PC-Sprint negative: joysticks don't function

[Chuckphd53]

I looked at the jr joystick schems, this is the very early and typical 555 type timer for joystick interface,
I have not looked at the BIOS, but will extrapolate via the Schems, that this may be item of interest with reading the joystick.

If you had made a basic program reading the joystick, then I would like to know if ONLY reading the position every 1 sec or so, in other words slow it down to see if the actual timing of the Read X/Y pulse to determine position changes results.

IMO the hardware is fixed, so this is not impacted by the speed change, but the actual READ timing of the Pulse change could be!

Would be fun to get more results with a basic pgm that can be manipulated..

My test unit is a pain with only the floppy attached...

[Chuckphd53]

Ok did a quick basic stick program..

x1 = stick(0)
x2 = stick(1)
print x1
print x2

with the SPRINT brd OFF
Did a RUN
I adjusted the the trim wheels on the joystick to read
58
58

Turned ON the Sprint brd and the values jumped to:'
91
89

So there is a WRITE (trigger to start the 558 chip) and READ latch the results criteria in the bios I am assuming that
affects the values....

I am sure there are folks out there that have dissected the Bios that may offer a conclusion.

[bagelswitch]

Yeah, Trixter pointed this out earlier in this thread (and I confirmed myself despite the obvious credibility of the source ) - if you write to, and then read from (repeatedly and immediately following the write) port 201, you can see that bits 2 and 3 will be on for a variable duration (according to the position of stick A) before returning to off.

The BIOS routines clearly do something processor-speed-dependent (cycle/iteration/whatever count when waiting for the high->low transition) to map these durations to reported stick position.

[Chuckphd53]

Probably a clock cycle timer/counter which is now impacted by the Sprint brd,
just gonna have to use the On /Off switch for these types of issues....

[OldComputerGuy]

Some old games that are timing dependent may also not run if the system is running too fast. Like when you try to run a 8088 or 286 era game on a fast 486 or Pentium. It rarely works as intended. So speeding up the CPU may have further reaching consequences. That was why early PC's with a turbo mode were switch selectable. Some software wouldn't run in the accelerated turbo mode.

[Trixter]

That was why early PC's with a turbo mode were switch selectable. Some software wouldn't run in the accelerated turbo mode.

The facts are true but the motivation is not: The reason turbo buttons were added wasn't for games, but for business software whose copy-protection methods wouldn't work on faster systems (even though the software itself did). Before turbo buttons were commonplace, some BIOSes would slow down the system for floppy disk reads, then speed up back once the int 13h call returned.

despite the obvious credibility of the source

Never assume I'm 100% correct. On a different platform, I contradicted my own information from as little as 2 years ago :-( so I guess my memory is starting to go? :-P