Xbox One PadHack REAL triggers fix

[tegezee]

I have a MadCatz SoulCalibur V Fightstick that I padhacked with xbox one. I had everything working flawlessly except LT & RT. I've tried everything, resistors (10ohm, 100ohm, no resistors, & different points. I've since taken the entire stick back to stock to try and figure it out. I've tried the mod with the top pcb's & couldn't get anything to work with the PS pcb connected. Below is some pics of my setup & fightstick. I got my solder points from HERE & HERE I only want it to work for xbox one!!!

Thanks in advanced for any help...

PS fightstick


The inside


Top pcb


Another pic of the top pcb


Xbone controller (TOP) solder points


More xbone controller (TOP) solder points


Xbone controller (BOTTOM LT) solder point


More xbone controller (BOTTOM RT) solder points

I

[RDC]

Forget wiring the Trigger up to any PCB or the like, as nothing that's out now is made to work with the new Trigger setup.

You can duplicate the Trigger by wiring up your new button to the LT/RT line, and then the other side of each new button goes to a 10ohm Resistor, then Ground.



If that doesn't work, then you've probably killed the Trigger inputs on the MCU, because I've done this and it does work.

[scarface4200]

Do you have a picture of where the lt/rt line is? Thanks!!

[RDC]

Yes, and if you look at the pinned topics in this section you'll see them.

[snowcolt17]

got a little too close didnt you..

[tegezee]

I got it solved any-who & it works flawlessly. All thanks go to GrammatonKlerik for this solution.

As promised here is the Trigger fix. 

The gist of the fix is to tie a 3.3v source to a 10K resistor.  You then splice the other side of the resistor to two wires.  One wire solders to the bottom pads of the RC filter and the other wire goes out to the signal side of your RT/LT buttons.  The two IC's on the PCB make a real nice non-conductive pad to hot-glue the 10K resistors to.

Although I removed the Sensor, the Bypass Cap, the RC filter and the cubes, the ONLY required things to remove for this fix is the RC filter, i.e. one tiny cap and one tiny resistor.  The easiest and safest way to remove these is to lie your soldering iron's tip across the top of the element so that it heats both sides simultaneously.  Make sure you tin your tip with a bit of solder to get the best heat transfer.  If you do it right, the little guys will move off the pads quite easily.  If you have some solder wick I recommend cleaning the pads off with the wick to get the surface nice and clean.  When that's done, just put a small dab of solder on the two bottom pads of the RC filter as that is where you'll tie to.

A word to the wise, even though I removed my cubes I recommend that you do NOT, there is no real reason to unless you do something stupid like I did which was cut the little white straps that go on either side of the sticks.  Don't cut those!  Those tie directly to the potentiometer in the cube and when cut will easily move causing all sorts of weird directions to happen.   If you're the type that likes a challenge or likes the clean look with no cubes then use 5K resistors, not 10K.   Although the potentiometers in the cubes are indeed 10K, that's measured across the entire pot but we are replacing the pot with two discrete resistors so we should be using 5K to mimic exactly what the cubes are doing.  If you don't have any 5K's lying around don't worry about it, it doesn't really seem to make that much of a difference but if you're going to go out to buy some resistors you might as well get the 5K's.

http://forums.shoryuken.com/discussion/comment/8591784/#Comment_8591784

LT


RT

[Modded Matt]

do not modify your OP when people have responded to your request. I dont have time to look over the entire post. I restored it so the thread makes since. All I am going to say is there are some very..questionable practices going on here.

for newbies, dont glue to the main IC's....  they are heat sensitive.  the xbox one controller is filled with open space.

[tegezee]

do not modify your OP when people have responded to your request. I dont have time to look over the entire post. I restored it so the thread makes since. All I am going to say is there are some very..questionable practices going on here.

for newbies, dont glue to the main IC's....  they are heat sensitive.  the xbox one controller is filled with open space.

Then glue it somewhere else (I did), but hell it works like a charm.

[GrammatonKlerik]

All I am going to say is there are some very..questionable practices going on here. for newbies, dont glue to the main IC's....  they are heat sensitive.  the xbox one controller is filled with open space.

I don't think it's questionable at all.  Obviously these IC's had to survive a reflow profile (see below) which peaks at 260 *C and that's after enduring a 60 soak at almost 200 *C.  Consumer grade hot glue melts at ~127 *C and industrial high-temp glue melts at ~194 *C; both well below the reflow temp of the solder.  You are, of course, free to glue it to where ever your heart desires.

In either case, this fix is very stable as it turns what used to be an actively sourced trigger into a standard pull-down trigger.

[GhoSt]

I don't think it's questionable at all.  Obviously these IC's had to survive a reflow profile (see below) which peaks at 260 *C and that's after enduring a 60 soak at almost 200 *C.  Consumer grade hot glue melts at ~127 *C and industrial high-temp glue melts at ~194 *C; both well below the reflow temp of the solder.  You are, of course, free to glue it to where ever your heart desires.

Keep in mind I have personally ripped components straight off of PCB's removing hot glue, admittedly not IC's, but its good practice to make things simple and safe to remove.

[Modded Matt]

There are so many questionable practices in this post. your comparing a reflow of the whole board, to applying direct heat to only a small portion of the IC. not the same. While all of this may work fine without any issues, gluing to the IC's is bad, gluing your solder joints on top of the joint or as close to it as you have is bad, standing the resistors up off the board is bad, your wiring practices are bad. you may have fixed one problem, but have opened up other potential issues.

bad is not a great word here because there is nothing actually wrong with it, its just poor practice.

[RDC]

There is no point in making that huge mess or any reason at all to remove any components for anyone wanting to duplicate the Triggers, or make an arcade stick with the XB1 controller.

That method does not turn it into any kind of pull down Trigger at all, as that's what it already is. The 10k Resistors are needed because the RC filter is removed, which takes the Hall sensor out of the circuit and that's what kept the Trigger lines pulled Hi, and only when the MCU needed. Removing the RC filter does nothing at all but take the Hall Sensor out of the circuit, which has an output of ~1.6v, not the 3.3v that the 10k Pull-Up is letting in there. That method is flawed and more work than necessary, not to mention possibly damaging to the Analog lines over time.


As has already been stated in the first reply to this topic, just solder the new buttons to the LT and RT lines, then the other side of them to Ground, done. The 10ohm Resistor is optional, and only 1 of them needs to be used if it is. No removing components, half the wiring, and no Resistor unless you want to use it.

[geekonaut]

Keep in mind I have personally ripped components straight off of PCB's removing hot glue, admittedly not IC's, but its good practice to make things simple and safe to remove.

I'm an engineer who has used hot glue extensively in literally hundreds if not thousands of prototypes over the course of 12+ years. I've never destroyed an IC with hot glue (though I've destroyed my share of ICs through plenty of other means). These ICs are capable of taking these temperatures for short periods of time if they are not powered. (Yes, the chips do reach some of these temperatures while going through reflow. I've seen temp readings on these chips while in reflow.) Occasionally you'll run into odd things like NAND chips losing their memory while reaching these temperatures (that was a pain to deal with), but these are the exception not the rule.

I've also learned techniques to getting hot glue off without destroying the board itself. Rubbing alcohol actually makes hot glue release from whatever it's attached to. Take a Q-Tip soaked in alcohol, and press it into where the hot glue meets the other material. Use tweezers, or something with a corner to begin easing the hot glue away from the surface. Apply the alcohol periodically as you peel the hot glue off. I can normally remove hot glue that's holding 30 AWG wire in place that was soldered to a via, without ripping the wire off.

If used carefully, hot glue is strong enough to protect your modifications. If treated carefully, it is also removable from even delicate mods. As always, your mileage may vary. I've seen plenty of other people destroy prototypes using other very common techniques that work well for other people. The biggest skill in all of electronics is patience, and this is no difference. Just be patient, and, in my experience, use hot glue often.

[GrammatonKlerik]

your comparing a reflow of the whole board, to applying direct heat to only a small portion of the IC. not the same.  While all of this may work fine without any issues, gluing to the IC's is bad

If you don't think every single component on that board is going to reach at least 127 *C after soaking in an oven for 3 minutes at temperatures greater than 150 *C then we will have to agree to disagree because there isn't much more to say.  It really is a non-issue; geekonaut offers up the same findings.

gluing your solder joints on top of the joint or as close to it as you have is bad

That rule of thumb applies when your solder joint is on a pad.  The rational behind it is two-fold. The first is that thermal stress will develop as the hot glue contracts as it cools; this stress can be enough to pull the pad away from the trace that it's tied to or at the very least make it very fragile.  The second is that you've now put the focal point of any torque applied to the wire directly over the pad which will easily rip it off in large part due to the thermal stress.  When you are gluing to a package like I am, this is another non-issue.

standing the resistors up off the board is bad

By that assessment, every single tantalum and/or electrolytic cap I've seen on a PCB is "bad."  I mounted the resistor vertical because the pitch of the through holes is less than the width of the resistor.

your wiring practices are bad.

Why?  I think it would be a lot more helpful if you explain what you would consider the "good" way  so that we can all learn.  Saying something is "bad" without offering up a solution helps nobody.

you may have fixed one problem, but have opened up other potential issues.

I'm not foolish and/or arrogant enough to think that I'm infallible, but I think I offered up enough justification why this is not the case.  If you don't agree, please let us know how you would do it.

That method does not turn it into any kind of pull down Trigger at all, as that's what it already is.

I believe you're confusing active-low with pull-down.  The trigger as it comes from MS is indeed active-low and my fix did not change that fact.  What my fix does is turn what used to be an actively sourced trigger into a passive pull-down trigger much like all the other buttons on the pad. 

Removing the RC filter does nothing at all but take the Hall Sensor out of the circuit, which has an output of ~1.6v, not the 3.3v that the 10k Pull-Up is letting in there. That method is flawed and more work than necessary, not to mention possibly damaging to the Analog lines over time.

Correct, removing the filter was necessary to remove the hall sensor from the circuit.  Using the 3.3V source, although not identical to the hall sensor's VOH, was intentional and is not flawed as the I/O pins on the freescale MCU are rated up to 5V; so no possibility of damaging the analog input pins.  I originally used the 1.8v source from the analog cubes as that is much closer to the 1.65V sensor output, but I noticed when the circuit was pulled to ground via a trigger press that the 1.8v source dropped a few hundred millivolts; it just couldn't handle the load.  Switching over to the 3.3V prevented this behavior.

As has already been stated in the first reply to this topic, just solder the new buttons to the LT and RT lines, then the other side of them to Ground, done. The 10ohm Resistor is optional, and only 1 of them needs to be used if it is. No removing components, half the wiring, and no Resistor unless you want to use it.

Sorry RDC, but out of all the things in this thread that could potentially cause damage or harm, your recommendation of how to implement the new LT/RT buttons is the worst offender and I'm fairly certain is a large reason people are seeing premature failure of their sensors.

These hall sensors are of the linear/ratiometric type which have actively driven outputs via an op-amp/comparator.  Note that this is in contrast to the other two types of hall sensors, switch and latch, which are NOT actively driven but rather use an open collector output. By pulling the circuit to ground you are asking the sensor to drive a 100 (or 110) ohm load which is HUGE compared to the MEGAohm+ MCU pin it normally drives.  Driving a 110ohm load is asking the sensor to source ~15mA, that is A LOT.  I don't know about you, but I've looked at many linear sensor data sheets from mouser,digikey & newark and I've not found a single sensor in a SOT-23 package who's ABSOLUTE MAXIMUM output rating was greater than 10mA.  In fact the majority of the sensors were rated to no more than 3-5mA, only one was rated to 10mA; so at BEST you are over-driving the sensor by 1.5x leading to its failure.   Go ahead and verify this, you'll see that I'm correct.

Without the datasheet for the exact hall sensor MS uses, removing the sensor from the circuit via my fix is the only way to guarantee you are not going to get a trigger that fails.  It's also the only way to revive a pad who's sensor has failed using your method which was the impetus for my fix (as mine failed too.)  If you strongly feel that this isn't necessary, then I would suggest you revise your circuit to use a mandatory 330ohm resistor in lieu of the 10 ohm.   With a 330ohm resistor, the sensor will see a 430 ohm load (including 100 filter resistor) when pulled to ground.  This load reduces the source current to ~3.8mA which still might be outside of its tolerance but probably not by much.   This lighter load also means that the MCU will only see 1.27V when the trigger is pushed but that's actually OK because the MCU's threshold voltage for a trigger pull seems to be about 1.31V;  I was able to locate this threshold value by using a trim pot.

Hopefully you can see that I put a lot of thought into my fix before I posted it (hell, I didn't even post it here) and I have a fairly good idea of what I'm doing.  I'm not here to play who's MSEE e-peen is bigger, I just want to contribute to the cause and I believe I've done so.  I have a great deal of respect for the work you guys have done and without RDC's scans I would not have been able to figure out the fix as quickly as I did.

[GhoSt]

This is being watched, just make sure you respect other members (all of you) when you offer up feedback or constructive criticism.
Lets keep if friendly . It'll be great if all parties can agree/confirm on the raw information provided to ensure we are not offering any misinformation.
btw, 10inch e-peen.
-GhoSt.