• Please review our updated Terms and Rules here

CBM Pet 4032 DOA

im uploading it now on to youtube.

http://youtu.be/K6Oz98ILI9M

I'm fascinated by the code from the ROM , how is it edited and what code is it?

but it works really well, how ever I'm not sure what its telling me though...apart from good 16K ram and showing the CRT running. i presume it means the CPU is running as well?


Ive take some measurements on the CRT again and the regulator i thought had blown that was pulling 18v is now pulling 25v and getting pretty hot the spec sheet shows a max of 24v so maybe it is failing after all.

not sure.

When the PET is up and running i will check the CRT board in more detail , measure the caps , regs and resistors - something is failing. from a quick visual i can see the back of the tube is discoloured from the 22mm neck end up to about halfway up and then it changes back to the silvery colour that tubes should be - perhaps its just worn out?


Anyway any thoughts on the RAM problem ? I've studied the schematics and I'm stumped to be honest.
 

Attachments

  • crt.jpg
    crt.jpg
    86.7 KB · Views: 1
Last edited:
With Petester ROM fitted

The cpu sync is LOW and pulsing - according to my logic probe set to cmos.

Pettest2 continually loops so one will get ~25% duty cycle pulse on SYNC at it fetches an OP Code every four cycles or so. But with the normal ROMs in, does it hang with no SYNC? I'm glad you have a logic probe, but use it in TTL mode as the PET is filled with little TTL chips and a few NMOS big chips that have voltage swings closer to TTL than the newer CMOS which swing almost rail to rail (ground to Vcc). TTL swings from ~0.3 V to 4.4 V. With high loads of many input gates on an output, a TTL HIGH might be at 3.0 V which is fine.

12- 3.11v high/low & pulse on logic probe
13-3.11 high/low pulse
14-3.11 high/low pulse
15-4.25 stuck hi

So A15 is the problem?

PETTEST runs in the $F000 address space and does a lot of reading and writing in the $8000 Video RAM space so note that the most significant address line (A15) will mostly be set HIGH. But it should transfer data patterns to low RAM which I would expect would cause A15 to go low once in a while. Recheck A15 with logic probe to see it blinks a little. Can you set the blink time for the pulses on your probe or set it to 'pulse catcher' mode where it latches on detecting a pulse?


With commodore kernel fitted / ready or random graphics displayed

12=0.04v hi
13=0.13v hi
14=3.13v hi
15=3.14v hi

I don't understand this data. Do you mean the pattern starting with A15 is 1111 to 1100. You have A13 and A12 at 0 volts but say HIGH?? I would think all the lines except A15 would blink briefing before getting stuck somewhere.

I need to tear myself away from this fun troubleshooting for a while for a family birthday party. But I wonder if it is worth a shot to see if we can get into the Monitor mode by holding the Diagnostics input at ground on power up? the signal to be grounded can be found at the expansion port J9-pin20 or at the User Port J2-pin5. If the 60 Hz interrupt is not running, we will not get a cursor.
-Dave
 
hi Dave

enjoy your birthday the PET can wait ;-) its been in a garage for 25 years

i'll set the problem to TLL and have another look.

Ive just been to Maplins and bought a probe - even though i have one, i can't find mine anywhere
 
with pettest running

A15 is pulsing all the time, both hi & low are on too.

logic set to TLL , its a basic probe ..

A12 high / low quick pulse
A13 ,A14 as above
a15 Hi showing and pulse every six seconds
 
Ive take some measurements on the CRT again and the regulator i thought had blown that was pulling 18v is now pulling 25v and getting pretty hot the spec sheet shows a max of 24v so maybe it is failing after all.

not sure.

It has failed. Turn off the PET! Disconnect the cable going to the monitor which should turn off the high voltage, but note that the 22VAC that feeds the 18VDC regulator goes directly from the big transformer to the CRT on a separate twisted pair cable. you would have to unsolder it from the transformer if you want to run the PET without power to the CRT. But Fix the regulator before applying power to the CRT. Be careful of the hi voltage in the CRT area.
 
ok Dave

will do

cheers

Ive ordered a new part so that will take a few days to arrive....

nice to have a break from it :)
 
im playing with the video ram..trying some from an Apple machine

Syntax error in 0

just came up
 

Thanks for posting. Now I know what is happening. I thought the 'Random' pattern you posted earlier was from normal power up with proper ROMs not the PETTEST EPROM. With PETTEST the pattern is not random but is a display of all PETSCII characters in order. So that is why I thought the A7 address line was somehow tied to D7. That issue is settled. When at the real ROMs are installed at power up, do you get a flash of random characters on the display before the screen goes blank?


I'm fascinated by the code from the ROM , how is it edited and what code is it?

It is 6502 Assembly code. One would use an ASCII text editor to create the source code file, then use an Assembler program to take the source file and create Load file. This file contains the proper 1s and 0's (binary machine language code) along with the starting load address for the PET to properly load the file into RAM as a program (.prg file). However in this case, there was only a need to have a binary image that would be burned into an EPROM targeted for the $F000 memory space.

I'm not sure what its telling me though...apart from good 16K ram and showing the CRT running. i presume it means the CPU is running as well?

It says a lot, The CPU and data paths are fine. Most of memory is probably fine although PETTEST can be tricked with addressing problems. What it then tells us is that the problem is in reading the ROMs. Either a ROM is bad or there is an intermittent connection or there is perhaps a tricky Addressing issue.
 
ah okay , no it was pettester :)

okay so it could be roms ? I have either eprom or original roms except for the kernel which is dead - I burnt 901465-22 to a 2532 rom

its all I have ...I wonder if that is the problem here

yes I usually get the full screen of characters so a second then whatever it wants to display
 
im playing with the video ram..trying some from an Apple machine

Syntax error in 0

You may have found a clue. Remove the upper 16K chips (A4, A6, A8...A18 ).

Check all RAM sockets for solder splashes, cold solder joints, etc. We may have two address lines intermittently touching.

This is real fun for me, but must get going....
 
Last edited:
ah okay , no it was pettester :)

okay so it could be roms ? I have either eprom or original roms except for the kernel which is dead - I burnt 901465-22 to a 2532 rom

its all I have ...I wonder if that is the problem here

That should be fine. Use 2532 for all ROM sockets except a 2716 can be used for the Editor ROM (2K file) in the newer 4000/8000 series PET. In older PETs a 2716 must be used for the Editor ROM.

yes I usually get the full screen of characters so a second then whatever it wants to display

That is quite normal. It means the scrambled screen is being cleared properly under program control before the PET runs amuck out of program sequence for some reason.
 
good somethings normal then

what does the chirp differences mean?

short one or double one?
 
The waviness in the picture is mains ripple and is due to a bad capacitor. The "discolorisation" on the CRT is of no concern at all.

Is the "crackling" a "snap-snap-snap", an occasional "snap", or the same static electricity sound you typically get when first turning on or off a CRT (except in your case whilst it is running)?
 
hi

the crackling sound is only until the screen is initialised then it stops completely.

bad caps on the crt board i pressure ?
 
If I understand right, that crackling could well be normal.

There is one bad capacitor causing the waviness, and it's either on the board under the CRT, that is, not the one attached to the back of the CRT itself. Or it's in the PET power supply, I'd have to review the schematics to see which. But you can find it pretty quick by using a high impedance AC voltmeter or oscilloscope, by checking for high levels of AC at the positive lead of all the electrolytics whilst the PET is powered up, very carefully of course.
 
If I understand right, that crackling could well be normal.

There is one bad capacitor causing the waviness, and it's either on the board under the CRT, that is, not the one attached to the back of the CRT itself. Or it's in the PET power supply, I'd have to review the schematics to see which. But you can find it pretty quick by using a high impedance AC voltmeter or oscilloscope, by checking for high levels of AC at the positive lead of all the electrolytics whilst the PET is powered up, very carefully of course.

I've just changed the power supply one , so it could well be on the crt board :) , these thing are 30 years old now

As for the crackling noise at start up , under normal circumstances you would only hear it for a second - but watching a few youtube vids of pet they don't appear to do. You would swear the flyback is sparking but it doesn't appear to be.

mike.
 
Last edited:
Back
Top