• Please review our updated Terms and Rules here

2001-8 Chicklet Pet Fault Finding.

Pet Rescue

Experienced Member
Joined
Dec 15, 2010
Messages
484
Location
Staffordshire, England
Hi all it's been a while since I've had a project.
I have recently aquired a 2001 chicklet pet with the 320008 assy board.
It was a non working machine that showed garbage.
All four voltage regs are around 5v (4.95v & 5.05v etc)
I removed the PIA's and all 6550's except I1,I2,J1,J2. I subbed the 6502 in another pet to test and found it was duff!
Both PIA's worked and so did the VIA in the test Pet. One of the 6540 rom's had a bent leg in H1.
After straightening and fitting a new 6502 still had garbage screen.

I swapped the video ram with other 6550's and up came the boot screen with 1023 bytes free.
Happy days!!

I then determined that 3 of the 18 6550's were faulty by substitution.
I populated the Ram banks with what was remaining and had 3071 bytes free on the boot screen.
Anyway after cleaning the keyboard and writing a small program on it lost the cursor and some more memory.
The keyboard input still works but just no flashing cursor.

I also then found another 5 6550's had gone faulty.

I subbed the 6520 with a known working chip but still no cursor but the keyboard still worked.
The 6520's were again tested in the test pet and worked ok.

What could cause the cursor to disappear but with the keyboard function still intact.
Are 6550's prone to fail like that or may something else be blowing them.
As stated the 5v from the 4 regulators are all sitting at 5v.
 
What could cause the cursor to disappear but with the keyboard function still intact.
Are 6550's prone to fail like that or may something else be blowing them.

Put your best RAM parts into the J1 and I1 sockets. You need good zero page to store system variables and good RAM from $0100 to $0200 to store the stack for the PET to work.

Double check all sockets! They cause bad connections.
 
Put your best RAM parts into the J1 and I1 sockets. You need good zero page to store system variables and good RAM from $0100 to $0200 to store the stack for the PET to work.

Double check all sockets! They cause bad connections.

Hi Dave I was counting on your reply it's good to hear from you, it's good to see Mike back too!

Thanks for your input, it was working fine even loaded from tape etc when I decided to type in a program and it decided to play up.
It was only on for about 30 mins when the cursor disappeared which led me to believe it had froze but upon pressing the keyboard
the characters appeared as normal.

I can't get over how hard it is to insert the chips into the sockets.

So you are leaning towards a bad RAM? Do they fail easily due to it's age or have I just had bad luck?

Can ROM cause this issue too? (thinking about making an adapter to read the 6540's)
 
A disappearing cursor is not an I/O problem. Certainly not if you can type.

If you press RVS and type, do you get reverserd characters? If not, you have one signal missing in the video circuit.

When i first got my PET, it had the opposite problem. That line was stuck high, so everything was always reversed. There was no cursor for that reason.

With missing zero page RAM, it's amazing it would do anything. Same goes for missing stack memory. But I suppose normally you only need the top few bytes of that.
 
With missing zero page RAM, it's amazing it would do anything.

Good idea about reverse video signal.

Also, typically when a RAM chip goes bad, only a few locations of memory go bad. If it happens to be where the interrupt handler uses zero page, cursor blinking or keyboard scanning problems may crop up.
 
I can't get over how hard it is to insert the chips into the sockets.

Get proper insertion and extraction tools or you may bend or break pins.

70176423_large.jpg40J743_AS01.jpgmos1416.jpg4618389.jpg



So you are leaning towards a bad RAM? Do they fail easily due to it's age or have I just had bad luck?

Those old NMOS 6540 and 6550 memory chips run hot and are prone to failure. Also Commodore used cheap sockets which contribute to 'memory like' problems. However the PET will run OK with 1K of RAM. So I would check KC's recommendation of using the RVS key and tracing the screen data bit 7 (SD7) out of the video RAM which controls reverse video.

If you have too many memory problems with the 6540 and 6550 parts, you will need to consider getting a RAM/ROM replacement board from Tynemouth.
-Dave
 
A disappearing cursor is not an I/O problem. Certainly not if you can type.

If you press RVS and type, do you get reverserd characters? If not, you have one signal missing in the video circuit.

When i first got my PET, it had the opposite problem. That line was stuck high, so everything was always reversed. There was no cursor for that reason.

With missing zero page RAM, it's amazing it would do anything. Same goes for missing stack memory. But I suppose normally you only need the top few bytes of that.

Thanks for the input KC9UDX I pressed the RVS key and typed but the text remained normal.

Good idea about reverse video signal.

Also, typically when a RAM chip goes bad, only a few locations of memory go bad. If it happens to be where the interrupt handler uses zero page, cursor blinking or keyboard scanning problems may crop up.

Dave with best RAM still don't get a cursor and sockets test out fine.

Also when typing in " " i.e print " " etc and press clr home you normally get an inverted heart but it comes out as a normal heart
 
Thanks for the input KC9UDX I pressed the RVS key and typed but the text remained normal.

So KC is right. It is possible that the SD7 output of RAM C4-pin13 is held low due to bad 6550 or something shorting it. So replace C4. Do you have a way of tracing this signal with logic probe or scope?
 
So KC is right. It is possible that the SD7 output of RAM C4-pin13 is held low due to bad 6550 or something shorting it. So replace C4. Do you have a way of tracing this signal with logic probe or scope?

Hi Dave I've just had the logic probe on the output of pin 13 SD7 on C4 the video RAM 6550 and it's pulsing.
I followed it to pin 2 on C1 the 74ls74 which I believe acts as a Master/slave flip flop circuit from what I've read.
But the output is constantly high on Pin 5 which I would expect to be pulsing if the clock on pin 3 is pulsing
and pin 2 is pulsing or am I misinterpreting the circuit?

video circuit.jpg
 
Last edited:
I wish I had my schematic handy. I don't recall a flip flop in that circuit. There is some glue logic associated with that line though, it was a 7400 series chip that was bad in mine. Maybe a NAND gate or a bus transceiver or something.
 
I wish I had my schematic handy. I don't recall a flip flop in that circuit. There is some glue logic associated with that line though, it was a 7400 series chip that was bad in mine. Maybe a NAND gate or a bus transceiver or something.

KC I found the schematic on zimmers, it goes out of the video ram onto a 74ls244 but before that it goes to the 74ls74.
If i am reading it right on that flip flop the high output of one goes into the input on other and ends up being held low.
That then goes to some more logic in the circuit a 74ls00 and then a 74ls08 and finally a 74ls20.
 
I wonder if the flip flop has something to do with blanking.

Do this: turn RVS on, clear the screen with HOME, and fill the screen with spaces. Better yet, write a program to do this, as this PET will not repeat the space bar for you.

With a whole screen full of reversed spaces, you should have a larger duty cycle with SD7 on.

Follow the signal from there. If you lose it at the flip flop, either the flip flop circuit isn't working, or that chip could be bad. It could also be a short after the flip flop, so remove the chip after it if you can.
 
I wonder if the flip flop has something to do with blanking.

Do this: turn RVS on, clear the screen with HOME, and fill the screen with spaces. Better yet, write a program to do this, as this PET will not repeat the space bar for you.

With a whole screen full of reversed spaces, you should have a larger duty cycle with SD7 on.

Follow the signal from there. If you lose it at the flip flop, either the flip flop circuit isn't working, or that chip could be bad. It could also be a short after the flip flop, so remove the chip after it if you can.

I'm afraid the spaces don't reverse, I can't get any reverse characters at all as RVS won't work even in inverted commas clr home comes up as a normal heart, but I like your idea!

How did you find SD7 was the reverse signal?
 
You don't see them reversed, but if you type with RVS on, they are indeed reversed in RAM, unless bit 7 is getting lost in RAM.

All PETSCII screen characters 128-255 are reversed. That is, bit 7 is on for all of them. The PET takes advantage of this by only needing a 7 bit character ROM, with 128 characters. The 8th bit, bit 7, is used to reverse the video. It's labeled as such somewhere on the schematic.

(Technically the opposite is true, the character set is that way because the PET hardware is that way. It's actually pretty slick. The Apple ][ gets all the credit for Woz's genius, but there's a whole lot of genius in the Commodore 8-bits too. They really did think of everything.)
 
You don't see them reversed, but if you type with RVS on, they are indeed reversed in RAM, unless bit 7 is getting lost in RAM.

All PETSCII screen characters 128-255 are reversed. That is, bit 7 is on for all of them. The PET takes advantage of this by only needing a 7 bit character ROM, with 128 characters. The 8th bit, bit 7, is used to reverse the video. It's labeled as such somewhere on the schematic.

(Technically the opposite is true, the character set is that way because the PET hardware is that way. It's actually pretty slick. The Apple ][ gets all the credit for Woz's genius, but there's a whole lot of genius in the Commodore 8-bits too. They really did think of everything.)

KC I'll give it a go and report back later on, thanks for the help. They had their knockers but commodore had a lot of clever people working for them, I wonder at the time whether they were truly appreciated!
 
If i am reading it right on that flip flop the high output of one goes into the input on other and ends up being held low.

From that description it sounds like the 74LS74 flip flop C1 output pin 9 may be bad (always low). The flip flop is used to latch the reverse video information for each character time during each horizontal scan.
 
Well Dave and KC I removed the 74ls74 this morning which seemed to be a bit more difficult than normal.
I cut the pins etc but they were still hard to remove, even half a pad came away (re glued in place afterwards).
After fitting a socket and new old stock 74ls74 I flicked the switch and Yessss!! the cursor is back!!
I got the logic probe out and checked it in normal operation and yes the output should have been pulsing.

Then it decided to die again and the screen froze. I turned off and on and ARRRGH!!! the garbage screen.
This Pet is a right royal pain in the @R$£!!!!
 
Look at it this way: parts are failing whilst you are prepared to work on it; not when you take it to the pub to show it off.

Now would be a good time to check your power supply for correct voltage and ripple.

Also, maybe when you were working on it you flexed the board just enough to make an intermittent connection show itself.
 
Look at it this way: parts are failing whilst you are prepared to work on it; not when you take it to the pub to show it off.

Now would be a good time to check your power supply for correct voltage and ripple.

Also, maybe when you were working on it you flexed the board just enough to make an intermittent connection show itself.

KC the voltages on J8 are 15.2v AC brown wires and 9v dc between red and black on both reds.

The pet is now showing the classic screen for bad video RAM.
It looks like the 6550's decided to give up the ghost with one last finale of a Boot screen and cursor.
The power supply looks ok as the 5v looks steady, maybe I'll get the scope out to check the ripple.

I will have to look at an equivalent part maybe pet vet or 2114 adapters

P.S Just scoped the 9v and 5v dc both have very little ripple.
 
Last edited:
The pet is now showing the classic screen for bad video RAM.
It looks like the 6550's decided to give up the ghost with one last finale of a Boot screen and cursor.

I will have to look at an equivalent part maybe pet vet or 2114 adapters

This additional trouble will make memories of fixing the PET even sweeter when you get it running.

How long was it working before the freeze? Minutes or an hour (thermal question)?

The garage screen could mean bad video RAM, but it also could just mean the PET is not running in program or stuck. Check Sync (F3-pin7) on the 6502 chip for pulsing. If pulsing at least we know it is fetching instructions. If not pulsing it is stuck and probably means it hit a bad ROM and saw a KILL instruction.

If later you decide to replace the 23 main memory chips, you may want to look at a British replacement memory board that I have used with excellent results: Tynemouth. Tell him it is for the 2001-8 style socket if you have the black socket (for the 6502).

In any case save the 6550 RAMs as spares for the video RAM. Hang in there.
-Dave
 
Back
Top