Modified Logitech 3D Pro, no centre detent.

Alright I started to get to the point where having the detent in the cyclic was being a pain in controlling the Gazelle and Huey. I was starting to feel seemingly random rolls to the right in the Gazelle when tracking to a point on the horizon, my thinking was the sensor on that axis was glitchy so i ordered a new 3D pro there actually a great low cost joy stick.

Well after getting some stick time in a Robinson R44 Cadet I felt the centre detent function wasn't right but the cyclic needed some well stiffness or resistance to movement but there was no sense of a "detent" in the R44's cyclic action.

So I experimented with the old but well served 3D pro and removed the detent mechanism and it's really a great feel. :thumbup:

It's not without it's drawbacks, the stick needs to be fairly vertical when the joystick is initialised otherwise the X and Y movements may not be correct in sim.

The odd thing here is after some re learning of cyclic control it becomes more natural and I'm thinking there maybe was some pilot induced oscillations caused by the control over the centre detent.

Your stick is well limp, but the dust gromet does a great job of giving the stick some resistance to movement.

The ever so close trim that the Gazelle had near centre detent at 100K++ with a slight increase in altitude is gone, "YOU CANNOT LET GO OF THE CYCLIC" you will likely die.

So now I'm relearning the control of the cyclic, next that annoying detent in my pedals and the new stick. :pilotfly:

BTW next time == more hovering. :helpsmilie:

Interesting. Solved my helo woes with a MSFFB2 off Ebay. But I only fly the Ka-50 so far. I will say when holding the trim button in the ka-50, the stick goes completely limp with zero resistance and I was initially pretty amazed at how easy it was to control this way. I was using an unmodified Extreme 3d Pro prior to that. From what I can tell, the completely limp stick is more in line with an actual cyclic than a center spring.

In the video above that cyclic looks like it's plenty loose and just sits where the pilot holds it without fighting him at all. I'm not sure if that varies between choppers. I mean.. when not holding the trim button there are strong spring sensations holding it to wherever I've trimmed which is accurate afaik.

I guess I'm saying I'm not surprised that removing your center detent made helo flying easier.

Yes the cyclic just has a little resistance to movement and yes it is like only 3 fingers to control and no detent to move across it's position is where your balancing the aircraft. Error correction mostly consists of just bilps (nudges of the position of the cyclic from balance point position) of the cyclic in the direction you want the correction to assert this also works in manoeuvring the aircraft.

After I had the introductory trial flight :thumbup: I worked on reducing the saturation (28 on cyclic x y) till I had got to none which was interesting initially in the Gazelle but more realistic.

The interesting point I found is flipping the muscle memory from a position over a detent to a position of the cyclic over the balance point of the aircraft and adjusting the rest position of of my wrist and using this position to kinda trim or neutralise the cyclic position.

Can"t answer the trim question, for the cyclic it was just a resistance to movement from what ever point it was at to what ever point you put it and so on no springyness in the action. AFAIK

Except maybe the delay between cyclic movement and aircraft reaction and predicting the correct amount of control. ;)

The other thing that helped was oculus rift, well at least in my mind I can almost fly these things. :megalol:

If anyone is interested I took some photos of the new 3D pro being modded, it's an easy mod and will void your warranty for sure. :music_whistling:

I also noted that the 3D Pro could be a cheap (I mean low cost) almost complete flight sim controller.

Oh BTW the R44 Cadet had aircon and more instruments :thumbup:

Okay the Logitech Extreme 3d Pro how to remove the centre detent form the joy stick action guide is here finally!

The mod is fairly straight forward with only the removal and re fitting of small wiring assembly and PCB connectors the only real technical requirement.

Start by removing the screws on the joystick grip noting that the screw in the pivot point of the trigger switch (PTT?) is somewhat smaller than the rest of the black screws from the joystick handle and also from the base which you need to remove latter.

https://forums.eagle.ru/attachment.php?attachmentid=160789&stc=1&d=1492314214

You’ll need to push the rubber shroud down to get to the screws in the base of the joystick handle.

https://forums.eagle.ru/attachment.php?attachmentid=160789&stc=1&d=1492314214

Carefully un-clip the tang at the bottom of the joystick handle.

https://forums.eagle.ru/attachment.php?attachmentid=160789&stc=1&d=1492314214

Carefully disconnect these connectors to seperate the the PCB.

https://forums.eagle.ru/attachment.php?attachmentid=160791&stc=1&d=1492314214

Undo the eight screws at the base of the joystick, this will allow the upper housing to come apart from the base. For the moment just loft the upper housing enough to disconnect the connector show in the picture, there is a fair bit of wiggle room here.

https://forums.eagle.ru/attachment.php?attachmentid=160792&stc=1&d=1492314376

Once connector has been disconnected lift the upper housing away from the assembly. Locate the the 4 screws that retain the the joystick yoke being careful not to damage the wires to the X and Y sensors.

https://forums.eagle.ru/attachment.php?attachmentid=160793&stc=1&d=1492314376

Lift the yoke up and away from the the base this will leave a spring and a bush remove these and set aside (you probably won’t use them again) replace the yoke and screw it back to the base.

https://forums.eagle.ru/attachment.php?attachmentid=160794&stc=1&d=1492314376

Replace the boot observe the keyways for orientation and location. Bring the upper housing into position and re connect the connector to the PCB. Check the boot is in position and the cables are in there correct position and that they wont be cleved off or something catastrophic.

https://forums.eagle.ru/attachment.php?attachmentid=160796&stc=1&d=1492314505

The upper housing also forms part of the USB cable clamp with the upper housing in place put the 8 screws in the base to attach the upper housing. Check the boot is correctly located. Then set the joystick on its base and assemble the grip assembly. Connect the assembly and locate the PCB and switches in the grip housing and reconnect the connectors. Carefully assemble the joystick grip it will key together at the top and the tang at the base will keep the assembly sort of together whilst you fit the screws, observing the location of the smaller black screw at the trigger pivot point.

https://forums.eagle.ru/attachment.php?attachmentid=160797&stc=1&d=1492314505

I found that you may need to hold the joystick centred when it's initialising on startup so if the cyclic behaves oddly restart DCS withthe cyclic centredand you should be good to go.

Need to centre the joystick when starting up.

You also need to retrain your muscle memory for cyclic control but worth it IMHO.

Enjoy! :thumbup:

As an aside the Logitech joystick would make a great low-cost sim controller I'm thinking of experimenting with extending the cables to sensors and switches wiring to utilise anti toque pedals and collective control also extend the wiring to the yoke and grip and make a cyclic. Could even extend the grip cables as well. mmmm

Extended cyclic

Okay early days but the old logitech is being frankensteined.

Here is the prototype extended joystick for cyclic. I'll wire it up and hopefully have a go at remote mounting the "slider" control and improvise some form of collective next.

The shaft from the joystick is about 15.3ish mm so I found this 20mm Gas Pipe type tubing plastic coated steel tubing with a nylon (something) lining with an inside diameter of 15.4 ish mm. :thumbup:

Plan to use a small Dektite as the dust boot and provide some resistance to movement, that's the plan.

Now for a collective.

<edit>

The seat and frame are also coming along pretty well.

Alright I've started the Collective, still very much a WIP hopefully will be able to remote mount the slider from the 3D housing and devise some linkage to operate it. :)

The threaded rod is my first attempt at a friction clutch for the collective. mmmm

Currently only using aluminium tube as bushes not sure how many flight hours I'd get before they might need replacing.

The short aluminium tube on the threaded rod is taped at M8*1.25 which works well.

The short tube is a improvised wad cutter to put a hole in the stick on felt protector to use as a friction clutch.

The Proto Collective, cont.

Okay just about need to start a V-pit thread, the collective is coming along pretty good I now need to get the friction clutches working but I'm still not 100% sure how that will work hence the prototype. :)

I'm trying to keep construction to basic materials and basic tools you don't know how many times I've been tempted to get the TIG out. :smilewink:

Bearings are a well actually just bushings of 12mm aluminium tube not sure how many flight hours I'll get but easy to replace.

https://forums.eagle.ru/attachment.php?attachmentid=161508&stc=1&d=1493501559

Plan is for 2 friction clutches either side of the assembly, I also need to set the length of the collective and having decided to keep it fairly generic that gives me some freedom in how it gets set up.

https://forums.eagle.ru/attachment.php?attachmentid=161509&stc=1&d=1493501559

After the clutches are done it's time to setup the potentiometer from the 3D pro and some form of linkage still thinking about that one however. :music_whistling:

Okay this is getting spooky anyway hope it helps others it's interesting and I'm have as much fun doing this as I am flying the Huey and Gazelle. ;)

This concept didn't work to well :(

https://forums.eagle.ru/attachment.php?attachmentid=161560&stc=1&d=1493622768

shyt happens.

https://forums.eagle.ru/attachment.php?attachmentid=161560&stc=1&d=1493622768

Okay the potentiometer from the 3D Pro used for the slider is gear reduction drive, going through an arc of about 90deg with maybe 110~120 on the external lever. My collective has an angular movement of about 55deg so this is doable. It also means that the voltage range the electronics is looking for is about 1/3 to 2/3 VCC as the pot is still ~270deg rotation.

Volia!

https://forums.eagle.ru/attachment.php?attachmentid=161561&stc=1&d=1493622768

Okay now need to do some more work on it. :thumbup:

Yes! Yes it works a treat I have both the extended cyclic and a collective using the 3D pro joystick. Now I need to set up a button box on the cyclic using the inputs available on the 3D pro joystick and put the hand grip om top of the cyclic and connect that to the main PCB.

Scratch 2 Huey's and 2 Gazelles so far, just a matter of dialling in curves and getting used to using a "real" collective. But it does make flying that much better. :thumbup:

Some pictures soon, must fly. :pilotfly:

Alright just for completeness on the 3D Pro tare down for here, I'll put together a V-SimPit build for well those on a budget using the 3D Pro for cyclic and collective whilst it has provision for torque pedals the construction could require more advanced tooling.

Turns out the approximations of angle and resistance of the potentiometer works fine for the collective and yes I have been busy re-learning how to fly or at least miss the ground on occasion both Gazelles and Huey's have been broken. :music_whistling:

With the new Alps pot wired in it needed about 100deg rotation and centering it's relative position and volia a working collective.

Add some friction and it's basically pretty good but there is a way to go to make better, just saying.

I'll move on to the V-SimPit and maybe change to other interfaces if it works happy to share, if anyone wants details of this or aspects of this build send me a PM or reply here or maybe both.

Heres the start of it.

:D

Alright I come to the conclusion that I need inputs for the VSimPit so first is a control head for the collective, which I should be able to map the button functionality of the 6 buttons on the base of the 3D Pro controller PCBA.

Mock up of the control head.

https://forums.eagle.ru/attachment.php?attachmentid=167043&stc=1&d=1502059923

Nothing fancy just some aluminium angle.

https://forums.eagle.ru/attachment.php?attachmentid=167044&stc=1&d=1502059923

Well today went well with this,..

https://forums.eagle.ru/attachmhttps://forums.eagle.ru/attachment.php?attachmentid=167135&stc=1&d=1502194721ent.php?attachmentid=167134&stc=1&d=1502194721

:thumbup:

Okay this is drifting off a little but I am still using the Logitech 3D pro electronics and the modified gimbal as cyclic which has been a great mod for sim flying also the collective I built has been somewhat transformational for sim flying.

Recent events have seen some modifications to add button box on the collective as well as a motorcycle steering damper, this actually makes the sim collective feel somewhat like that of a Robinson R44 in actual use and feel and that is not by accident. :D

I have also allowed for throttle action not yet implemented but working on it. ;)

Thanks to Hansolo for pointing these out. :thumbup:

Because of the additional load on the bushes I have investigated using bearings, again I wanted this to be doable using fairly basic hand tools (the TIG comes out in the next project. :music_whistling: )

Prototyping use of bearings instead of bushings, although I have to say the bushes are holding their own with maybe 20 plus hours flying since adding the damper to the system.

My next consideration will be the use of some form of HID interface and I am perplexed as to whether or not to use something already available Leo Bodnar etc or roll my own with Arduino or similar.

Thanks for reading and if anyone wants info PM me. Now for some more flying. :pilotfly:

Okay I have added dampers from RC cars (removed the springs) and the cyclic feels somewhat better. Pictures soon more testing with Gazelle. :joystick:

Alrighty then several hours of testing latter, this mod feels surprisingly good! Maybe not so surprising, given the damper on the collective mod turned out to be a great improvement.

I have had 4 dampers on the cyclic although they where miss matched pairs, looks like I bought out the entire towns hobby shops yes 2 of them out of stock however my thinking here is that this only needs to prove that it's functional. And it's functional or I need. :helpsmilie: Maybe both. :D

Currently I don't have the weight of a "grip" just a cyclic and three finger grip when flying the Gazelle or the Huey. This allows the cyclic to be moved from one position to another with a slight amount of force and then that new position becomes well it's new position. I may need to revise this a bit for when a grip is added however currently you can simply let go and it stays where it's set, not that you should leave the cyclic unguarded.

There is enough play in the gimbal and struts and mounts to allow fine bias of the cyclic X and Y control before moving the pistons in the dampers. The feel is closer to how it felt in the R44. :thumbup:

Awesome work! I bet building it is as much fun as flying it!

Awesome work! I bet building it is as much fun as flying it!

Yes it is fun, including 4.8Hrs in an R44 for research of course. :thumbup:

I don't know if this little nugget will help anyone or not .. but.. I have the CH Pro Rudder Pedals which have a centre (or if you prefer center) detent.

This can be annoying when trying to calmly and smoothy Hover in the Huey or any other Chopper.

So for each of my chopper mods/aircraft I have set a "3" dead zone (all directions) which is just enough to get past the detent, therefore nullifying it's nasty effects. In fact, I do the same for the P-51 Rudder too, allowing me to adjust my aim more accurately and smoothly.

But with regards to the Stick/Cyclic, I can surely understand why you'd want to remove springs and detents. It would eliminate the need for Trim and Trim Reset functions ... wouldn't it ?

Yes and it feels more like actual helicopter controls (well an R44 at least) also I have removed the center detent and springs in my combat pro pedals and I quite like the feel of them like that.

However I need to build toque pedals and add some form of resistance to the travel, likely another motor bike steering damper (or 2) just to give a nice feel.

No I don't use trim the cyclic stays where I leave it not that you should let go of it. :D

Well between work and stuff I've been busy at Frag labs :P

The ole 3D Pro meck got some shims fitted to remove the slop and play in the gimble Ummm cyclic feels smooth as silk now. :D

Also had to make L brackets with 2 screw mounting for the dampers as the single screw version was prone to loosen but the new brackets work great and so far have not loosened.

And viewed from the front here.

I am however working on a DYI cyclic and here is the prototype, reckon I'd get 10's of thousands of virtual flying hours out of these bearings although I think I can scale down the mechanism to maybe about half the size.

Next torque pedals and a proper V Sim Pit. :pilotfly:

Ah ha now experimenting with UGN3503 's for cyclic and other inputs stay tuned. :D

Okay got an opportunity to make up a test jig to test out using hall effect device and based on what I found with HDD magnet there seems to be 2 magnets set end on end with a polarity change in about the centre. i had also followed some information that Sokol1_br pointed out and I've come up with this prototype that gives me a range of 0.8V to 4.3v and seems quite linear over about 60 deg rotation.

Of course the Hall Effect device doesn't sit there in testing i need to make a bracket to locate it as well as some way to set it's center position and distance.

Okay a bit more work here at the Frag Lab with a final prototype and it seems that I have about 30ish degrees of angular rotation which is good for my prototype cyclic and perhaps torque pedals. The other takeaway here is you can forget millivolt resolution with a range of .8x to 4.2x volts and several mv drift over time/temperature at best so 10bit ADC resolution and dither the LSB (9 bit resolution) will help by keeping steps resolution to roughly 5mv steps and help average/filter the control voltage.

So here is the test bed with adjustable X,Y,Z positioning of the sensor.

:D

Cool stuff, been looking to making my own collective afer the chores of DIY-ing my attic is done and my mancave is finished.

Will follow and ask questions, maybe give adivce, who knows :) ?

Cool stuff, been looking to making my own collective afer the chores of DIY-ing my attic is done and my mancave is finished.

 

Will follow and ask questions, maybe give adivce, who knows :) ?

All the above welcome. :D

Heck I'm just making it up as i go,.. :)