DIY Head Tracker For A Tenner

I leave mine sitting on it's side on top of my PC for days on end - naturally when I get it ready for use, it needs a button press, but do you mean you're having to do gyro bias calc - a button press won't sort it, right? I'll bring it to Rob's attention next weekend, we're meeting up for a bit of a codeFest/buildFest on numerous things :)

No mines fine with a button press after lying down. The symptoms I get from lying down are the occasional bounce wobble which the button press fixes.

I made a simple hanger for my headset out of a bookend, holds it upright. With that I never have to press the button. Just pick it up, put it on my head and press the opentrack center button on my joystick.

I don't think there is actually anything wrong with my EDtracker. It works amazingly well :)
 
Weird stuff happens if I place it on it's side on the desk for long periods. I don't think it likes being long periods lying off center by 90 degrees. I started perching it upright against my joystick and now have a simple gangar/stand that holds it upright (I use opentrack and it's center keybind) and the EDTracker 9150 works for weeks without having to press the reset button.

Once I got my ED Tracker running it never gets powered off and this 9150 ran for months without needing even a recenter and it was left on its side for days using the original code. With the newest code it will start to jump about after 24 hours or so and its always left in the same place so dont think its a new magnetic source affecting it.
 
I leave mine sitting on it's side on top of my PC for days on end - naturally when I get it ready for use, it needs a button press, but do you mean you're having to do gyro bias calc - a button press won't sort it, right? I'll bring it to Rob's attention next weekend, we're meeting up for a bit of a codeFest/buildFest on numerous things :)

Yes, it exhibits instability - But does not do it all the time after power up, say 25-50%. You put it on your noggin, reset the view in the GUI, and you get small rapid vibrations with the head slowly turning. I can't easily get access to the button on the device (switch from Hobby components is shorter in length than your version on youtube so does not poke out of the hammond box) so always use the GUI control to reset the view.

When this happens, then I have to perform a bias calc to solve it (with it upgright and static) and all is well.

The EDtracker is powered up and down with the PC, and is often powered down/up with it on its side.

I always pre-check it before using it, its not a big deal.

Its going to be probably be hard to find - today it did not do it, saturday it did.

This small problem aside, the 9150 is perfect! Thanks.
 
Last edited:
Auto-center isn't needed on 9150's; it's "built in" automatically. It uses the mag to compensate for drift, so any code-based "guessing" is not needed ;)

As well as being redundant, I suspect there'd have been no room for it code-wise anyway :)
 
Has anyone tried using this in Star Citizen, i tried but not sure of how to set it with gimbaled weapons, and focus look ect. I know maybe not the place to ask here, but if someone could pm me with some setup details for SC it would help a lot.
I also want to use my joystick and Ed tracker together in SC.
 
Last edited:
Thanks for the response.
In fact, I buy the parts on Hobby Components (9150 on Breadboard)
After assembly :


Did you know if the position of MPU (USB Arduino left and MPU right) has an impact on the yaw tracking? Because its work well but when I turn head to far to the left (usb left conf), the yaw axis turn automatically to 180° to the right. I thought maybe it's the position of the MPU that is not below the Arduino board no?
 
You have the sensitivity too high and it's just wrapping around. If you need to turn your head that much but don't want it wrapping, then turn the sensitivity (scaling) value down for yaw ;) should sort you out
 
Ok you can confirm me the position on the board has no impact on the tracking ? cuz I was planning to desoldering and do like the video XD (MPU below the PCB below the Arduino board).

On Edtracker, the yaw sensitivity is setted on 0.25 but after I configure this on opentrack.
Mayby its the sensitivity in the magnometer tab?

thx for your help!
 
Auto-center isn't needed on 9150's; it's "built in" automatically. It uses the mag to compensate for drift, so any code-based "guessing" is not needed ;)

As well as being redundant, I suspect there'd have been no room for it code-wise anyway :)

Hi

Any idea on current lead times for the mag verion? i broke my non mag ver stupidly and cant get used to playing with out head tracking! lol

Thanks for the awesome work and many enjoyable hours of play you guys have added to the game for me!

Cheers

Pete
 
Can't wait to receive mine, package tracking says it's in Helsinki already :)

While waiting, because EDTracker is only 3DOF, I started thinking if it would be possible to have some way to add left/right movement depending on how much yaw there is. I plan on using the EDTracker with some WW2 flight simulators, and it would help in checking my six if some lateral movement was added towards the end of the yaw curve.
 
Can't wait to receive mine, package tracking says it's in Helsinki already :)

While waiting, because EDTracker is only 3DOF, I started thinking if it would be possible to have some way to add left/right movement depending on how much yaw there is. I plan on using the EDTracker with some WW2 flight simulators, and it would help in checking my six if some lateral movement was added towards the end of the yaw curve.

I tried this in DCS to try and get as realistic an "over the shoulder" view as possible. i.e. you will naturally lean out as you look backwards. I added some X to the yaw in Opentrack and for the first few goes it was looking good, probably a very close approximation to what a real A10 pilot would see. The problems soon started however as tiny errors built up and I ended up well outside of my seat looking forward. I have built a three IR LED+PS3 camera rig to track XYZ and I think that's the way to go. Indeed, Brumster has already said that the hopefully forthcoming EDTPro will have three LED plugs specifically to provide XYZ with a suitable camera.

When you get your EDTracker, have a go yourself in Opentrack, you might have better luck.
 
I'm looking at my b@stardised prototype right now actually, with 3 infra-red LEDs sprouting out of it :) not pretty looking, but got it working with Opentrack. Needs some refinement; the intensity is sufficient for up to about 1 metre but then the LEDs drop out. Need to drop down on a resistor on the board I think, woomph a little more power into them :D
 
I'm looking at my b@stardised prototype right now actually, with 3 infra-red LEDs sprouting out of it :) not pretty looking, but got it working with Opentrack. Needs some refinement; the intensity is sufficient for up to about 1 metre but then the LEDs drop out. Need to drop down on a resistor on the board I think, woomph a little more power into them :D

Are you running 2 instances of opentrack for that? I saw that facetraclnoir has multiple inputs, but I've never seen it on open track.

Got my EDTracker last week and its amazing, so smooth and the magnometer seems to be working flawlessly.

I run mine through opentrack by default now as I like the roll :)
 
Opentrack has two inputs, main (EdTracker) and Auxilliary (Pointtracker1.1)

At least in Opentrack 2.2 stable anyway :)

I too like the roll, very underestimated axis ;) very useful for keeping situational awareness in combat and when making the last turn into the station.
 
Last edited:
Has anyone considered using face tracking for the translation tracking? It's far from accurate enough for rotation, but could it work for this purpose?
 
I tried this in DCS to try and get as realistic an "over the shoulder" view as possible. i.e. you will naturally lean out as you look backwards. I added some X to the yaw in Opentrack and for the first few goes it was looking good, probably a very close approximation to what a real A10 pilot would see. The problems soon started however as tiny errors built up and I ended up well outside of my seat looking forward. I have built a three IR LED+PS3 camera rig to track XYZ and I think that's the way to go. Indeed, Brumster has already said that the hopefully forthcoming EDTPro will have three LED plugs specifically to provide XYZ with a suitable camera.

When you get your EDTracker, have a go yourself in Opentrack, you might have better luck.

I wasn't aware that you could add X to the yaw in Opentrack. I'll definitely try this.

I happen to have a trackclip and and a modded webcam available, and I'll be giving the combination a go if the EDTracker-only option doesn't work
 
I wasn't aware that you could add X to the yaw in Opentrack. I'll definitely try this.

I happen to have a trackclip and and a modded webcam available, and I'll be giving the combination a go if the EDTracker-only option doesn't work

Badly phrased :)
What I meant was that in the "mapping" section of Opentrack, go to options tab and then you can see the output remap section, bottom left. in addition to yaw>yaw, pitch>pitch etc. there are also the translational axes, X,Y,Z. Just select X and remap to Yaw (leaving Yaw mapped to Yaw obviously). In the X mapping tab, just play around with the curves until you get the effect you want. I had it feed in X when my head was at 10 degrees (in game 45degrees) but there is no reason you can't start feeding in the X in tandem with Yaw. As I say, I found the errors in the X translation built up and I ended up sitting on the right side panel (uncomfortable!) YMMV.
Have a go and let us know how you get on. If the EDT pro does come out, I will happily butcher my new Corsair wireless headphones to embed IR LED's to get the extra 3DOF but at the moment it's another wire to have hanging off them :(
 
Back
Top Bottom