Giving your Sided Winder Force Feedback 2 Steroids

Nice! That looks like a really great set-up!

It's been slow going on my end... I've only just got together all of the parts and equipment I need and am kind of stumped on what to do now.

Here's what my chair currently looks like with my Warthog attached:

The idea is to be able to just remove the Warthog, which is bolted to the chair with some wing nuts, and just put the sidewinder in it's place when I need it... Like so:

Obviously now the next tasks are to increase the force feedback, extend the stick, attach a new grip and get it to work with the teensy.

The resistors I got are exactly the right type. Exactly the same size and they should fit nicely on the existing ones on the Sidewinder board.

Just for future reference for anyone else that wants to attempt this, the correct form factor for the SMD resistors is 1206 and the correct desgination is 1R0 or 1R00.

I don't think I'm going to have much trouble with this and the job of soldering on the extra resistors seems pretty straight forward.

The question now is, how do I get these two to work together?

I have no idea how I should attach the wires. I've opened the stick and some of the buttons are connected to a small board in the stick while some connect to the board of the original joystick directly.

So I'm asking myself, do I rewire every button to connect to the teensy directly, or do I keep it as is using the board in the stick and the original wiring... Can the teensy deal with that???

Many of the tutorials I see use a breadboard, but I just want to use the teensy and take advantage of its small size.

Before you start soldering the stick to your teensy I highly recommend you get a multimeter. There is more then one way to do this, but if I were going to do it I would map out each switch on the stick first with the multimetere. The breadboard is nice because you can easily make changes while you write your script and test the stick but then I noticed it doesn't look like you got the teensy board with the pins so a breadboard wouldn't do much for you any ways. With the multimeter just test the switch first than solder it to the teensy board so you know you did it right. I wouldn't plan on using the board in the stick. I would rewire each button directly to the teensy.

Looks like you are on the right track though, thanks for sharing!

Alright, I'm now over the greatest hurdle, which was to get the buttons to work with the CH grip and to extend the wires!

I decided to visit my uncle, who is an electrical engineer, on the weekend and he was kind enough to help me with my project.

In the end, it was much easier than I expected, but getting the help saved my a lot of time.

We mapped out the buttons using a multimeter which was very simple. Turns out the board in the stick does nothing but to keep some of the buttons in place. In the end the wires coming out of the stick are still one for ground, and one for every button.

The two four way hats are also treated as buttons, which made things very simple.

We ended up using the teensyduino joystick example and copy and pasted a few extra buttons to the script so that we have 14 in total. So most of the work was already done.

My uncle came up with the brilliant idea of rather than to extend all of the wires, why not put the teensy inside the stick rather than the base, and just have the USB cable as the extension. The teensy++ 2.0 happened to be just small enough to fit inside the grip.

This way we were able to keep all of the buttons wired up as they were, shorten the wires and solder them to the teensy.

I'm also working on a mod to convert the pots in the Sidewinder to hall sensors.

May I ask you why you want to change the pots to hall sensors? My understanding is that MSFFB2 uses optical sensors instead of pots. To me those seem to have good accuracy and don't suffer from spiking or wear which is the usual reason to change to hall sensors.

Yesterday I tried to change one of my MSFFB2 original pots to hall sensor just to try to get rid of the deadzone in the Sidewinder. The hall sensor worked fine but didn't solve the deadzone problem. Other than that I haven't figured out why hall sensors would be preferable and I removed the whole hall sensor mod I had made. By the way I used Allegro A1302 hall sensors and those seemed to work fine with the Sidewinder.

How has your stick been holding up with the increased force mod? I have a PeterP style frankenstick with two sets of MSFFB2 and a Cougar grip and a counter weight. The Cougar grip + counter weight is so heavy that the force seems to be quite weak and I was thinking of doing your increased force mod too.

Sydost, does MSFFB2 have hardcoded deadzone? I know it has a little play on one axis, but never researched for deadzone.

I never observed any kind of deadzone on my MSFFB2.

Changing optical senzors to hall pots is a downgrade. You deliberately give up the linearity and full 1024 bits precision for an non linear movement and not a full resolution precision of hall senzor.

May I ask you why you want to change the pots to hall sensors? My understanding is that MSFFB2 uses optical sensors instead of pots. To me those seem to have good accuracy and don't suffer from spiking or wear which is the usual reason to change to hall sensors.

 

Yesterday I tried to change one of my MSFFB2 original pots to hall sensor just to try to get rid of the deadzone in the Sidewinder. The hall sensor worked fine but didn't solve the deadzone problem. Other than that I haven't figured out why hall sensors would be preferable and I removed the whole hall sensor mod I had made. By the way I used Allegro A1302 hall sensors and those seemed to work fine with the Sidewinder.

 

How has your stick been holding up with the increased force mod? I have a PeterP style frankenstick with two sets of MSFFB2 and a Cougar grip and a counter weight. The Cougar grip + counter weight is so heavy that the force seems to be quite weak and I was thinking of doing your increased force mod too.

I never observed any kind of deadzone on my MSFFB2.

 

Changing optical senzors to hall pots is a downgrade. You deliberately give up the linearity and full 1024 bits precision for an non linear movement and not a full resolution precision of hall senzor.

Its pretty interesting that you guys brought this up... I have been testing the hall sensor setup I made. Part of it might just be my lack of experience of making such a sensor but I noticed zero improvement and given the bulky mounts it was requiring, I decided to scrap the idea all together. I was unaware that the sensors used in the SWFFB were optical sensors and not pots but that makes sense given their size. There is a bit of a dead zone that I noticed gets amplified when you extend the stick on SWFFB. However, now that I have experience with an extended warthog stick, I don't think a slight dead zone is all that bad for the SWFFB. The reason being, the SWFFB doesn't have a definite precise center like the warthog so having a slight dead zone is almost necessary. The one thing I thought hall sensors would fix is the slight throbbing sensation I felt from the force feedback (after doing the resistor mod). I thought this was due to the pots flickering but now that I know its not a pot but an optical sensor I think the throbbing is due to the increase current to the motors. All electric motors have a pit of a vibration to them when under load and I think the resistor mod amplifies this. In the end, the throbbing/vibration is only noticeable when you are trying to fly level, once you actually ably load to the stick and the FFB tension takes over you don't notice it anymore. As for durability, my stick has been holding up pretty well since I did the resistor mod. It would be interesting to see what Peter P's doubled up version feels like in comparison but I suspect his method is a much more robust way of getting more force from the stick than the hack I did. I am working on a counter weight version of my stick which I think will have a whole feel of its own. The only problem is that I am also trying to use my warthog with it and I am finding that it is going to take some crazy amount of weight or a little more creativity on my part to balance everything so the stick will remain centered when the power is cut to the force feedback.

If you did the resistor mod with the doubled up mod I bet that would work really well with your cougar grip. Since I have two SWFFB this might be the approach I take if I decide I still want more force with my counterweight.

Sydost, does MSFFB2 have hardcoded deadzone? I know it has a little play on one axis, but never researched for deadzone.

Yep, it has or at least both of my two units have. I think it is hardcoded in to the drivers. My understanding is that it used to be adjustable with the software provided with the stick but unfortunately Microsoft didn't update the program for newer operating systems.

I never observed any kind of deadzone on my MSFFB2.

Hmm... well that is weird. I actually asked the simprojects.nl guy does he have that deadzone and he too said that he hasn't noticed such thing but he changed the original pots (or optical sensors). That is why I thought I should try the hall sensor mod.

Changing optical senzors to hall pots is a downgrade. You deliberately give up the linearity and full 1024 bits precision for an non linear movement and not a full resolution precision of hall senzor.

I wouldn't say it is a downgrade. I would say it is pretty much equal or it can be made equal. There is the linearity issue, but you can easily make it almost linear (something like 98% linear which is enough for my controllers) by using only the linear part of the sine curve the hall sensor produces (depends on the magnet placement). Also it is rather easy to get the full resolution (1024 in this case) from a hall effect sensor just by using strong enough magnets to oversaturate the sensor to have full resolution available on the range you intend to use.

I have done my hall sensors with this method http://www.mycockpit.org/forums/showthread.php?t=22267 . I mean that I have two magnets on the rotating axis and then I have put the sensor inside the tube.

There is a bit of a dead zone that I noticed gets amplified when you extend the stick on SWFFB. However, now that I have experience with an extended warthog stick, I don't think a slight dead zone is all that bad for the SWFFB. The reason being, the SWFFB doesn't have a definite precise center like the warthog so having a slight dead zone is almost necessary. The one thing I thought hall sensors would fix is the slight throbbing sensation I felt from the force feedback (after doing the resistor mod). I thought this was due to the pots flickering but now that I know its not a pot but an optical sensor I think the throbbing is due to the increase current to the motors. All electric motors have a pit of a vibration to them when under load and I think the resistor mod amplifies this. In the end, the throbbing/vibration is only noticeable when you are trying to fly level, once you actually ably load to the stick and the FFB tension takes over you don't notice it anymore.

I think we opened up 3 new different topics here. I didn't test for deadzone, that's why I asked. But the feeling you talk about in quoted text is another topic I wondered about. Even without joystick powered up I noticed non-linearity in travel. It is logical if you think that we use brushed motors with ?probably? 2 poles. I noticed that feeling is very different(better) when using 2 mechanisms bound together. With power you should feel the difference even more. I would like to contribute to this topic further, but my franken-PeterP-2MSFFB stick is still not finished. I use one MSFFB2 to play and have one frankenstick without handle for testing. I am working on this for almost a year but I still didn't found time to finish it.

There is a bit of a dead zone that I noticed gets amplified when you extend the stick on SWFFB. However, now that I have experience with an extended warthog stick, I don't think a slight dead zone is all that bad for the SWFFB. The reason being, the SWFFB doesn't have a definite precise center like the warthog so having a slight dead zone is almost necessary.

The deadzone isn't actually as big as it feels when flying. It seems to be about 1,5cm or half an inch measured from the top of my stick. But I find it is a bit annoying especially when flying level. For example when accelerating from low speed and trying to steadily push nose down to keep level flight there suddenly is a small part of stick travel where nothing happens. Initially I had the hall sensors on my top FFB set as roll and pitch input through the Cougar for extra resolution. I felt that it was very easy to fly precisely in the center portion of the stick when there wasn't this deadzone on the inputs. Later I found out that the Cougar is only 10-bit system like Sidewinder so I didn't get any extra resolution with that and to make it work well it would have needed a lot of adjustment of the hall sensor mod.

At the moment my next plan is to use left over axis channels on my pedal's Bodnar board for pitch and roll. I am going to take that input from the Sidewinder's own optical sensor so same position info is going to the Sidewinder and Bodnar board. That way I will have 12-bit resolution and get rid of the deadzone if I want to. So it will (hopefully) be almost perfect solution. It might be a good idea to consider doing the same thing with the Warthog as it also has quite a bit more resolution than the Sidewinder.

The reason why I am so keen to get rid of the deadzone is that I believe there would be a more defined center position with the FFB if the FFB would receive accurate position in the center of the stick.

If you did the resistor mod with the doubled up mod I bet that would work really well with your cougar grip.

I think so too. I don't have the resistors yet so need to wait a while. I am a bit put off when thinking about slodering those tiny resistors :) .

I am a bit confused by your explanation with using the extra axis channel from your pedal board. Are you talking about splitting the same optical sensor and wiring it to the the axis channel of your other board? If so how will that effect the centering force on the SWFFB2 unless you wiring it back from the your other board to complete the loop. I am under the impression that if you want the FFB to work properly the axis has to be part of the same board but perhaps I am mistaken. Does that make sense? I am probably missing something here. In either case I agree with you, if the DZ can be reduced then there would likely be a much more precise center position. Also have you notice the force almost feels like it is stepping up as it increases. This would likely be improved with a higher resolution sensor as well because I am assuming the current to the motors is a linear relation to the degree of motion from the optical sensor.

Probably my lousy explanation and this one probably ain't better. Yeah of course the best solution would be to have the SWFFB2 with better resolution too but I think that is not possible.

What I meant is that you can use a board with better resolution for inputs to DCS. That way you will get more accurate flying inputs. The force feedback will be the same as before. From DCS control setup you would set pitch and roll assignments to the warthog board and leave the SWFFB2 axis assignments empty. If you are unfamiliar with this one you can test it by plugging the SWFFB2 and Warthog to the computer and putting the assignments in DCS the way I described. You can fly with the warthog but the SWFFB2 will still do all the FFB effects even it isn't assigned as a control. In a test like this you will mainly see shaking when stalling and changing center position when trimming.

You can use any input device in DCS for pitch and roll and DCS will send the same FFB commands to all FFB devices that are connected to the computer.

So it was only a solution to have better resolution for DCS.

And yes I was thinking about splitting the same optical sensor to the other board.

Whiplash Thanks for your tutorial! I had my Stick already extended, but it really dimmed FFB. I was actually surprised how much more bang it got from 1 ohm resistors. I tend to fly with Force trim off for better feel, but when I do need trim, it holds beautifully now.

Hi, thinking of dropping my Warthog handle onto my MSFF2 stick, I love the Warthog but imho it does not compare to the feeling of flying with the MSFF2, thus thinking I might reach the best of both worlds, the biggest question now being the weight of the Warthog handle ?

Does anyone know if the weight of that handle will require me to do a mod like the OP has outlined here ??

I have nearly an entire MSFF2 in spares, mainboard, motors, cogs etc ? so I could in theory take a little risk with such a thing ?

Probably my lousy explanation and this one probably ain't better. Yeah of course the best solution would be to have the SWFFB2 with better resolution too but I think that is not possible.

 

What I meant is that you can use a board with better resolution for inputs to DCS. That way you will get more accurate flying inputs. The force feedback will be the same as before. From DCS control setup you would set pitch and roll assignments to the warthog board and leave the SWFFB2 axis assignments empty. If you are unfamiliar with this one you can test it by plugging the SWFFB2 and Warthog to the computer and putting the assignments in DCS the way I described. You can fly with the warthog but the SWFFB2 will still do all the FFB effects even it isn't assigned as a control. In a test like this you will mainly see shaking when stalling and changing center position when trimming.

 

You can use any input device in DCS for pitch and roll and DCS will send the same FFB commands to all FFB devices that are connected to the computer.

 

So it was only a solution to have better resolution for DCS.

 

 

And yes I was thinking about splitting the same optical sensor to the other board.

That is interesting, I might have to play with that. Instead of wiring it to my warthog though I would likely use a teensy board.

Whiplash Thanks for your tutorial! I had my Stick already extended, but it really dimmed FFB. I was actually surprised how much more bang it got from 1 ohm resistors. I tend to fly with Force trim off for better feel, but when I do need trim, it holds beautifully now.

Awesome, glad I could help.

Hi, thinking of dropping my Warthog handle onto my MSFF2 stick, I love the Warthog but imho it does not compare to the feeling of flying with the MSFF2, thus thinking I might reach the best of both worlds, the biggest question now being the weight of the Warthog handle ?

 

Does anyone know if the weight of that handle will require me to do a mod like the OP has outlined here ??

 

I have nearly an entire MSFF2 in spares, mainboard, motors, cogs etc ? so I could in theory take a little risk with such a thing ?

The Warthog handle is extremely heavy compared to that of the sidewinder. I'm not sure how the motors would handle it. You would definitely need a counterweight, which would require some kind of custom mounting for your stick. PeterP made a mod in which he used the motors from two sidewinders to power an extended Cougar grip, but this is much more challenging.

http://forums.eagle.ru/showthread.php?t=83814&highlight=sidewinder

Alternatively you could try getting your hands on a CH Fighterstick, which has exactly the same layout, minus the two-stage trigger. With a little bit of ingenuity you might be able to mod another trigger stage in, using one of the buttons from the Sidewinder.

The advantage to the CH grip ist that it is made of plastic and much lighter than the Warthog. It's not cheap, but you might be able to get a used one.

Here's my modded CH Combatstick (has a few less Hat switches than the Fighterstick):

You can see how the buttons are now connected to the teensy, with the USB cable running out.

I want to put the whole thing together now, but I also need to replace the light sensing diode with some sort of a switch.

Probably easy to do, but I'm unsure of how I should connect it.

The existing circuit seems to have two sensors, that are facing each other and the stick is activated when the circuit is broken. Is this correct?

Here's a couple of pictures of the front and back of the board:

The black part which is hanging off is the pot for the rudder axis of the sidewinder, which I'll no longer be needing.

Ideally, I just want to unsolder the appropriate wires and solder them to a switch that I can attach to the base.

Would anyone be kind enough to enlighten me on the correct method for doing so?

Ideally, I just want to unsolder the appropriate wires and solder them to a switch that I can attach to the base.

 

Would anyone be kind enough to enlighten me on the correct method for doing so?

I tested it earlier and if I remember correctly you should put the switch between white and grey wires. Sorry, cannot confirm it now as I don't have those wires attached at the moment. The system keeps the power on all the time when the circuit is open. When the circuit is closed it cuts the power. If you want I can test it to confirm the correct wires but it requires a bit of additional work.

Here is the wiring diagram for the stick if that helps http://www.simprojects.nl/ms_siderwinder_ff2_hack.htm

I tested it earlier and if I remember correctly you should put the switch between white and grey wires. Sorry, cannot confirm it now as I don't have those wires attached at the moment. The system keeps the power on all the time when the circuit is open. When the circuit is closed it cuts the power. If you want I can test it to confirm the correct wires but it requires a bit of additional work.

 

Here is the wiring diagram for the stick if that helps http://www.simprojects.nl/ms_siderwinder_ff2_hack.htm

Thanks! I think that makes sense. And the dark red wire is the ground I assume?

Thanks! I think that makes sense. And the dark red wire is the ground I assume?

I'd say the dark red is +5V. The pink one is ground but ground isn't used at all in the sensing diode system.

I'd say the dark red is +5V. The pink one is ground but ground isn't used at all in the sensing diode system.

Ah ok, this is the power for the light sensing diode? So I wouldn't need the dark red wire at all if I incorporate a switch? Just hook up the grey and the white and it's good to go?

Just hook up the grey and the white and it's good to go?

Yeah, that is the way I would do it. Only grey and white needed for the switch.

Dang Boris, that looks great. I'll have to look at the circuit later but what sydost said looks correct.

No need, it's done!

What Sydost said is correct. The grey and white wire and voila, the stick is now activated with a toggle switch.

I've tested it in DCS and everything works the way it's supposed to. I'm very happy with the end result! :joystick:

Thanks Whiplash and all the others that were of assistance!

Very impressive! You put that all together really quickly. Be sure to update us on how it's holding together.

Looks very good. Enjoy!