Anyone planning/did a Teensy T.Flight Hotas X mod?

[Neotec]

Well, yes, I did buy a T.Flight Hotas X, and well, it's not awesome but it's ok. What bothers me the most is the way the ADCs are used:

There are 2 10 bit ADCs on X+Y, the rest is 8 bit ADCs. Each ADC gets stripped of 25% of its range on the controller. So you have a 12.5% deazone in the center and 6.25% deadzones at maximum. This results in 768 and 192 different values for the 10 and 8 bit ADCs.

I already wrote a small DirecInput->vJoy bridge that allows me to 'gamma' correct each axis (to have an exponential value distribution) and this already adds a lot to its sensitivity and usefulness. Nevertheless, the ADCs and the internal digital deadzones on throttle and rudder are quite annoying.

What I am thinking about now it to rip out the current controller and replace it by a Teensy 3.2 (~$20). The Teensy has 16 bit ADCs with 13 bits usable. Removing the least significant bit here to reduce the noise will still result in 12 bit resolution per axis.

The plan is now to write the firmware in a way, that it's possible to configure hardware deadzones and axis curves and have them persisted in the device.

So, the software part is not what bothers me, but I did not yet open the T.Flight to check how everything's wired and if I can get away by just replacing and rewiring the controller inside the stick unit.

My question now is, has anybody experience with replacing the controller (be it a Arduino/Teensy or a BU)?

Edit: Corrected some assumptions because of wrongly configured input ...

[Loriath]

1. Used to have a T-Flight.

2. The rest you said is WAY over my head.

3. I got a better stick.

4.Profit???

[Neotec]

1. Used to have a T-Flight.

2. The rest you said is WAY over my head.

3. I got a better stick.

4.Profit???

Well, yeah, huh. If my budget (married, 2 children, 2 cats, just moved, only one earning money in the family) would have allowed to buy a 4-5 times more expensive X52/X55, then I would probably have done so ... (has been hard enough to justify this stick and a new monitor )

The profit in the end would be that for 1/3 (or so) the price of a X52/55, some grease and time I would end up with a pretty decent stick that I could use for as long I can't justify another (more expensive) HOTAS setup that suits my needs

[Loriath]

Well, yeah, huh. If my budget (married, 2 children, 2 cats, just moved, only one earning money in the family) would have allowed to buy a 4-5 times more expensive X52/X55, then I would probably have done so ... (has been hard enough to justify this stick and a new monitor )

The profit in the end would be that for 1/3 (or so) the price of a X52/55, some grease and time I would end up with a pretty decent stick that I could use for as long I can't justify another (more expensive) HOTAS setup that suits my needs

I wasn't give much of a choice in the matter. I dropped the T-Flight and it didn't work any more. So I had to do something, and no one in Switzerland had the T-Flight in stock, and couldn't give me an ETA. So I broke down and hunted till I found a sale on the X52Pro for just a tad over 100 Swiss.

With your apparent skills you probably could have fixed it. My ability in that department is "No Lights, Windows doesn't detect it, No Work." So recycle it.

[Avago-Earo]

Would the changes you propose improve things? I use a T-Flight and have never owned a HOTAS before so I'm curious to what I could be missing. Do other set ups use more Bits? Would this be an advantage in combat when wiggling the stick?

I do notice the dead zone but have learned to account for it and wonder how much of a difference it would make to my combat (skill taken into account) if I had a 'higher level' HOTAS. I say combat as that's where I imagine a more responsive stick and throttle would be an advantage as opposed to docking. I too can't afford a better HOTAS at present.

Interested to see if you do experiment and if so, what you find out.

[Gorf]

I don't like the deadzone and seem to struggle with getting the ship turned when the stick is at its limits - I thought the calibration function of Windows was supposed to sort all that out, though?

[smartroad]

Well, yeah, huh. If my budget (married, 2 children, 2 cats, just moved, only one earning money in the family) would have allowed to buy a 4-5 times more expensive X52/X55, then I would probably have done so ... (has been hard enough to justify this stick and a new monitor )

The profit in the end would be that for 1/3 (or so) the price of a X52/55, some grease and time I would end up with a pretty decent stick that I could use for as long I can't justify another (more expensive) HOTAS setup that suits my needs

Meh, just sell a child on the Imperial slave market

[Neotec]

What the Windows calibration is supposed to do is determining the minimum, maximum and rest/idle values so it has a reference. This is just to make sure that centre is centre and that games get the whole possible value range.

In the case of the T.Flight, calibration should have zero effect, as it can not possibly fail in creating correct values (because of the deadzones in the centre and at maximum/minimum).

There's this T.A.R.G.E.T software that Thrustmaster has, that works with selected sticks but not with the Hotas X. On one screenshot you can see an axis curve (well, line) that looks exactly how the T.Flight is set up internally. My guess is now, that T.A.R.G.E.T will modify the joystick controller to provide hardware deadzones and curves (on supported devices) and that the T.Flight either has this functionality disabled or features a totally different controller.

I really don't understand why they decided to have such a large deadzone in the centre for throttle and rudder (and how they did it). Cheap potentiometers may have troubles at minimum/maximum positions but never in the centre. So I can understand that they limit the maximum/minimum values ... but centre?

Yes, the modification would increase the precision (5x higher precision on X/Y, 20x higher on all other axes) and it will eliminate all deadzones (so that you definitely need to set one for the centre in-game, which is normal and expected).

And, yes, the T.Flight is still a cheap plastic stick, but replacing the controller/ADCs will definitely turn it into a way more precise instrument that will also be a lot more enjoyable.

I think I will just order a Teensy to get comfortable with the development itself (I need to check if I can have it configured as a HID Joystick and still be able to send/receive custom packets), order some grease and then just disassemble the T.Flight to identify all connections and how it's wired, draw a schematic and then decide whether it's doable or not. Having the Teensy won't hurt anyway

Still, if anybody has experience with such modifications or has even drawn a schematic (or knows of anybody who did) please speak up. Each and every input is highly appreciated.

[Zadkiel]

Each and every input is highly appreciated.

I see what you did there.

[Neotec]

So, the Teensy3 has arrived, had to setup the build system first. Everything's fine now and I'm already feeling comfortable on this device.

The Teflon spray also improved the handling of the stick a lot.

Next thing is prototyping the connections, testing what works how and only then I will start cutting and soldering.

Shouldn't be that hard to re-solder everything

Here are some images from the inside:

Throttle 1:

Throttle 2:

Stick: