Lesson DIY Force Feedback Flight Yoke

Thanks for that. I've ordered the 1" to allow for variance.

One of the timing belts is unfortunately sold out.
A 6R23M167060 (501)
http://shop.sdp-si.com/catalog/product/?id=A_6R23M167060#

I did find a comparable one, although I did read that not all comparable belts are made alike. Some require sanding to decrease the rigidity as I understand.
https://www.bbman.com/catalog/product/501-3M-06

Which did you end up ordering?

You are correct. Not sure which belt is sold out,, Is it the long one or the shorter one.. The longer one you actually have to cut (no longer forms a circle).. The shorter one need to be the correct size.. It was the long belt which I purchased from a guy on ebay. The designer of the ffb yoke has it wrapped at too tight an angle and it wasted a lot of torque from the motor.. The 3mm pitch is not very common. He did not use the 2mm pitched belts I believe because the accuracy of the 3d printed gears could not be printed.. With todays resin/laser printers I think it would be possible.. Yes it was me who ended up shaving the belt to make it as thin as possible and hence more flexible to fit around the smaller gears..

The one that is sold out would be the longer of two (501mm) which yes, is cut for the elevator function. If you found the longer of the two much too tight, may I ask what you substituted for it? The shorter of the two was alright?
I will be doing the resistor mod as well so I guess you went with something that helps with the added torque.

Hopefully I can get both belts off ebay where the shipping is cheaper than the engineering supply companies I linked above. The shipping for the belts far outweighs the item cost so I'm not in much of a position to experiment unfortunately with those

I did order the shorter belt from the guy in England. It has been about 2 years ago so maybe he was out of stock of the long belt even back then. It is not that the belt was too tight it was the fact that it was not flexible enough to go around the radius of the smaller gears without causing a bit of binding. Shaving the back of the belt down to the kevlar or fiberglass belting did not kill its strength but did make it more flexible.. You will find you will actually have to tension the belt pretty tight. I slightly modded the front belt mount to allow it to set a finer tension . You will see what I mean when you do your final adjustments. I got the material for the long belt from China on Ebay. The seller is gone but this was the description. GT Arc 3GT Open Synchronou Timing Belt Width 3/6/9/15mm Neoprene Small Backlash.. I think you want the 6mm width.

I finished my vrflightsim DIY FFB Yoke some months ago. Did anybody find a solution for the hardcoded deathzone of the Sidewinder Force Feedback 2 Stick? it is really annoying.

What is a deathzone. Do you mean deadzone?? There should not be any. What driver software are you using.
Is it xpforce? I use my diy yoke on a daily basis and have never had a problem.

Sure deadzone
Yes, a deadzone. like 1.5 cm yoke travel on pitch axis, and maybe 5 degrees of yoke rotation on roll axis. cant get rid of it. it is not like Play or something with the belts or gears, outside of middle positions i get instand reaction from movements in both directions (left right, in/ out). DIView doesnt show any change of the pot values during travel through these deadzones. and at least the Bochem Multi-rotation pot for the roll axis has no build- in middle position without resistance change during movement. The old Win-XP Sidewinder software had an adjustement function for the deadzone,but it makes the deadzone only smaler. and after changing back to win 10 the adjustement is gone. and it is definetly not a "did you try to calibrate" kind of problem.

for the PITCH axis

I did some measuring today on my yoke. Deadzone on pitch is 4mm. Deadzone on roll is +- 3deg measured.
This was done using a electronic protractor.
Here is a screenshot of the calibration screen used to do measurements . Note be sure to click on raw data.
I set the yoke to zero zero on the x/y axis.

I did the Callibration a dozen times, still thought i made a mistake. But the Yoke, and my spare SWFFB2 still refused to collaborate. But now i know i am not alone wtih this Problem. Looks like i need another FFB Controller for the Yoke Hardware. The Deadzone makes Final Approaches in the new DC-6 , lets say, a "little bit" harder. At least for me.

I understand the calibration attempts. What I am asking is what is the raw data saying.
Set the pitch to zero (raw data )and the slowly move the yoke in/out until it reads a value greater or less than zero.
I measured this pitch deadzone at 4mm you know +-2mm.
With the 10k pot it should be very similar to mine.
Going to a 5k pot should tighten it up a bit more although I personally have not tried it.
I am sure the stock pot value is indeed 5K. I don';t know why the designer chose a 10k value.

Oh sorry i forgot ot put the numbers in here:
x goes from -512 to 511
y goes from -512 to 511

I dont think that a different pot will change anything. The Force Feedback wants to center the stick a a given (hard coded) value. Try to callibrate The stick "off center", and the PC software will show the "new" middle position as ok. But the stick centers at given resistance values.

All right.. I have a few hours of board level modding experience with the Microsoft Force Feedback 2.
I am pretty familiar with its electronics.
I will ask once again.
Pitch only test
Remove power to the motors so They are not pulling or pushing the yoke.
Go to the config screen and set the raw data to zero.(by pulling or pushing on the yoke)
Place a mark on the yoke shaft at the point where it lines up with the front of the housing.
Push the yoke slowly until the raw count # goes to 1 or slightly more. 2-5 is still ok
Make a mark on the yoke shaft again.
Pull the yoke until it comes back to zero and through to -1,-2, or so.
Mark the shaft again.
The distance between the 3 marks is your yokes deadzone.
The deadzone distance on my yoke is about 5mm +_

The Distance between the first and third mark is 8,24 mm

so 1,5 cm was a lie. i am a liar was just estimating without propper science

I would say the electronics are ok. Did you do the resistor modification on the circuit board to increase the torque? On the pitch timing belt have you thinned the belt by shaving half of the rubber off the back of the belt. And lastly on the big printed roll axis gear did you use a zip tie to secure the belt to the gear?

I made a short Video showing movement of my yoke. made some Belt modifications. The Smal Roll axis Belt is modified to an open version, each end atttched with a zip tie. The Pitch Belt has a slightly different Routing with two Pulleys guidance, and both ends at the same plane, forward end in a custom Belt Tensioner.
https://youtu.be/tqUGJMpgSUI

EDIT:
Both Belts are from the same Origin, just cut into size.
EDIT No.2:
No Resistor Mod.

Thank you for the video.
I like your pitch axis belt tensioner. I have one on mine and it is a necessary addition to the original design.
The roll axis belt is an excellent idea since that closed belt is often not available.
Do you have the program ForceTest .exe?
It is a utility program for the FFB2 joystick. (link below)
Everything looks really good on your build.
The pitch axis is the weak link in the design.
Its fault is that it does not return to the zero position (elevator neutral)
This is greatly helped by doing the resistor mod.
Thinning the pitch belt makes a big difference also.
This is accomplished by taking a dremel mototool with a grinding stone attachment and carefully removing the rubber layer on the back of the belt until you just can see the kevlar reinforcing threads in the center of the belt.
This makes the belt much more pliable and can go around the small gear without inducing so much drag.

The original designer used 3mm pitch belt which helped induce this problem. he said he did this because the 3d printed gears could not be printed accurately enough for 2mm pitch timing belts.
With todays resin printers such gears could easily be made.
But the design is actually getting quite old and the designer seems to of abandoned the project.

Sorry for the long post.
Please reply if you have Force Test. exe installed on your PC.



https://www.dirks-software.ca/downloads/ForceTest.exe

My first "Pre Alpha" version used the original routing of the pitch Belt. My main Problem was not Belt-slip (as mentioned in the Inventors Forum Thread), it was friction in the Linear Bearings caused by off-center load of the forward part of the Belt. Having both directions pulling in the same plane fixed the friction Problem. My Yoke returns reliably to the center position +this freaking deadzone +maybe 1-2 mm. The Resistor modification is the next project after changing the PLA (everything white in the Video) against PETG parts. PLA is so weak, that the roll belt chaffs against the inner side of the big black gear because the Belt tension puts the motorbracket to an angle.
Grinding the pitch belt is not an option for me because i am surely not able to grind it as smoth as the original back surface. And an uneven surface must go over two pulleys. would annoy me, i guess. My Belts have a very thin Aramid (Is that also the name in english?) layer and are very "not so rigid". I can easyly bend the belt back to back.
The Force-Test software works well with the yoke. It moves on its own in every position.