|
|
Post by blighty on Feb 7, 2009 23:15:57 GMT -5
hi all,
Has anyone got any info on why the 6 button joypads don`t work with a overclocked machine? & any info on a fix?
Love the site btw.
bLiGhTy.
|
|
|
|
Post by jlf65 on Feb 8, 2009 1:38:31 GMT -5
You are not supposed to poll the joypad more than 60 times per secound (SEGA recommended doing it once a vert blank at most). If you poll the pad more often, it may or may not return proper data, depending on who made the pad. With most 6 button pads I've tested, you get garbage. So if the overclocking affects the vblank rate, it's quite likely that 6 button pads would start to fail.
If they are doing the polling based on the vblank rate, switching to PAL mode would help. Instead of 60 Hz * overclock value, you're doing 50 Hz * overclock value.
|
|
|
|
Post by Tiido on Feb 8, 2009 8:54:12 GMT -5
when overclocked, the pad can start missing TH transitions as they happen too fast, and shit happens then.... most my 6button pads get autofire effect when my MD2 is overclocked, and it does not matter if a game polls the pad once per frame or more than once.
|
|
|
|
Post by jlf65 on Feb 8, 2009 17:36:33 GMT -5
when overclocked, the pad can start missing TH transitions as they happen too fast, and shit happens then.... most my 6button pads get autofire effect when my MD2 is overclocked, and it does not matter if a game polls the pad once per frame or more than once. I agree that it's probably the width of the transition - most folks just use two nops for timing. But it DOES matter how often you poll a 6 button pad. I've only got one pad (third party) that will return good data if polled more than once per 60Hz period. The other five return garbage. Most controllers do have some kind of timeout period for the data packet. That way 3 button games always get 3 button data. 3 button pads aren't that way - I have yet to see a 3 button pad with any issues with the polling period. It's just 6 button pads that have this. |
|
|
|
Post by Tiido on Feb 9, 2009 9:10:03 GMT -5
6button pad should have a timeout, I have not really measured how long the timeout is, but its short enough on most of my pads so you can read the pad effectively at least 5 times per frame withuot any crap going on.
|
|
|
|
Post by jlf65 on Feb 9, 2009 18:34:03 GMT -5
How old are your pads? Maybe this is something recent since my pads are as old as the Genesis II. We're talking 15+ years old.  |
|
|
|
Post by Tiido on Feb 11, 2009 2:08:52 GMT -5
One original pad I have should be 15 years old (MD2 age), then there's at least 10 year old clone pad and few at least 5 year old clone pads... all of them become troublesome with some overclocking, mostly autofire effect. I haven't really done any individual testing, and don't really plan to...
|
|
