Genesis/CDX: pressing a pad button for a LONG time question

[keropi_666]

Hiya!!!!


I was playing Thunderforce and Contra on my Multimega and megadrive1, and I naturally kept pressing "B" for a long time continuously to shoot ... but I noticed that after 4-5 minutes of pressing, the button just does not react... I need to leave it, press another and re-press it...


Is this normal? Tried both a 3button and 6button genuine sega pads, they behave the same.... Is something broken?


Thanks!!!!

 

[keropi_666]

I did some testing with Gunstar Heroes , it works fine, it seems this is game depended...

But I noticed something else: with the CDX, after pressing the button for ~5mins , and leave it, there is a 2-3secs delay, meaning that the character still shoots after I release the button for 2-3secs, like there is some sort of input buffer or something and it empties... something fishy going on I think, since this only happens with the CDX, the megadrive1 I have seems fine and instant.... :? :shock:

 

[ExCyber]

This is kind of weird; it sounds like you're somehow charging/discharging a capacitor. I can't say it's impossible that such capacitance exists somewhere inside the CDX, but the part where it gets weird is that B doesn't have its own controller port pin or register bit; it's shared with A and periodically switched by the game (unless the game doesn't read A or Start during gameplay, which seems unlikely), so it seems like any capacitance there should affect both A and B.

 

[keropi_666]

hmmmmm sounds like a rough explanation... the multimega was out of use for more than a year+ , maybe it needs to work a little...

Strange things happen in old machines and caps, if they don't get used for a long time, they need to work for some time to be fully OK... I have seen it lots of times in older computers like amigas and atari....



edit: yep, something fishy...

I was playing castlevania this time and I paused it for ~5mins or so... I unpaused and B was not responding, pressed it 4-5 times. Then pressed a couple of time A, it worked the second time and then B was working too. I did pause again on purpose and after 5mins I tried C. It worked first time OK , A+B where stuck again... B does not work EVER, unless it gets unstuck with pressing A twice. I wonder what are the controller caps on the mobo:

edit2: I have tested with probotector: Pad1 has the problems described above, whereas pad2 works perfectly. No "memory" effect, no stuck , no nothing.... I also noticed that pad1 with the button B pressed, acted like I have pressed button A at the same time (change in fire mode, it has a screen visual to indicate that) and it appears that actually pressing it while B is pressed it does nothing , so all point that port1 has something wrong...

 

[keropi_666]

I have concluded that Joyport1 has problematic A+B buttons. I have opened my multimega , and I see nothing wrong visually...


some pics I took are here: http://rapidshare.com/files/293213770/k-multimega.rar


for now I have no idea what to check, maybe those black thingies next to the ports?

 

[Chilly Willy]

Those "black thingies" are EMI filters - they are there to keep the port from emitting too much EM noise to pass FCC testing. They are usually some combination of resistors, caps, and coils that "round-off" the corners of the digital signal going to/coming from the port. The sharp corners of square waves are what generate the most noise, as a square wave has quite a few harmonics in it. They shouldn't cause the issue found in the thread, but I've seen weirder things. They can usually be clipped and bridged across without affecting the operation of the device. People often do that with the EMI filters on VGA color signals (RGB lines) because it gives a sharper signal without the filtering.