• Please review our updated Terms and Rules here

Pet 2001 3032 boots in graphics mode and I can't switch it in text mode

Joined
Nov 18, 2012
Messages
17
Hello,
I'm new on forum and french so please be tolerant :).
So I have a computer PET 2001 3032, it powers on correctly, but in graphics mode, that is to say for example when I type something, the screen shows the graphics characters (circle, line...) and not the text. Moreover the boot message "### commodore basic 4.0 ###" is also writing with graphics characters (I transpose it on paper in text mode with the equivalent keys and it's correct).
I tried to change ROM, to disconnect keyboard... but I didn't succeed to have text characters.
Have you got an idea to solve this problem?
Have a good night!
Jonathan
 
So I have a computer PET 2001 3032, it powers on correctly, but in graphics mode, that is to say for example when I type something, the screen shows the graphics characters (circle, line...) and not the text. Moreover the boot message "### commodore basic 4.0 ###" is also writing with graphics characters (I transpose it on paper in text mode with the equivalent keys and it's correct).

Jonathan,
There is a control signal called 'graphics' that tells the Character Generator ROM to use the graphics character set. The signal may be stuck high or the circuit trace may be open. The source of the signal is pin 39 of the VIA (C5). It goes to the CG ROM (F10) pin 19. As a test, you might ground F10-19 and see if the video is correct. If so, then find out if the 6522 VIA is bad or the circuit trace is open.

http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/2001N/320349-8.gif

http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/2001N/320349-3.gif

-Dave
 
Hello Dave,
ok I changed the 6522 VIA and now I have the correct boot text with good Ram. But again when I type something it's still the graphics characters which are writing on screen (sometimes I can type several text characters but for two seconds only).
I can't test the circuit because I haven't equipment at home, but I will ask my father to lend his scope to me.
Thanks a lot for your help.
Jonathan
 
Jonathan,
Perhaps the real problem seems to do with a stuck shift key as I forgot you must have a graphics keyboard (I have a business keyboard). You can tap lightly on the shift lock, and if that does not help, then with power off, re-seat the connector (J5) going from the keyboard to the main board.
 
Hello Dave,
I tried but nothing. In fact, it's more complex because if I power off computer and I immediately power on again the boot message is written with graphics characters. But if after to switch off I wait 3-4 minutes, I power on and boot message is ok.
I think it's a heat problem (because moreover when the boot message is ok, I can type on the screen several text characters but only for 5-6 seconds) but what chip could be the fault?
Have a good day!
Jonathan
 
I think it's a heat problem (because moreover when the boot message is ok, I can type on the screen several text characters but only for 5-6 seconds) but what chip could be the fault?

It might be an IC like one of the video RAM chips (F7 or F8 ) although normally one would see many more bad characters on the screen. Do you see any reverse video characters also? Are your video RAM chips on sockets? If not, try to "piggyback" a 2114 RAM chip on top of F7. Be careful not to short adjacent pins.
-Dave
 
Last edited:
Hi Dave,
thanks for the advice, there is no reverse video characters, but ok I will try on these days to change the video Ram.
Kind regards
Jonathan
 
It might be an IC like one of the video RAM chips (F7 or F8 ) although normally one would see many more bad characters on the screen.
-Dave
Hi Dave; every time I argue with you it turns out you're right, but that doesn't stop me ;-)

I don't think the video RAM is involved in the text/graphics selection at all, is it?

AFAIK it's controlled only by bit 1 of the VIA (CA2) at location 59468, available at C5 pin 39; it goes directly to pin 19 (A10) of the character generator F10 to select the upper or lower half of the ROM and the corresponding character set.

So I would think that as you suggest there is 1) a problem with the IC sockets C5 (pin39) and/or F10 (pin19), 2) the CG ROM F10 or even the replacement 6522 VIA, or 3) the trace connecting the two.

Jonathan, try wiggling or removing and replacing F10. Failing that, carefully lift F10 pin 19 and alternately connect it to ground or Vcc (pins 12 and 24) as Dave suggested.
 
Last edited:
Hello Mike,
thanks for your message, I tried to move the F10 and to change it also (I had a second 901447-10), I tried to connect 19pin to ground or Vcc, but nothing it's exactly the same.
Do you think it could be a problem of voltage?
Can I try to connect directly 19pin to the 39pin of the 6522?
Have a good day!
Jonathan
 
Ok I have improved the computer's problem in connecting the 19pin to 24pin (I thought 24pin was also the ground) and now I have text characters. I didn't have time to try Poke or other command because my wife wants I stop my experiment for this day! :p
So I will continue tomorrow morning (or this night on the sly :p )
If somebody has got other idea, please don't hesitate to help me! :)
Have a good day and thanks Dave and Mike!
Jonathan
 
Well, that's some progress, but I think Dave was probably right all along; now that I think about it, pin 19 selects between UC/GRPH (low, 59468:12) and LC/UC (high,59468:14) and UC vs GRPH is selected by the shift state and bit 6 of VRAM.

I suspect that where you previously had graphic characters you now have lower case instead and whether you shift or not may not make any difference (although the PET should at least be usable). I think I'd agree with Dave and suspect the video RAM chip F7, specifically pin 12.
 
Well, that's some progress, but I think Dave was probably right all along; now that I think about it, pin 19 selects between UC/GRPH (low, 59468:12) and LC/UC (high,59468:14) and UC vs GRPH is selected by the shift state and bit 6 of VRAM.

Mike & Jonathan,
Yes I think we are on the right track now with the video RAM or the screen refresh logic as the problem. I was also thinking that Jonathan is seeing lower case letters in his temporary fix. He is just forcing the character set for the business keyboard.

Also it must be noted that the 'temporary fix' should be disconnected. F10 pin 19 should not be tied to a hard +5 Volts. If the signal 'graphics' is properly connected to its source C5 pin 39, it is the CA2 line of the VIA which in the PET is programmed as an output and is defaulted to a logic low. Tying this line to +5V might damage the output driver MOSFET in pin 39 of the VIA.

I got us off on the wrong foot with my initial guess that the signal 'graphics' might be the problem.

Sigh, every time I think I fully understand the little PET, a question like this comes up and I get confused all over again. :confused:
 
Hi Mike and Dave!
ok I read your messages, thanks for your technical explanations.
I also think it's a video ram problem because there is still a character's problem after several seconds. I will test to put a heatsink on the RAM or other chips.
I have disconnected the connector pin19-pin24 in order to not damage the MOS.
Thanks again for your advices!
Merry Chrismas and happy new year!
Jonathan
 
I have disconnected the connector pin19-pin24 in order to not damage the MOS.
Sorry, I probably should have emphasized the
carefully lift F10 pin 19
I.e. bend out the pin a little so it does not go into the socket; better yet, use a spare socket in between and bend its pin instead to avoid possibly breaking the ROM's pin.

Or just wait for the RAMs ;-)
Merry Chrismas and happy new year!
Jonathan
And the same to you! And a quick successful repair and lots of fun with your PET!
 
Back
Top