I bought it. No mention of previous owners unfortunately. I guess it's a miracle I have it after it ended up in a crusher. If only it could speak. Sn 1007205. It certainly is intriguing.
Who ever replaced it did an excellent job on the soldering. Looks the same as surrounding ics, no damage
 
On the topic of 74154 IC's, I have a large collection of them and have used them over the years. One thing I have noticed about them, in general, they appear to be a very reliable IC. I'm not sure why, but one guess is that the thermal and mechanical properties of the large IC body helps the die inside stay cooler and protected.

Any IC can fail of course,and I was surprised to see that in a Desperado PET repair, I think it turned out that a 74154 in one of his machines was defective, that is the only failed one I have ever seen.

By 1993, I would have guessed that the 74154 would be well out of production, and the 74LS154 would probably about to go. I think, for this National semiconductor 74154, the date code would relate to the -743 and it is a 1974 vintage part, rather than from 1993. The 9311 is probably not a date code. If you look in the TI TTL data book , in the cross reference section, you will find that 9311 is simply the Fairchild equivalent number of the 74154.
 
RetroGadgetMan said:
Just looked at my G2 again. Is that a date code of wk 11 1993 or is that number for some thing else. I will scope it tomorrow.
View attachment 1251561
Click to expand...
When you enter the garbled screen, (since you have the tynemouth board), hit the shift and the "clr home" button, and see if half of the screen flickers. If it does, just hit it repeatedly until the screen clears itself out. I had a tynemouth board with a similar problem, and doing that fixed it. If it doesnt flicker id also probe around the sn74ls157n chips that are right above the cpu, (there should be three of those), and them being faulty is known to throw a wrench in the video ram circuitry
 
Hugo Holden said:
On the topic of 74154 IC's, I have a large collection of them and have used them over the years. One thing I have noticed about them, in general, they appear to be a very reliable IC. I'm not sure why, but one guess is that the thermal and mechanical properties of the large IC body helps the die inside stay cooler and protected.

Any IC can fail of course,and I was surprised to see that in a Desperado PET repair, I think it turned out that a 74154 in one of his machines was defective, that is the only failed one I have ever seen.

By 1993, I would have guessed that the 74154 would be well out of production, and the 74LS154 would probably about to go. I think, for this National semiconductor 74154, the date code would relate to the -743 and it is a 1974 vintage part, rather than from 1993. The 9311 is probably not a date code. If you look in the TI TTL data book , in the cross reference section, you will find that 9311 is simply the Fairchild equivalent number of the 74154.
Click to expand...
Didn't the failed one in desparados machine die from a catastrophic power event on one of the voltage regulators?
 
Mike Brixius said:
Didn't the failed one in desparados machine die from a catastrophic power event on one of the voltage regulators?
Click to expand...
That seemed likely, due to the number of failed chips, including one that had a very rare case of two inputs on a TTL gate being shorted together, suggesting some sort of melt down inside the IC.
 
Hugo Holden said:
On the topic of 74154 IC's, I have a large collection of them and have used them over the years. One thing I have noticed about them, in general, they appear to be a very reliable IC. I'm not sure why, but one guess is that the thermal and mechanical properties of the large IC body helps the die inside stay cooler and protected.

Any IC can fail of course,and I was surprised to see that in a Desperado PET repair, I think it turned out that a 74154 in one of his machines was defective, that is the only failed one I have ever seen.

By 1993, I would have guessed that the 74154 would be well out of production, and the 74LS154 would probably about to go. I think, for this National semiconductor 74154, the date code would relate to the -743 and it is a 1974 vintage part, rather than from 1993. The 9311 is probably not a date code. If you look in the TI TTL data book , in the cross reference section, you will find that 9311 is simply the Fairchild equivalent number of the 74154.
Click to expand...
That all makes sense.
 
genobombino said:
When you enter the garbled screen, (since you have the tynemouth board), hit the shift and the "clr home" button, and see if half of the screen flickers. If it does, just hit it repeatedly until the screen clears itself out. I had a tynemouth board with a similar problem, and doing that fixed it. If it doesnt flicker id also probe around the sn74ls157n chips that are right above the cpu, (there should be three of those), and them being faulty is known to throw a wrench in the video ram circuitry
Click to expand...
Thanks for the info. My Tynemouth board is working fine on my other pet.
 
Ok so G2 not much going on. 🤔 I expected to see more activity. I have only taken pics of pins with waveforms. The ones I have omitted are flatlined and move up just shy of 5 segments on the scope.

Pic of board to show what is socketed for this test.
16736971181778877400711553916254.jpg_20230114_114518.JPG_20230114_114502.JPG
 

Attachments

  • _20230114_114551.JPG
    _20230114_114551.JPG
    137.6 KB · Views: 2
You will need the NOP generator in to check the 74154. Having read what others have said, it seems like it would be an unlikely failure of the chip. Apologies for sending you down a rabbit hole.
 
In order to test the 74154 (as eight088 has stated) you need the NOP generator running.

This should result in all of the 16 outputs being driven in sequence as the CPU continuously 'reads' instructions from the entire memory space.

Dave
 
daver2 said:
In order to test the 74154 (as eight088 has stated) you need the NOP generator running.

This should result in all of the 16 outputs being driven in sequence as the CPU continuously 'reads' instructions from the entire memory space.

Dave
Click to expand...
Ah ok.
 
You must log in or register to reply here.
Back
Top