SidKick pico thinks it a Drean C64

[MarkMccabe]

I built mine using the Clone pico. Tested it on a NTSC 326298 5 pin video C64 (the oldest revision). On all of them I changed the bus timing to 9 and the phi2 to 13 to get to the menu consistently, but 2 of them have slightly garbled menus and won't save changes. These two also always thinks they are installed on a Drean C64. Funny thing is both work perfectly, pressing P in the config prog plays music, and they pass all diagnostic cartridge tests for the SID and control ports.
All the other ones I built come up fine and let me set them to 6581's play the music fine, but fail the control port and SID diagnostic cartridge checks, even with the paddle offset set to 20 during setup. What gives? Why does the garbled one determined to remain a Drean work perfectly while the ones with a good menu do not?
Oh, and my test machine has the original ROMs, VIC, CIA's and PLA (no aftermarket or modern replacement parts) if that helps explain my conundrum. I am also using the SIDKick with the external DAC option (which I have not installed, and using the V0.211 firmware if that helps explain things here.

[frntc]

#70 (reply in thread)

[frntc]

Plz note the paddle values (with Game Controller Tester or the config menu) of those which do not pass the diag and let me know.

[MarkMccabe]

I tested 4 Sidkicks, the menus were mostly flaky when using the default 15 and 11 timing with a paddle offset of 20. I changed the settings to 13 and 9 with a paddle offset of 20 and the config menu were mostly steady.

Using the NTSC 326298 5 pin video C64: All 4 Sidkicks failed the diagnostic cartridge tests for SID and the Control port. If I installed an ARMSID or original SID chip the computer passed all diagnostic checks. Upon removing the diagnostic cart and dongles I checked the paddle values in the config utility and they read FF. For some reason, this (maybe all???) 326298 board would not run the game controller test gct.prg file. It would start to load and then freeze up with a solid blue screen. It did not matter if I had a SIDkick, ARMSID, original SID chip, or nothing at all installed. Games however worked fine when either loaded from the SidKick's program list directly, or from an SD2IEC adapter. Kind of odd, perhaps it is because this board revision is the oldest C64 one? I don't know enough about the gct-prg to hazard a guess.

So next I moved on to trying again, this time using a 250466 mainboard and the same 13 and 9 with a paddle offset of 20 settings. All 4 cartridges fail the diagnostic cartridge tests for the SID and control ports. The config menu shows FF for the paddle values (still with no paddles connected). The game controller test gct.prg runs fine and showed all paddle values as 255 (again with no paddles connected).
I got the same results regardless of if I set the SID up as a 6581, 8580. or 8580+digiboost. SID2 was set to NONE and I did not have a DAC connected to any of the Sidkicks. The config menu in all cases showed the correct NTSC clock values, and for the paddles I had RAW selected.

Should I repeat the tests on the 250466 mainboard with the paddles connected and report the results? Was any of this helpful to you?

UPDATE: I think I may have discovered the reason why the gct.prg faisl on the old 326298 mainboards. Will elaborate in next reply.

[MarkMccabe]

Why the Game controller Test program gct.prg does not work on the oldest revision C64 revision 326298 according to google's AI.
The early Commodore 64 326298 mainboard has a faulty reset circuit that prevents the gct.prg game controller test program from running correctly. This is because the reset line is controlled by a 556 timer chip in a way that interferes with software that attempts to access it.
On these early boards, the 556 timer pulls the reset line high and supplies a strong current to keep it there. This prevents any external device, including certain diagnostic cartridges or software that relies on the reset line, from pulling the line low to initiate a reset.
The 326298 mainboard has a known issue with the reset function. Programs or cartridges that trigger a reset, like GCT.prg, can cause the computer to hang, display a blank blue screen, or behave erratically.

Now I don't know enough about the gct.prg to determine if this is true or hogwash but I suspect it is not true.

[frntc]

GCT cannot pull the reset line, such "AI answers" are nonsense ;-) (use any other program to read paddle values, also the config tool is fine). Now if you connect the diagnostic harness to the joystick ports, what values does it read?

[MarkMccabe]

I did not think there was anyway for a software reset to trigger the 566 timer. :)
Here are the results using the control port dongles on 4 different SidKicks installed on my NTSC 326298 5 pin video C64 (all set to a 6581, no second sound chip) and the one that thinks it is a drean:
SIDKICK4 X=3B Y=3B with paddles set to Raw or Median
X=5B Y=5B with paddles Smooth
SIDKICK3 X=28 Y=27 with paddles set to Raw or Median
X=48 Y=47 with paddles Smooth
SIDKICK2 X=28 Y=29 with paddles set to Raw or Median
X=47 Y=48 with paddles Smooth
The Drean X=3B Y=3B with paddles set to Raw, Median, Smooth, or "Tings"???

Without the dongles installed they were all FF
I used these dongles from PCB Way

[MarkMccabe]

I also ran the test using a cheesy little basic program on the same NTSC 326298 5 pin video C64, with the dongle connected and had these results (in decimal):
Sidkick4 Port1 X: 96 Port1 Y: 97 Port2 X: 95 Port2 Y: 98
Sidkick3 Port1 X: 77 Port1 Y: 77 Port2 X: 79 Port2 Y: 78
Sidkick3 Port1 X: 77 Port1 Y: 78 Port2 X: 77 Port2 Y: 78
ARMSID Port1 X: 82 Port1 Y: 82 Port2 X: 81 Port2 Y: 82
6581SID Port1 X: 97 Port1 Y: 100 Port2 X: 97 Port2 Y: 100

The whacky Drean one...
Drean Port1 X: 98 Port1 Y: 100 Port2 X: 99 Port2 Y: 98

All were 255 without the dongle installed.
Attached is a copy of the simple BASIC program I used..
Basic Program.txt

[frntc]

The diagnostic carts test for paddle values between 80 and 120 ($50 and $78), i.e. the SKpico values are close but outside that range. I'd try offsets of 20, 30, 40 and see if that works.

[MarkMccabe]

Thank you for the help!!!