• Please review our updated Terms and Rules here

3032 repair

Gary C

Gary C

Veteran Member
Joined
May 26, 2018
Messages
2,300
Location
Lancashire, UK
Not really for anything other than blogging another PET repair without adding noise in the forums

SO, a 3032 PET with a thin green line on the screen. First thing to check, do I have a vertical sync pulse on pin 3 of the video connector J7, nope. Good Horizontal and video but no vertical. Result.

Page 6 of drawing 320349 shows vertical drive generated by a succession of gates. H10, no activity on inputs, H8 not doing anything because the NEXT line isn't doing anything.
So to page 7 and another logic train to follow.
G8 showing no activity on Q, but does have a clock. Its a D type so should be latching what is presented on pin 12, and thats not doing anything either.

Back through H5 & G2, still no activity but oddly, the 8 bit latch G3 has activity on pin 13 but nowhere else. This bit is a bit circular but basically two counters step through the video and latch the address onto the video bus and also sync to the vertical frame.

Its noticeable that CLK1 is present which is producing a pulse at QA and that is fed to CLK2, but there is no counting going on on QB,QC & QD :) Fault found (well fault 1 of X I imagine knowing my luck)
Both 4 bit counter chips removed and both test as bad on my logic tester, but I'm not totally convinced that it can test counters properly.

New 74177's ordered but I did note that an alternative device was fitted, an S8281A but its just an alternate part.
 
Last edited:
F2 & F4 both faulty. Replaced and I now get a screen full of characters :)

But still not starting. PETTEST installed but no change.

Sync on CPU pin 7 is pulsing so its trying, but I notice that pin 20 of the EDIT ROM is not being pulsed so its not being addressed. Time for NOP to be installed and a check that the ROM select decoder D2 is working which I suspect it isn't.
 
NOP generator fitted and the decoder is working fine. Hum

So why am I not seeing the F000 select pulse on reset, or am I missing it. Time to install the tynmouth board with its reset switch which is a handy test not only in that I can map out the DRAMs and the ROMs but also reset it many times while watching selected signals without having to wobble the PSU.
 
Well, yes there was a chip select going to F000, interestingly with the tynemouth board in and the ROM/RAM mapped out, it would run the test pattern ok, and chip select on F000 remains active, however the data bus was just a mess. Most peaks about 2V.

Something is dragging it down. All ROMS removed and all support chips, powered up and Ha! now running the test pattern and the data bus looks good. All ROMS back in, no problem, 6522 back in C5 and bang, bus is pulled down :)

So 6522 required. Spare one handy and fitted :)

and onto next problem, PETTEST works fine but not with the original ROM
 
ROM 901447-24 stuck in my programmer and reading zilch. All zero's (it should read as a 2716). New 2716 burnt and yay, the PET message is now on display.

But the keyboard doesn't work :(
 
Ok, nothing too bad. BCD to decimal decoder has failed.

Which is a shame as I need the keyboard working to check the spare 6522 but time to order some parts.
 
Arrrrrrrrrrgh !

You would have though I would have learnt by now.

The decoder is open collector and works fine, its the keyboard thats dodgy :)
 
Carbon pads all given a rub down with some clean paper and every key has sprung into life.

Its noticeable that the edge connectors on this board are all gold plated (unlike my 8032-SK) but are filthy so the next job is to give them a clean and test.

The plastic case has come up a treat though. From where the monitor was mounted, you could see the original colour but after cleaning, the line has all but vanished and the case is back to virtually its original colour.
The magic ingredient was Mr Muscle oven cleaner ! Has lifted the dirt right off where detergents and hypochlorite's didn't. Had to then give it a real soak to remove any residual, but I have not seen a PET this white for ages.

Will try it on my smokers C64 next.

So, just some minor cleaning, testing of ports and it will should be ready to deliver back to NWCM for display on Saturday.
 
Oh well

Failed again.

Display is working, just showing uncleared memory. CPU is not running, no activity on Pin 7 even on reset.

Hang on, why is reset low ?

Capacitor C68 failed and reset timer not running. New one fitted and PET rides again.
 
And then it doesn't

Now its showing incorrect characters. If I do the poke 59468,15 it changes map and shows some sensible characters so I must have a bit out somewhere, but not a RAM fault as its interpreting the keypresses ok so in the display buffer or char gen ROM.

And now my 8032-SK has suddenly stopped reading from the disk drive. Half way though loading it throws an error and the device is only receiving data packets of 0. ATN is being pulled low, but no data is being sent. Looks like the 6520 isn't working :(

Oh, the PET gods, please smile on me !
 
So, looking at the startup banner, the O in Commodore has been converted into a left and top line. From PETSCI thats 0F translated into 4F or to put it another way, bit 6 is stuck high.

Into the drawing I know its not the actual SRAM because it wouldn't interpret the BASIC commands properly, so its either the 8 bit latch or the character ROM itself. The data on pin 8 of F9 looks fine as expected, but not on pin 9, so the latch is suspect
 
You must log in or register to reply here.
Back
Top