Building the KLyball 600D

[BillO]

A question for anyone that has experience with both the 600B and the 600D.

On my 600D, the first column of video seems to start 3 columns over from where the same first column does on the 600B. Like below:

Code: Select all

600B:
D/C/W/M?

600D:
   D/C/W/M?

Is this normal?

[bxdanny]

A quick question for anyone with working 600D: If the board is in 48-character mode, and you press the reset button/Break key, does it switch back to 24-character mode? I seem to remember that it does, but I'm not sure.

[BillO]

If I recall correctly, 48 char mode will survive a reset, but not a cold start.

[bxdanny]

If I recall correctly, 48 char mode will survive a reset, but not a cold start.

No, that doesn't sound right. The ROM was the same as on the 600B, so there is nothing in the code that would change the screen mode.

[BillO]

When you do a cold start, some code and vectors (locations of I/O and other routines) are written to RAM memory. The "switch" code for the 48 char mode modifies these vectors and loads up some of it's own code. If you do a reset and a cold start, all the changes will be over-written. If you do a warm start, or enter the monitor, none of that code is touched so the changes placed into memory should not be affected. Not sure what booting from disk would do as I never had a FD system back in the day.

I could do a test to very this but I don't have a copy of the "switch" code. If someone can give me a copy I'd be glad to try it out.

[bxdanny]

When OSI sold the C1P Series II systems, using the 600 Rev. D board, they had the same SYN600 ROM that the earlier C1P systems had, which means there was no support of any kind in ROM for 48-character mode, or for writing to the port (at $D800) that controlled which mode the system was in. You should be able to switch the hardware into 48-character mode with the command POKE 55296,1 and then see if it switches back on reset; I'm pretty sure now that it will. Probably the reset line is routed to a flip-flop or similar circuit that controls which mode the screen is in.

Regarding your earlier question about the video seeming to be shifted right, if I did the math right 3 character widths in 24-character mode represent 6 microseconds of horizontal scan. That's way too much to be due to the extra logic the video signal must pass through due to the existence of the two screen modes, I would think. As I said, the ROMs were the same, and the "D" of D/C/W/M should represent location $D365 on either version of the board. I never did a direct A/B comparison of the two Superboard versions on the same monitor, but I do remember that the horizontal hold usually had to be adjusted when switching a monitor between a C1 and a C4 (and of course, between either and broadcast TV). Since the resistors and capacitors (especially the latter) used in the RC circuits for sync pulse widths and such had wide tolerances (up to 20%), my guess would be that the difference you are seeing is simply due to variations in those component values. But I'm only guessing.

[BillO]

When OSI sold the C1P Series II systems, using the 600 Rev. D board, they had the same SYN600 ROM that the earlier C1P systems had, which means there was no support of any kind in ROM for 48-character mode, or for writing to the port (at $D800) that controlled which mode the system was in. You should be able to switch the hardware into 48-character mode with the command POKE 55296,1 and then see if it switches back on reset; I'm pretty sure now that it will. Probably the reset line is routed to a flip-flop or similar circuit that controls which mode the screen is in.

Ah, okay. I think I see where we are not on the same page. Yes, a reset will change the the POKE back to the 24 char mode. However, just doing the poke by itself will not make the 48-char mode work right. The lines will not fill with characters in a normal fashion. You also need the driver code to be loaded otherwise there will be gaps in the text lines and the sides of the screen will 'flip' around as the lines scroll. That code will be wiped out with a cold start.