Re: Joysticks typical polling freq.

From: groepaz_at_gmx.net
Date: Fri, 13 Dec 2019 16:58:00 +0100
Message-ID: <1866805.PJQU6Thypy_at_rakete>
Am Freitag, 13. Dezember 2019, 16:40:39 CET schrieb Jim Brain:
> On 12/13/2019 9:21 AM, groepaz_at_gmx.net wrote:
> > just to make it clear: its not the paddle (obviously, its just a resistor)
> > - its the SID that imposes this limit. there is a capacitor that will
> > charge in series with the resistor, and the SID counts how long that
> > takes.
> > 
> > the 512 cycles is also true for the VIC20. on the atari 800 the period is
> > much longer (~20ms), the same is true for the Amiga. On the VCS *afaik*
> > its even longer than that.
> 
> Egad, typing on my phone is atrocious... Oh well, you got the idea. 
> Groepaz is more precise, it's not the paddles itself, but the circuitry
> in the machine that takes the time. And, to add a bit more (on a real
> keyboard, this time), it does not sample for 512 cycles.  It clamps the
> POT line to ground for 256 cycles to fully discharge the external
> capacitor and then (if memory serves) allows the capacitor to charge for
> 256 cycles, counting the number of cycles (of the 256) it takes to reach
> some threshold.
> 
> Folks will correct if I am imprecise again, but I do know there are two
> states (the discharge state, the occurs for 256 cycles) and the sample
> state (which also takes 256 cycles).

thats correct :) interestingly, its different on the VIC20: the charging takes 
~440µs and the clamping to GND takes ~65µs - thats why the 1531 mouse does not 
work with it

-- 

http://hitmen.eu                 http://ar.pokefinder.org
http://vice-emu.sourceforge.net  http://magicdisk.untergrund.net

well, if you cannot do _ANYTHING_ creative in the demoscene, you end up as 
diskmag writer and write about how good others are. 
<Maali>
Received on 2020-05-29 23:55:27

Archive generated by hypermail 2.3.0.