• Please review our updated Terms and Rules here

Another PET 2001 motherboard needs your help...

fachat said:
Yes, that would be a good action to measure the delay between the 8MHz clock and the C8 input, comparing the working and the broken PET. If that really reveals that delay difference, I fear we have to take the risk and replace the C9. If not, then I'm running out of ideas as well....
Click to expand...

Thanks for the reply André.
some questions...
Do I use the 8mhz clock as a 'trigger' for C8?
Do I get the 8mhz clock from E2-5, or possibly E2-6?
Do I measure C8 clock from pin C8-13, C8-12, or possibly C8-2?

fachat said:
But a warning: there are different versions of the '93 IC out there, with different pin assignments. Please make sure that the one you got has the exact pinout as required.
Click to expand...

I got myself a SN74LS93N from Texas Instruments. You can see the pin arraignment here:

http://datasheet.octopart.com/SN74LS93N-Texas-Instruments-datasheet-151515.pdf

...as far as I can tell it looks compatible, just I'm not sure about pins 14 and 12 as the PET schematic doesn't say what these pins are.
 
Last edited:
I had suggested to Philip over a week ago that he replace C9, so I'm probably as eager to see the answer as he is ;-) . I also warned him about the different pinouts (both non-standard, an odd chip indeed), but as long as he sticks to an LS it should be fine; AFAIK within a given family they're compatible.

In any case, we must surely be getting close; this would make quite a blog considering what this poor PET board has gone through, including the miles it's travelled ;-) ; I've certainly found it interesting.
 
Nama said:
...as far as I can tell it looks compatible, just I'm not sure about pins 14 and 12 as the PET schematic doesn't say what these pins are.
Click to expand...
The '93 counter has two sections: pins 14 and 12 are the binary counter section, shown on the schematic just below the divide-by-eight section.

I think only the 74L93 has a different pinout from the '93, LS93 and HC93.
 
MikeS said:
I had suggested to Philip over a week ago that he replace C9, so I'm probably as eager to see the answer as he is ;-) .
Click to expand...

Yes, I agree with you and André that the timing of this ripple counter should be measured.

But we must remember that a very slow counter would make the C8-3 output a little skinny, but would not effect the amplitude issues that we are seeing on pins 3 and 2, the outputs of the C8 LS107 flip-flop. We must not lose sight of that.

There may be two different issues.

We missed a clue when Phil reported that when he lifted C8-3, he did not see a pulse there (I'll bet it was stuck at high). With an open drive on that circuit, the D8 NAND gate should see about 1.6V on that input and read it as a high. The gate would still work properly to issue a reset. So there should have been pulses on C8-3 with the pin lifted.

Therefore, something on the C8-3 circuit is holding the open line to less than a volt. That may explain why the amplitude of C8-3 is only around 2V when it switches high.

The other problem is why is the /Q output (C8-2) both floating at 1 V above ground and so sensitive to wires hanging on it?

If we solve those mysteries, the PET should be fixed.
 
dave_m said:
Then tack solder one lead to pin 7 of U8 (ground), and the other lead to one of the test outputs such as pin 3, 2, 12 and 13. Log the results and which tests seem to make the problem better or worse.
Click to expand...

I'm assuming U8 is actually C8...right?

Results:

Pin 2: Blank white screen with vertical black lines. Flashing cursor can be seen in wrong place so assuming problem still exists.
Pin 3: Inverted screen (black type on white background). Problem still exists
Pin 12: no change. Problem still exists
Pin 13: Screen fixed!

Now I know worst case is that I can fix the issue by attaching the cap between earth and pin 13 and not have a leg hanging out :-/
 
Philip,

I just want to say that I admire your determination (or you're a glutton for punishment) for sticking with this so long. I would have put a couple of mercury drops into the sucker by now :)
 
Nama said:
Now I know worst case is that I can fix the issue by attaching the cap between earth and pin 13 and not have a leg hanging out :-/
Click to expand...

Phil,
Well that's true :), but I think the value of capacitance I picked was too large to give us much insight. I wish I had asked for 33 pF. But at least we got a little good news.

You can remove the cap as we have other things to try.

When you get an opportunity, can you lift pin C8-3 once again (it's going to fatigue and break off pretty soon), verify that the leg is at a high (>3.6 V) meaning that we are incorrectly missing the reset pulses from D8-8 and measure the circuit voltage at either D8-10, D5-9, or C1-11. It should be 1.6V but I'm guessing it will be around 1 V.
-Dave
 
Results with 33pf cap:

- pin 12 and 13, same as with previous test. 12, no change and problem still exists. 13 shows good screen.
- pin 2 shows crazy repeating flashing cursor, but problem still exists
- pin 3 no change. Problem still exists

I'll do the voltage test now...
 
Voltage test:

C8 Pin 3 leg lifted: 123mV
D8-10: 2V
D5-9: 2V
C1-11: 1.8V

...All these readings were taken with C8 pin 3 still lifted.
 
Druid6900 said:
Philip,

I just want to say that I admire your determination (or you're a glutton for punishment) for sticking with this so long. I would have put a couple of mercury drops into the sucker by now :)
Click to expand...

In most cases I get more enjoyment from fixing old computers than actually using them...go figure!

It's like a good detective or 'who done it' story, and in this case I am Watson to your Sherlock.
 
Last edited:
Nama said:
Results with 33pf cap:

- pin 12 and 13, same as with previous test. 12, no change and problem still exists. 13 shows good screen.
- pin 2 shows crazy repeating flashing cursor, but problem still exists
- pin 3 no change. Problem still exists
Click to expand...

OK, this test indicates issues with pin 2 and pin 13.

Pin 2 controls the loading of the character ROM shift register and should have nothing to do with blinking cursors which is controlled by a software interrupt routine. I still think pin 2 is tied to something it should not.

Pin 13 is a sensitive point for reasons as yet unknown.
 
Nama said:
Voltage test:

C8 Pin 3 leg lifted: 123mV
D8-10: 2V
D5-9: 2V
C1-11: 1.8V

...All these readings were taken with C8 pin 3 still lifted.
Click to expand...

Phil,
This data has my head spinning. It is the exact opposite of what I was thinking.

Did you measure this data with a meter? If so, maybe you should re-measure with a scope as skinny 30 nS positive pulses at a 1 MHz rate might average out to a low DC reading (although 0.123 mV is awful low). A 2.0 V DC average reading could be a square wave, etc., and I am confused anyway :)

I wish I had also had you measure the signal at C8-13 (reset). It should be an 8 MHz clock in this test condition since the NAND gate that drives it ( D8-8 ) has one input is open (high) and 8 MHz on the other input, but possibly it is low, keeping the C8 output constantly low?? Maybe there is a problem with the D8 Nand gate?? This is a never-ending problem. :)
-Dave
 
Covering my bases, here is my kludge fix just in case we never get to the bottom of the real issue.
I've had the board up and running Android Nim for 3 hours now with no issues seen with this cap in place.

http://web.me.com/lord_philip/other_computers/20012.html#84

Dave,yes I did the voltage reading with my digital multimeter. I haven't had time this evening to do any scope test as I've had my hands full with my 2 year old son!!!!
 
You must log in or register to reply here.
Back
Top