Ps3 rapid fire studying and understanding project open source

[Modded Matt]

yea, that sucks!!

I will start testing thsi weekend. other than MW2 and WAW wuld you like feedback for?

I can test it for halo...oh wait, no I guess I cant LMAO

[hyper999]

awesome thanks matt

@RDC i downloaded my disable hex from here again and saw no problem when pressing any combination of r1 and r2 ???

[RDC]

If R2 is pressed all the way here the R1 Disable 'skips' between R1 and /\ or shifts over /\ completely. Then if R1 is pressed the same thing happens on down the line, there's a bug in there somewhere.

This may not be happening on the version of controller you're using also, since the Daughter boards on them vary slightly, but it does do it here on this MSU PP 4.05 version. Really bad video of it, but it shows what's happening.

RDCXBG - hyper999R1DisableGlicth.mp4

[Hazer]

OK. Here is a simple RF at 6 SPS. This uses the counter method. Very simple toggling of the TRISIO with each counter reset. I know it works, I can see it through the PC. But two things need to be tested:

1. Is it choppy or stable in a game.
2. If you switch the power supply, will it slow down enough to notice?

I have a feeling it will be noticeable, and this test should prove it.

[RDC]

1. It's not choppy in the Demo of MW2 or WaW with the pistol.

2. It is a different speed based on USB or battery power.

[Hazer]

Here is the TMR2 test. I simulated it, but did not program it and test it. It should work, but yell at me if it does not.

This uses the TMR2 interupt at a frequency of 500 usec. So it is timed to occur inside every button pulse in the pulse train. This way, I can guarantee it will consume 20 cycles in the interupt and plan for it during the button pulses. I have added NOPs everywhere to force the timing to be perfect under each condition. The ISR will take 20 usec no matter what happens during it (I simulated it repeatedly to add the right number of NOPs) and then did the same thing during the wait loops. Using the simulator, I set breakpoints at each call of the wait loop, and added just the right amount of NOPs so that each call was 500 usec apart.

So now this new code should run independant of the power source.

[RDC]

@ Hazer - I'm not gonna yell, but I don't even know where to start with how tanked that one is..

It starts off looking like it should on the scope, for anywhere form 1 to 4 seconds or so, then it holds the entire COM2 line Hi until a button is pressed kinda hard (almost all the way) and it doesn't R/F but maybe once in every 10 presses or so and even then only a few shots, if that, then the R/F is still off time when it does work. It's the same on Battery or USB also, so you should be able to see it on the scope there without even touching a button, just fire it up and watch the COM2 line for a few seconds, it's right jacked up.


@ hyper999 - That 'bug' that's shown in the video I posted, it only happens on USB power, on Battery it's solid.

[Hazer]

Yup. Just found the big hole. I left the possibility that SHADOW could get updated between switching of the TRISIO inside the pulse train. Since I was using that to XOR TRISIO, if it changed it would leave the output on after the pulse train was finished and then stay that way since it would never return to the comparator.

I fixed that, noticed other possible bugs and tested it. The only thing that bugs me now is that from what I can see the speed is 3 shots per second, not the 6 I was calculating at (and even simulated). I will have to doulbe check the PR2 value later.

For now, its still a valid enough code to test the power theory.

[RDC]

Now that one is more like it there Hazer.

The speed is 6 SPS here, least in 2 the Demos I have anyway, and the same on both USB and Battery.


An interesting side note about that deal with pressing the button too hard and it only firing once with the R1 disable code and even when using a semi-automatic weapon when the R/F is on, is that if you're on an automatic weapon most times the R/F is probably going to make it stutter (depending on the gun of course) but it will fire at it's normal rate if the button is pressed harder, so that 'issue' kind of works for the best if you use the button properly.

[hyper999]

ok heres my disable code
you can disable any combination of buttons on com2 by changing the value of buttonsR in the asm file
it is currently set to disable r1 in the asm and hex files attatched

[Modded Matt]

Hazer's latest code:
the full game results are the same, kinda slow for MW2, but fires very smooth and consistently. The speed is not dependant on the power source.

I too had an issue with the button sensitivity, if you slam the button, it seems to skip the RF sequence completely, however a normal squeeze, or soft press of the button activates RF everytime.

this is surly on the right track

[RDC]

you can disable any combination of buttons on com2 by changing the value of buttonsR in the asm file

..and then recompiling it into a new .hex file to program onto the PIC that is.

I tried it on default (R1), then just R2 and then all buttons at the same time and it seems good, that bug is gone now too.

[geraldrubalcava]

all right, well i will use mplabs to recompile the asm.

hazer which pin does your current code use for com2? if the same i will yest.

also i have the full game of both waw and mw2, so if there is any specific gun you want me to test like, all semi-auto? or other in a video then i would be glad to do that

hyper, your code fixes the speed from batt/usb?
Post Merge: September 18, 2010, 04:32:23 PMnice work on your ps3 controller modder matt! xD lol great for debugging. truly this is going in the right direction

also RDC, are you going to be working on your code or are you looking at hypers code at the moment? just wondering the status of different codes

[Hazer]

As I have said, I believe the reason that pressing the button too hard is due to hardware limitations of the PICs output pin current. This cannot be overcome with code. To prove it, someone put a 500 Ohm resistor between COM2 and pin 6 and see if the code works no matter how hard you push.

geraldrubalcava: Com2 is connected to pin 6 of the PIC (GPIO1).

My intentions were to show how to get the comparator to work, and that it has a much faster and stable reaction time. I also wanted to find a way to overcome the power problem. My last code needs some cleanup and has redundant instructions to get rid of the bugs RDC found. The intention is to show how to get this done.

So now here is the real question: Is anyone going to even bother looking at the code structure I provided? If not, just tell me what you want the chip to do and I can give it a week and do it for you. Simply give me the design requirements. I dont care about speeds. That whole 'fixed speeds' from the X360 days was retarded. THe final code needs to be self edittable. Hell, if the battery can provide enough power, I can make it configure through RS232 and not even bother changing the PC utility I made for the X360.

So whats the endgame here? Flex? Akimbo? Macros? RS232?

I dont have a PS3 to even test on. I cannot even charge the battery. I have to plug this into my PC USB port just to do a simple verification and look at scope signals. If you want to learn how to make specific functions, tell me what they are.

[geraldrubalcava]

well the code you previously posted will help out alot of people in this forum. so i dont think is useless. It will be a great asset here. You were able to fix the speed and power buggs. thats so awesome.

I agree the hole "preset" speeds. ridiculous. but according to game informer, the ps3's most popular game is mw2 for 5 consecutive weeks. so that might be a target. hum well it would be great to have akimbo too. but can the pic12f683 do rs232? or TTL. well if its reprogrammable i belieave it would be better on the ps3 because with com1 & com2 you can control most of the controller. so this might be a good thing hum also in dealextreme there are 1800mah batterys, so power consumption is not a problem.

well after all my mumbling...
if would be good if there are different codes. ie,
1a)one with modes,
1b) editable by Pressing the Change button for 3 sec then pulling trigger to change speed(for some people that might one it, may be me too lol)

2:) well it should include akimbo.

3) a code thats reprogrammable by rs232 or other form of pc communication. with a usb to rs232 adapter it would be great, but might be difficult for certain users.

so it might be kool to have two versions for the ps3. one for solid performance and other for people who have the materials to reprogram. im not sure if you would like to do this but i just generalized users to two category.

[hyper999]

hazer i glanced at the code structure from your tmr2 code but i struggled to understand it but i will give it a propper read later
also i think rs232 is overkill for this altho i would likke to see an rs323 code as i am very interested in learning how to use it
my plan was to make use of the awesome dpad on the ps3 controller to edit the settings in a similair way to flex

[geraldrubalcava]

i agree its overkill, but! i would still like to see this happen. i was thinking of adding a special port to my controller around were the leds are at. i was thinking of something similar to the contacts of a cellphone battery, then on a charging station have the prongs like a phones contacts. (if you get what i mean) then it might be as easy as just plug it in to the station, connect it to the com though usb serial adapter (build in to the station) then boom! lol.

But a better alternative would be something similar to flex like hyper said. but the idea with the dpad would be awesome! but i still believe we need a extra tack button on the bottom of the controller to change the modes , or put them into program mode or something similar. it might be easier to do what hyper said because if you are able to read com1 line then thats a good way to reprogram. hum...

hyper, how about if the d-pad increases by 10 sps when you push up and decreases 10 by pushing down, & pressing left and right change the speed by 1.

it should have a led. so i believe we should brainstorm in how we think we might need the RF.

[hyper999]

my plan is to have it like so
hold tact then use left and right to cycle modes, down for quick on off, and up to enter programming mode

[geraldrubalcava]

actually that is a great idea! but would during gameplay will move the game? but actualy i would prefer "pressing tact for 3 sec it turns off" Quick press, turns on/switch modes. Programing mode is a quick press with a tact and on the up dpad to enter programing.

or like having 3 pre-program modes & a 4th mode with programing. and it saves to the eeprom.

hum but your idea is still awesome! just have to work if it will interrupt the gameplay, and or how fast you can turn off and on. it does matter during gameplay to not interrupt to turn it of or switch modes.

[hyper999]

yh actually that would be better just tap the tact to turn off/on

@hazer im up to speed with your code now and am gna try and intergrate it with my latest code but am struggling to think of ways i could repeatadly read a swith to add modes and such without totally destroying the timing. well actually thats not true i know how to do it but am unsure on how to write it atm but il get there eventually atm im basicly atm trying to deal with mode change, leds and switch debouncing in <500us
also i have some ideas to test and would really apreciate it if you posted your counter test source code

[geraldrubalcava]

well i would be happy to test your codes hyper. as for assembly, i started studying ... but im still lost lol

well the ides of using dpad to modify/ reprogram a mode would be great! but as i heard, ps3 code is much smaller than the 360 code, so there might even be enough room for all sorts of creative stuff.

hyper i would suggest to get mw2, or in November cod7 (black ops). because it might be the funnest game yet. tey will try to compete with mw2 multiplier and also include "special co-ops" maybe National Socialist zombies

[Hazer]

The tact button test would easily be placed during the time the comparator is waiting for the falling edge of R2.

After that, you would create  flag that lets the code know you are in program/switch mode and use that to test button presses on COM2 and COM1. In order for that to happen, you will have to change the wait loops to half the time so you are testing the COM lines in the middle of the pulses.

I am personally biased to the Flex code, so I will not offer any opinion beyond that. I will say however that changing the config through RS232 is MUCH better than any kind of onboard editting code. If you get to the point where you can pull your speeds from the EEPROM, I will show how easy it is to add the serial coms using just 2 resistors.

But first I will need to test that code with the 2.8V power supply from the controller. It should work, but testing it is better than guessing.

[hyper999]

anyway on a side note just thought i should point out my aims for this code. I am doing this project for fun and as a test of my newly learnt assembly skills so i cannot promise any deadlines. I plan to make a code like bmu and then end up with a code similair to flex but a bit more user friendly, but i will release the source code whatever i end up with. and on the subject of rs232 i would like to learn how to use r232 for a number of other unrelated projects i want to try 

Post Merge: September 20, 2010, 01:32:58 AMkk am working on making a bmu like code from hazers counter test now
Post Merge: September 20, 2010, 01:37:38 PMhere is a 1mode test i have no idea how it will work as i have no means of testing atm but connect like so
pin8 - gnd
pin6 - com2
pin5 - tact switch - gnd
pin3 - negative side of led
pin1 - power

use like so tap the switch to switch between off and 6sps, simples
and please feel free to check up on the code

[geraldrubalcava]

ok i will test your code tomorrow i already had it set up like that so this is ganna be a good base. hum well if hyper is ganna work on a bmu code for a ps3 i guess Hazer could make a rs232? i think it would be a good idea

do you want a vid for conformation that it works? well for the ps3 there might b a lot of possibilities in store.

[hyper999]

kk cool dnt really need a video though unless it does something real crazy just test that it turns on and off and that when its on tht it shoots at 6sps
and for rs232 i think the plan is for hazer to test if the controller battery can provide enough voltage (not mah ) for it to work. after that i think (hope lol) that he is going to teach me how to add it to the code when i get to the stage of loading speeds from the eeprom, which if the code i posted above works, should be within a week.
as for possibilities they are pretty much endless and the 12f683 can provide more than enough memory for this.

[geraldrubalcava]

so the code above is a able to read the speed from the eeprom? aweosme! well the ps3 controller puts out 3.7 so im not sure if it might ne enough for standard rs232. But for TTL if im not mistaking it can work with lower voltages. humm
well then i guess vids are not needed lol exacly what part of cod6 is the demo? what guns can you try out? well il get back to you on the results.

[Modded Matt]

I will do a quick test on the on/off led this evening. with my test controller (pictured above) it will only take a matter of seconds to test just about anything. I ported vdd,gnd,com1,com2, l1,l2, r1,r2 out to the test board mounted cleanly on the back of the controler. with that setup, I can wire/rewire anything you guys throw at me and the controller is still completly usable. you cant even tell its there

[hyper999]

this code doesnt load from the eeprom but if it works its easy enough to add if my code for the reading the button works i need other people to test this as i have no tact switches atm and only have the cod 6 demo i just need to know if it turns off and on by pressing the button and if when its on it still fires at 6sps

[Modded Matt]

ok, the speed is about the same, but its very choppy. the button does nothing.  the led does nothing

[geraldrubalcava]

agree the tact switch does not do anything. the led nether. its choppy but i guess that is what you would expect of 6sps. but it was a even choppy so it might just be the speed.