Commodore PET CBM 4016 ASSY 8032080 RIFA burn and more

[DrAlis]

I can not see what uc3 pin 2 is connected to. no trace visible for me top or bottom.

 

[daver2]

It probably is routed under the chip.

I am slowly understanding...

A job for tomorrow now.

Have a good night.

Dave

 

[daver2]

Do you know what the history of this PET is?

Has it ever worked (as far as you know)?

Dave

 

[DrAlis]

Yes let's continue tomorrow. I have a simpler task - I am practicing schematics with a model f keyboard. It should take you 5 seconds to fix. If you'd like to take a look at my new post in IBM PC / XTs. Thanks.

 

[DrAlis]

Do you know what the history of this PET is?

Has it ever worked (as far as you know)?

Dave

Previous owner said always worked. I mean, we almost had it working here together and got PETTESTER to run. Our result was only bad RAM. It is only after the RAM swap it turned bad. Must be a silly handling postage accident.

 

[daver2]

OK.

I will re-read our posts again to see where we got to.

I can see some modifications that should have been made to this board at the factory, but they do not appear to have been done in their entirety for some reason.

I will have another look tomorrow when I have my laptop available and not my phone.

Dave

 

[daver2]

OK...

It appears that there were a number of modifications to this board (that should have been implemented within the factory) described on https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032080-1.gif.

I can see some of the modifications on your board, but the UC3/3 to +5V track appears not to have been done for some reason.

It might be worth CHECKING these modifications against your board (to see which ones have been done and which ones not). However, DO NOT implement any that have NOT been performed.

As we almost had this machine working, then it is likely that what Commodore had done in the factory was fine.

With PETTESTER working (pulses on CPU pin 7 (SYNC)) can you check the following two pins on UC3:



UC3 pins 9 and 11 (/READ EVEN and /WRITE EVEN).

I suspect the pulses will be there - but you may have missed them last time around? With your new-found oscilloscope skills (sweeping the timebase) hopefully you may now see them? Report back.

Can you then move on to look at all the pins of UA2 (74166) - HIGH, LOW, or PULSE again. Don't forget to be patient and sweep the timebase if nothing appears to be on the pin. I have asked you for all of the pins on this IC. Some of the pins are connected to 0V and +5V (so will be a static HIGH or LOW). I will be using these as 'check signals' to see if your oscilloscope skills are improving!

We are on this schematic if you want to 'follow along': https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032081-08.gif.

Dave

 

[DrAlis]

Quick question: Isn't the schematic we are on the one that is incorrect - as it shows UC3/3 as a signal rather than high? Is there a schematic that has UC3/3 high? I'll get measurements later.

 

[daver2]

Yes, there is, but this is the univ2 schematic - and your board is NOT a univ2 (if it is an 8032080).

If you look at the link I provided to you (https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032080-1.gif) you will observe that the permanent HIGH copper track is cut and a wire link added. However, your machine does not appear to have this fix - so your board is weird! It did appear to work at some point - so I am inclined to leave it as it is rather than introduce more bodges...

Universal (yours) 8032081:



Universal 2 (not yours) 8032087:



Dave

 

[DrAlis]

UC3 pins 9 and 11: pulse (slow)
UA2 pins 1 to 16:

low pulse pulse pulse pulse low STRANGE low

STRANGE pulse pulse pulse low low INV1MHZ high

STRANGE is a flat but at 1.8volts...

 

[daver2]

If you have activity on UC3 pins 9 and 11 then that is telling me that PETTESTER is writing to (and reading from) video RAM. So our initial concern regarding observing no activity is now unfounded.

Most of the data pins on UA2 look good (implying the video RAM is providing varying data into the character generator and the character generator is providing data to the parallel to serial converter (UA2) correctly).

The output from UA2 is LOW - so data is going IN but not coming OUT.

I am surprised that UA2 pin 9 appears to be floating. This is the /CLR pin. I would expect this pin to be pulled HIGH (possibly to /INIT or PULL UP 1). Floating is not good. it is also not identified on the schematic!

HOWEVER, the biggest issue is no 8 MHz clock on UA2 pin 7...

This signal is sourced from UD3 pin 5 (SN74177). This is the /CLK16 / CLK8 signal (look at schematics https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032081-08.gif and https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032081-06.gif).

Can you check for 8 MHz on UD3 pin 5 please?

The clock signal should be routed via jumper J1. Is this jumper fitted?

Dave

 

[daver2]

As a bit of education...

If we know the signal we are supposed to be measuring is 8 MHz, this equates to a period of 125 ns.

Set the oscilloscope to something as close to 125 ns/div as is available, and then tweak the timebase (if you can see something) accordingly to observe the signal detail.

Dave

 

[DrAlis]

UD3/5 got 8mhz . this is at the fastest limit of my toy scope. jumper 1? where is that fellow? i attach jumper like photos

 

[daver2]

The layout of the components (and the links) are here: https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032080-7.gif.

Link 1 is installed and 2 is removed (see IMG_6640).

You need to follow UD3/5 through to UD2/7 (and see where this signal disappears).

Look on schematic https://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ/8032081-08.gif. You will observe that signal:



Goes to a number of IC pins on this schematic. Check these other ICs and pins for the 8 MHz clock.

Dave

 

[DrAlis]

Dave, all the connected ic pins - they are all floating - on gif 8. I looked at gif 6 where it is supposed to come from UD2/10. I attach an image. Even I got to laugh... Should we call it off? The schematic is bullocks, right?

 

[daver2]

You are not reading (or understanding) the documents I am pointing you at...

UD2 pins 10 and 11 SHOULD be cut and there should be a wire link underneath the IC (solder side) bridging UD2 pins 10 and 11. Is this link present?

If the board WAS working (and we observed some display output) then either something has gone wrong with an IC, or some of the underside links have somehow become disconnected.

This is also the problem with all things vintage. We have what we have. It is possible that Commodore re-issued the schematics (that don't exist anymore) or gave up and produced univ2...

Dave

 

[DrAlis]

No?

No need to mark them... they are marked by flux ? residue

 

[daver2]

I have just checked the schematic again.

For a pure 40 column PET, the link does not need to be present across UD2.

The 8 MHz clock should go into one 'end' of link 1 and the other end of link 1 should go to the IC pins on schematic -08.

Use the schematics, your multimeter (low resistance range) and your eyes to identify where the 'break' is in the signal path.

Dave

 

[DrAlis]

f*ck it i bridged it and i win
and f*ck the schematic apparently
excuse my french but we sank a lot of hours in this hellhole of a device

 

[DrAlis]

i cant believe we cracked that nut after almost a year and countless hours. we both better have some celebration tonight!

thanks for your help. i knew it was a handling issue... the solder bridge broke off. you can see it even in the post #5 still being there!!! amazing conclusion... i guess i pop the regular rom back in. holy smokes...