• Please review our updated Terms and Rules here

Another PET 2001 motherboard needs your help...

dave_m said:
I suppose there could be a corroded/high impedance section of trace that is between the output and the loads. By jumpering to one of the loads, you eliminated that section.
Click to expand...
But then why doesn't the jumper work when pin 3 is in the socket, simply parallelling the jumper? Baffling indeed...

Now the dilemma: Does Phil tack a jumper from C8-3 to one of the other pins, close the lid and call it a job well done, or keep searching for truth and enlightenment ;-) ?

Another (probably unrelated) anomaly: if I pull C9-9 (which is not connected to anything) high on one of my working PETs it creates exactly the same symptom as Phil's, but not on his working PET or on my other one.

Maybe this saga will inspire Eudimorphodon to hang in there a little longer ;-)
 
MikeS said:
Now the dilemma: Does Phil tack a jumper from C8-3 to one of the other pins, close the lid and call it a job well done, or keep searching for truth and enlightenment ;-) ?
Click to expand...

That's a tough call. Remember the old adage: "Better is the enemy of good enough". ;)

Another (probably unrelated) anomaly: if I pull C9-9 (which is not connected to anything) high on one of my working PETs it creates exactly the same symptom as Phil's, but not on his working PET or on my other one.
Click to expand...
Well that's not fair as pin 9 is an output, but you are probably on the right track. It has something to do with the clock circuits. Personally I would replace the 74LS107 on C8 with a regular, more beefy 74107. I've never liked low power Schottky.

Also, I would like to look closer at the reset going to C8-13. Or Phil perhaps should tell us to go to hell and button up the box! :)
 
I think Phil should compare on the scope how the signal looks like when jumpered. Especially measured directly on C8 pin 3, not on D5 pin 9. If it is still weak (compared to the working PET), I would think that it is not a trace problem. Then I would assume a small variation in capacitance and/or resistance going through the wire instead of the trace helps the one weaker clock pulse to trigger D5. But the source of the problem would still be the weak clock signal for which we have to find a reason.

On the other hand, if it is a trace problem, the signal on the scope should look fine at its output C8 pin 3, and weak at D5 pin 9.

André
 
dave_m said:
I suppose there could be a corroded/high impedance section of trace that is between the output and the loads. By jumpering to one of the loads, you eliminated that section. With the jumper, does the clock signal look better?
-Dave
Click to expand...

As MikeS mentioned, I don't think this is the case here as I have also tried putting the pin back in the socket but still keeping the jumper wire and it does not work!!! crazy.

MikeS, to answer your other question, I'd prefer to keep going and try to get to the bottom of this problem, nothing will be learnt unless that is done, and besides, who wants a rare and classic 1977 Japanese PET 2001 motherboard with a chip with a leg sticking out and a jumper wire soldered to it...not me.

Worst case, I give up later.
 
so to recap, I should check:

• Signal quality of C8 pin 3 with jumper installed
• D5, D8, C1 inputs compared with working pet board.
• Reset going to C8-13
• D5, D8, C1 supply voltage.
• Signal quality on C9 pin 1, C9 pin 11 and D8 pin 8 (do I use DIS.ON as a trigger?)
 
Nama said:
MikeS, to answer your other question, I'd prefer to keep going and try to get to the bottom of this problem, nothing will be learnt unless that is done
Click to expand...

That's the spirit! But remember to blame Mike if your probe slips, you short out Vcc and your board bursts into flames. :)

Really, just take your time at this point. We do not want "Technician Induced Fault".
 
Nama said:
so to recap, I should check:

• Signal quality of C8 pin 3 with jumper installed
• D5, D8, C1 inputs compared with working pet board.
• Reset going to C8-13
• D5, D8, C1 supply voltage.
• Signal quality on C9 pin 1, C9 pin 11 and D8 pin 8 (do I use DIS.ON as a trigger?)
Click to expand...

This is a good plan.
-Dave
 
MikeS said:
Maybe this saga will inspire Eudimorphodon to hang in there a little longer ;-)
Click to expand...

*snicker* I've been watching this one with interest. I'll try not to lose heart.

Speaking of which, I do have an update on that. I have, finally, managed to wring some visible life out of the 4032. (It's still deep in a coma on life support, possibly with a broken spine, but it's *twitching*.)
 
Nama said:
• Signal quality of C8 pin 3 with jumper installed
Click to expand...

So far I can't see any noticeable difference. I'll keep looking

So here's a weird thing. With C8 pin 3 leg pulled out of the socket and a jumper running from this leg back to pin 3 of the socket that the leg was in, the machine works!!!

Stranger and stranger.

Phil
 
Could it be that the 30cm length of jumper wire is effecting the timing ever so slightly so that the boarder line pulse works?

I've tried a new 74LS107 in the socket and it also does not work. I also tried the new 74LS107 with the leg bent up and jumpered to see if the effect is the same as the original IC and it is. So, no change with a new 74LS107 installed.

I'll try tinning the new IC leg and see what that does.
 
Nama said:
So far I can't see any noticeable difference. I'll keep looking

So here's a weird thing. With C8 pin 3 leg pulled out of the socket and a jumper running from this leg back to pin 3 of the socket that the leg was in, the machine works!!!

Stranger and stranger.

Phil
Click to expand...

I think the effect (that one weaker pulse that we're chasing) is just a "side effect" of some capacitive/resistive effects on an already very weak pulse. That "side effect" may be triggered by minimum changes in e.g. resistance by using the wire instead of the trace or not.

But in my opinion the real problem is the generally weak signal. If the signal itself were strong enough, then those changes (e.g. using wire or not) would not have any effect at all.

Looking closer at http://web.me.com/lord_philip/other_computers/20012.html#50 where you compare the clock pulse between the working and the broken PET, you can see how weak the broken PET's signal actually is. If both have the same Volts/cm the signal is only 2/3 of a good signal. If it's 2V/cm the weak signal barely scraps the TTL "high" level, which could explain why some pulses go missing if an even very small
disturbance (like wire compared to trace) happens. If it's 1V/cm it's a wonder that the pulse is even registered at all.

Looking at http://web.me.com/lord_philip/other_computers/20012.html#61 where you compare the Q and /Q outputs of D8, we see that both signals are "weak". The positive pulse is weak as it barely goes over 1V (is it the same V/cm for both signals?) as we have seen before. But the negative pulse is also weak in that it does not even go below the TTL "low" threshold of 0.7V. And it does not look like a power problem, as the negative pulse starts from a high voltage level (assuming 1V/cm).

Can you repeat the measurements at http://web.me.com/lord_philip/other_computers/20012.html#61 and http://web.me.com/lord_philip/other_computers/20012.html#62 on the good PET? (and note the V/cm as well for both channels?). If my assumption is correct, the signal should be stronger and especially wider.

If that is so then it is a problem with the timing of the D8 input signals, not with something on the outputs.

André
 
I put a comparison vid between the PET board in working config (C8-pin 3 out and jumpered) and non working (C8-pin3 back in the socket).

http://web.me.com/lord_philip/other_computers/20012.html#65

There are some small differences. The peak voltage on the working setup looks slightly more, and the non working config seems to have more negative voltage below GND (is that even possible).

Also, I just wanted to confirm that these are the inputs I should check and compare between boards for extra load:

D5 - pins 1, 4, 8, 9, 11, 12
D8 - pins 1, 2,4 , 5, 9, 10, 12, 13
C1 - pins 1, 2, 3, 4, 10, 11, 12, 13

did I miss any?
 
Last edited:
..humm, I hope I haven't made a mistake. I think that is supposed to say C8, not D8 on my page for those images. I think I screwed up. I will check quickly.

UPDATE: Damn, really sorry about that. Thats a huge mistake. Yes it was supposed to be C8 not D8. I've fixed my page accordingly.

I'm just looking at the signals on the working PET for C8 pin 2 and 3 now.
 
Last edited:
Phil, that's great work! I especially like http://web.me.com/lord_philip/other_computers/20012.html#68 .Did you use some image manipulation to overlay those pictures?

But anyway, I think these pictures support my hypothesis that it is a timing problem.

Can you do this measurement again, but replace the by C8 pin 13? This is one relevant input pins for the C8 pulse output. Keep the C8 pin 2 signal, so we can use it as reference point for the signals.

I also would like to see how the C8 pin 12 and C8 pin 13 relate to each other. My assumption is that there is some different timing in the working and the broken PET.

Heads up, I feel we're getting closer...

André
 
fachat said:
Phil, that's great work! I especially like http://web.me.com/lord_philip/other_computers/20012.html#68 .Did you use some image manipulation to overlay those pictures?
Click to expand...

Yes, I used Photoshop. My day job is an Art Director, and many many moons ago I used to be a Photo-retoucher, so throwing a few layers together like this is fairly simple and takes about 10 mins.

So, just to confirm, you want me to do this again but comparing C8-13 and C8-2, also C8-13 and C8-12?

Phil
 
I've just posted the results, but I didn't use pin 13 as the trigger, hopefully this is ok.

Please take a look here:

http://web.me.com/lord_philip/other_computers/20012.html#69

here

http://web.me.com/lord_philip/other_computers/20012.html#70

and here

http://web.me.com/lord_philip/other_computers/20012.html#71

As far as I can tell only pin 2 has some real differences, but please take a closer look.

Also, have a look at this:

http://web.me.com/lord_philip/other_computers/20012.html#72

The screen on the broken board either goes crazy or displays a good startup screen when the CH1 probe is on C8-13 and CH2 probe on C8-2. Reversing the probes produces the bad startup screen!!! having a probe on pin 12 and 13 shows a bad start up screen. More weirdness.
 
You must log in or register to reply here.
Back
Top