Suncom SFS and Talon restoration and mods

Now I see the encoders as button presses, is it correct?

Crawling Chaos

Debolestis, I finished the retrofit of the Suncom pots. I done slightly different from the original DocFlyer recipe. I kept the upper metal plate of the pot (the face with two holes and the shaft slot) and used two layers of plastic sheet, for I do not found one tick enough to fill the gap between the magnet housing and the plastic encasing. It seems to work quite alright.

Crawling Chaos

Now I see the encoders as button presses, is it correct?

https://i.postimg.cc/7YTmgsx0/Anota-o-2020-06-01-040604.png

 

Crawling Chaos

Yes, encoder movement press sequentially one button in each direction.

I made a video of a test of my Suncom F15e retrofit. Although the video is very poor, it shows the X, Y, Z and rZ axis functioning. I think now is time to start putting things together.

I can post future steps here, on this tread, or create my own. What do you think, debolestis? Do you mind?

Does anyone with the Talon know if this is how the right hat switch is supposed to be? I know the castle and coolie are swapped, but is the right switch supposed to be gated like this?

Hi Zhutwo, in your pictures, the castle and coolie are swaped. In the F15, the coolie hat is placed on the rigth side and the castle is placed on the left side.

Crawling Chaos

Yes, I believe it is. I have a very old Suncom and I remember the hat is gated like that. Mine's probably not the same model as yours, as I have a red light where a hat is supposed to be, but the hat is gated like that.

Yes, I believe it is. I have a very old Suncom and I remember the hat is gated like that. Mine's probably not the same model as yours, as I have a red light where a hat is supposed to be, but the hat is gated like that.

I have the model you're talking about with the 1 coolie, and red light in the place of the right hat, and it's not gated. In case we're misunderstanding each other, when I say gated I mean like in a gated manual shifter:

The pictures I posted were of a Talon that I'm looking to buy, but it doesn't look stock and the seller doesn't know anything about it. Trying to get an idea of what's going on with the right hat. I know the castle and coolie are swapped, but the gates are really throwing me. It looks too clean to be a mod job.

Hi Zhutwo, in your pictures, the castle and coolie are swaped. In the F15, the coolie hat is placed on the rigth side and the castle is placed on the left side.

 

 

Crawling Chaos

Hi, I know the castle and coolie are swapped, but is the right side supposed to be gated like that? Also are you supposed to be able to swap the switch tops like that? I have the regular 1 hat SFS stick and I cannot twist it off on mine (the pictures are of a stick someone is selling).

Happens that when Suncom was released (90's) need rely basically in keyboard emulation - for Windows is just a 2 axis 2 buttons joystick, all buttons, HAT's send keypress.

I remember that Suncom was sued by Thrustmaster and probable due this can't use their FCS patent who have a POV HAT circuit.

So their HAT's are made for press keys, not for POV HAT, that have 4 buttons but 8 directions, in what 2 buttons pressed simultaneous generate a diagonal press in firmware.

So if you want that one of this two HAT work as POV HAT (in modern USB controller) need file openings in the diagonals, for pole move in this directions and press two buttons simultaneous.

http://theseger.com/projects/2014/04/converting-old-joysticks/

If you use ALPS RKJX... switches (used by VKB, VPC) that have 8 physical directions, can use in both HAT's, but for POV HAT need an special PCB for emulate that have only 4 buttons like a normal POV HAT (but 8 directions), check back in this topic Sergeype posts.

How strange, my SFS stick can do 8 directions stock.

Debolestis, I finished the retrofit of the Suncom pots. I done slightly different from the original DocFlyer recipe. I kept the upper metal plate of the pot (the face with two holes and the shaft slot) and used two layers of plastic sheet, for I do not found one tick enough to fill the gap between the magnet housing and the plastic encasing. It seems to work quite alright.

Do you have the DocFlyer mod hooked up electronically and working yet? How is the performance for you?

Hi, I know the castle and coolie are swapped, but is the right side supposed to be gated like that? Also are you supposed to be able to swap the switch tops like that? I have the regular 1 hat SFS stick and I cannot twist it off on mine (the pictures are of a stick someone is selling).

Oh, now I got it. Yes, the right side is gated. Mine is drilled to work with an analog thumbstick. They can be swapped easily, since they share the same footprint. I don't know how it is in the SFS.

Crawling Chaos

Do you have the DocFlyer mod hooked up electronically and working yet? How is the performance for you?

Yes. In fact I finished the retrofit and I am trying to find time to share it in a topic of my own. Well, I am still struggling with its deadzone, bu as I said, I don't have much time to spend debugging it. But it may have mechanical issues, as pointed by Debolestis.

Crawling Chaos

Yes. In fact I finished the retrofit and I am trying to find time to share it in a topic of my own. Well, I am still struggling with its deadzone, bu as I said, I don't have much time to spend debugging it. But it may have mechanical issues, as pointed by Debolestis.

 

 

Crawling Chaos

When I attempted it, the plastic knob piece would get rotated easily and cause excess friction. It's pretty hard to recreate the smoothness of potentiometer rails without factory machining. I'm currently attempting a solution based on putting an axle through the black potentiometer holder and just having the knob tip act as a lever.

Developed this method of mounting analog stick for hat switch to Suncom stick assembly without need for 3d printer access:

-Attach a machine screw to the underside of the stickbox using epoxy; try to get it as level as you can and make sure to envelop the entire screw head with a thick layer of epoxy so as to not rely on the metal to plastic adhesion.

-Put the screw through the hole in the PCB and secure with a rubber washer followed by a metal nut. The rubber washer will deform under pressure and level out a slightly crooked screw as long as your contact points above are uniform.

-If you want you can add a metal or nylon washer between the rubber washer and nut but you may run into clearance issues with the plastic structures on the stick that will require excavating.

-You can also add material under the stickbox to increase your contact patch, but the mounting pins (not the electrical contacts) provide enough resistance to the directions of force when being used. They also are conveniently level.

Going to order some of this style of potentiometer and see if it can be used as a rail to mount magnets and therefore provide smoother action in the DocFlyer mod.

zhutwo

This Suncom "USB'risation" keep the original pot's, and axes response (in 10 bits) is good.

Author create a custom code for Arduino that set pinky button as momentary shift for other buttons and POV HAT.

And little button under POV HAT as well base buttons as "modes" (5).

So can use ~80 buttons.

When I attempted it, the plastic knob piece would get rotated easily and cause excess friction. It's pretty hard to recreate the smoothness of potentiometer rails without factory machining. I'm currently attempting a solution based on putting an axle through the black potentiometer holder and just having the knob tip act as a lever.

In my case, the movement is smooth, but there is a HUGE deadzone. It might need a stiffer spring.

Crawling Chaos

zhutwo

 

This Suncom "USB'risation" keep the original pot's, and axes response (in 10 bits) is good.

I have it currently working fine with original pots using MMJoy2. The problem is it is impossible to find replacement pots in the same dimensions as the original, so a hall sensor solution is needed for the long term.

In my case, the movement is smooth, but there is a HUGE deadzone. It might need a stiffer spring.

 

 

Crawling Chaos

I think this has to do with the small amount of play inside the gimbal holes with the plastic nibs. You could try filling the holes with a generous amount of the thickest grease you can find to try to lessen some of the tolerance. I am noticing a lot of deadzone with the stock pots as well. The raw value tends to leave +/- 20 in the direction it came from when returning to center, which I am certain is due to the tolerance inside the gimbal hole.

I PMed GranBichus on the XPlane forum and he said he mounted the magnets directly to the gimbal, and attached the hall sensors to the potentiometer casings. This seems promising and should solve the issue of deadzone by eliminating the use of the nibs entirely. Some excavating of either the gimbal or base will be necessary in order to make clearance for the magnets however.

I PMed GranBichus on the XPlane forum and he said he mounted the magnets directly to the gimbal, and attached the hall sensors to the potentiometer casings.

Curious about this mod in Xplane is that the guy made a well detailed , illustrated topic, but don't show a unique picture of Hall sensor/magnet placement in gimbal.

Curious about this mod in Xplane is that the guy made a well detailed , illustrated topic, but don't show a unique picture of Hall sensor/magnet placement in gimbal.

Yeah, and then in that thread I tried to get his attention 3 times over a period of 3 years to find out how he placed his magnets/sensor and he missed my reply every time. Finally after 12 years I PM him and he just tells me how he did it. It is such an obvious solution when you think about it.

The question is:

... mounted the magnets directly to the gimbal,

On Suncom gimbal the magnets will not turn or sliding left <> right, or move up <> down, but oscillate in X, Y axes in front the sensor.

Installing Hall sensor in CH pedal (turning magnet) I discover that plastic pivot induce a small lateral movement on magnet that disturb sensor readings, requiring install bearing.

In that "Saturn ring" Suncom gimbal I think this will be accentuated.

In T.16KM, Warthog, X-56, the magnet oscillate, but the sensor is "3D".

Anyway post the results of your try. :thumbup:

I think it's like this:

With cylinder magnets oriented this way, the rotational disturbance between axis should not matter too much because the magnetic field is radially symmetrical from the reference point of the sensor. Crucially, if you look inside, the metal ball pivot point is embedded in the gimbal so that it is aligned with the axes.

Progress:

In order to make clearance for the magnet, these plastic gates must be removed:

After:

The alternative is to remove some of the receptacle ends, but that would make it very difficult to restore to stock function. I did slightly sand down the one with the slightly rounded end to make it easier to attach the magnets flat:

Note that this receptacle sticks out slightly more than the other (1-2mm difference). I'm not sure what the reason for this was in the stock design, but it may need to be compensated for slightly in hall sensor positioning.

Next step is to mount the hall sensors. This will require quite a bit of trial and error to get a good range of output values, so it's handy to have the gimbal detached from the rest of the stick/base. An ideal solution would allow for easy height/depth repositioning of the sensor while keeping it at a right angle.