What Have I Done...

[ChildOfCv]

BTW, which card do you have?

 

#1

 

#2

 

#3

 

None of the above?

[bluejay]

The first one, extended 80 column/64k.

[ChildOfCv]

Ok, so it looks like the actual switch to 80-columns, as far as showing skinnier characters, comes from a 3rd custom logic chip, UD1.  The 80-column enable comes in on pin 9.  Pin 7 might possibly block 80-column mode depending on other mode select issues though.  But without exact information on the chip, that's only a guess.  Anyway, pin 13 is where the character width is finally affected.  In 40-column mode this pin will be pulsing and should look like a voltage somewhere significantly above 0V.  In 80-column mode, it should be about 0V.

 

So again:

 

after POKE 49164,0,  read pins 9, 7, and 13

 

then after POKE 49165,0  read pins 9, 7, and 13

 

Make sure all readings are taken on DC volts.

[bluejay]

UD1 is the one labeled as HAL, right?

The socket covers the labels.

 

Hmm, this time I'm getting 4vdc, 3vdc, and 7vdc for pins 7,9 and 13 with poke 49164,0.

Poke 49165,0 gives 3vdc, 0vdc, and 0vdc, for 7,9, and 13.

55 minutes ago, ChildOfCv said:

Ok, so it looks like the actual switch to 80-columns, as far as showing skinnier characters, comes from a 3rd custom logic chip, UD1.  The 80-column enable comes in on pin 9.  Pin 7 might possibly block 80-column mode depending on other mode select issues though.  But without exact information on the chip, that's only a guess.  Anyway, pin 13 is where the character width is finally affected.  In 40-column mode this pin will be pulsing and should look like a voltage somewhere significantly above 0V.  In 80-column mode, it should be about 0V.

 

So again:

 

after POKE 49164,0,  read pins 9, 7, and 13

 

then after POKE 49165,0  read pins 9, 7, and 13

 

Make sure all readings are taken on DC volts.

 

[bluejay]

Meanwhile, are Atari paddles compatible with Apple //es with the paddle input?

NVM, It doesn't fit.

Edited December 21, 2019 by bluejay

[ChildOfCv]

  

6 minutes ago, bluejay said:

UD1 is the one labeled as HAL, right?

The socket covers the labels.

 

Hmm, this time I'm getting 4vdc, 3vdc, and 7vdc for pins 7,9 and 13 with poke 49164,0.

Poke 49165,0 gives 3vdc, 0vdc, and 0vdc, for 7,9, and 13.

 

There is something seriously odd about that then.  The shift register that controls each pixel has a 14MHz clock.  When pin 13 pulses high, it prevents a shift on that count, which is effectively divides the clock in half and doubles the width of the pixels.  So it takes a well-timed signal to get 40-columns, and no signal should give you 80-columns.  If anything is wrong, it should be stuck in 80-column mode with a space between each character.

 

I don't suppose you have a frequency counter or oscilloscope?

[ChildOfCv]

5 minutes ago, bluejay said:

Meanwhile, are Atari paddles compatible with Apple //es with the paddle input?

NVM, It doesn't fit.

They're electrically compatible, but you'd need an adapter to convert between the plugs.

[bluejay]

8 minutes ago, ChildOfCv said:

I don't suppose you have a frequency counter or oscilloscope?

Nup, wish I did tho.

[ChildOfCv]

I looked over the schematics again, and pin 13 seems to be the only possible way to affect pixel width.  Could you take another measurement of pin 13 just to be sure?

 

 

[Aunty Entity]

On 12/19/2019 at 3:29 PM, ChildOfCv said:

Just inventing diagnostics tests on the fly here, but:  Try putting the 80-column card in without the disk card, and typing PR#6.  See if it goes to 80-column mode.  Or put the 80-column card into slot 6 and see if PR#3 activates it.

Ahhh thats odd!  It shouldn't do anything except sit there sucking its thumb if you do that to it.  Is it just an 80col card or an Extended80col card with memory on it? Will it pass the onboard ctl-closed apple-reset self test?  Solid cursor means the 80col firmware is active, if you just hit esc it should have a plus in it.

[Aunty Entity]

On 12/21/2019 at 11:12 AM, ChildOfCv said:

They're electrically compatible, but you'd need an adapter to convert between the plugs.

No, the atari's are digital on/off switches, the II uses resistance for a range of values.  Most software will still work with the discrete values of 0 or 255 out of the joystick port, but others will just zip you straight from left to right with no possibility of sitting in the middle of the screen anywhere.

 

[deepthaw]

Have you gotten ADTPro setup? There’s diagnostic software to test your cards. Both apples, control, reset. Should show KERNEL OK if it sees no errors. 

 

 

[ChildOfCv]

7 hours ago, Aunty Entity said:

No, the atari's are digital on/off switches, the II uses resistance for a range of values.  Most software will still work with the discrete values of 0 or 255 out of the joystick port, but others will just zip you straight from left to right with no possibility of sitting in the middle of the screen anywhere.

 

He was asking about the paddles though.  The paddles use a potentiometer just like PC and Apple joysticks do.

[ChildOfCv]

7 hours ago, Aunty Entity said:

Ahhh thats odd!  It shouldn't do anything except sit there sucking its thumb if you do that to it.  Is it just an 80col card or an Extended80col card with memory on it? Will it pass the onboard ctl-closed apple-reset self test?  Solid cursor means the 80col firmware is active, if you just hit esc it should have a plus in it.

Yeah, we've since established that he's using a standard 80-column card in the aux slot.  It's a different number of pins, so you can't switch them.  If it were one of the aftermarket 80-column cards that were always dropped in slot 3, that would be a different story.

 

Well, the other problem is that IIe added a lot of 80-column-in-slot-3-specific glue logic that does funny things with I/O access.  If nothing is found to be plugged in to slot 3, the IOU and MMU will handle the requests themselves.  But if it thinks it sees a card in the #3 slot, it will turn off the #3 redirection.  As a side note, they also employ a sort of shadow ROM with the extra 64K that allows it to copy BIOS into RAM at the same memory location and then write-protect that RAM region.  It uses this to overwrite C800-CFFF with handling code for the slot being accessed.

 

What else is left is that the HAL is responsible for the timing related to the 80/40 switch.  The dot clock for the text mode font generator clocks for an 80-column screen, and in 40-column mode the HAL will also send a signal on every other clock tick to tell the pixel shift register to ignore that tick, which doubles each pixel width.  So it goes to 80 columns when the HAL shuts up.  This suggests that the HAL may be the problem, and replacing it may solve the problem.

 

[bluejay]

On 1/3/2020 at 7:36 PM, deepthaw said:

Have you gotten ADTPro setup? There’s diagnostic software to test your cards. Both apples, control, reset. Should show KERNEL OK if it sees no errors. 

 

 

Gotta try that! I was out of town and didn't check this thread for a while.

Just did it, and it says KERNEL OK, just like it did when I did the //e's diagnostics(solid apple, ctrl, reset)

Edited January 12, 2020 by bluejay