DIY Head Tracker For A Tenner

Njos

N
I tried 3 different cables but it was all the same.

Some news after trying what you suggested. Flashing the Calib sketch worked, flash went all right and all the leds started working as intended: green and red fixed + orange blinking but.... nothing else. The device's still unresponsive and all the readings are flat even if I can click the Reset Factory, Calculate Bias (no effect on anything though) and Toggle Sensor Mode to polling, which didn't have any effect either. Flashing EDTracker 2.20.7 sent things back to where they started: leds off (but the red one), device unresponsive and GUI basically blank. Tried old versions too, same stuff.

btw: win7 x64 home premium here, just in case it may be of some use
 
Last edited:

Polusk

P
Hi Everybody!

Some days ago I was crying since I was not able to run my EdTracker.
I built it from my self (with chineses components, but I tried different ones).
Symptom is:
1) The classical code works
2) the 9150 code doesn't
3) but I can see edtracker on "devices and printers'

Well some news:
I'm still blocked, but now rather than being blocked by the sketch directly integrated in Edtracker, I upload it from Arduino IDE.

I still have the same behavior. It is very difficult for me to debug since I don't have all JTag or other techno in hand.
In any case, I can see that it blocked (ie: no responses from Serial or from Edtracker UI) since it goes to inv_mpu.

Each time it calls a code like:
i2c_write(st.hw->addr, st.reg->pwr_mgmt_1, 1, data)

It freezes and I'm not able to do anything else.
For the ones who knows code from Invensense, they understand that it happends directly when calling mpu_init.
I'm not sure if this init is usefull, but commenting it doesn't help. It will crash at the next i2c_write call.

FYI: I don't know how to set MPU9150 at compile time.
When compiling from source it asking me to defines the MPU i'm using.
So currently I've just #define MPU9150 in the head of inv_mpu.c

As you can see, the issue looks to happen not in edtracker but in a library used by it.
Due to my lack in C/C++/ and the components themself, I'm a bit lost how to continue my study of what it is not working.

Any help would be very appreciated as I don't know which way to follow now.

Thanks!
 

Njos

N
Confirming that modding the sketch as Polusk suggested (and fixing that little son of a short-circuit I made) poured a whole new life into my 9150

IT CAN BE DONE!!!

young_frankenstein1.jpg
 
Last edited:

voon

V
Is the reluctance to look at the 9250, because the 9150 already works perfectly? I wonder if I should throw my 9250 EDtracker into the bin?
 

Antmax

Antmax
9150 works really well. The only minor niggle for me is that if you place it on it's side for any length of time the center drifts and it has to be reset. If you store it upright it stays true for weeks on end. I mentioned this on a reddit post recently and the response was that this was being addressed in an upcoming update.

Otherwise for me, the 9150 I bought runs flawlessly so I'm looking forward to the wireless one they have been developing and is currently in prototype stage.

Being able to clip it onto a wireless headest and run off to answer the door, phone, take care of fighting puppies etc without having to worry about a USB cable will be fantastic :)
 
Last edited:

Arnak

Arnak
Tracker not being seen in ED.

** Update **

The problem is solved.


Hi Folks,

Could I have a little help please?

The tracker is working fine in the calibration software but does not work with ED.

It is a pre-built version.

Should the mouse headlock be set to on or off?

I have tried both settings to no avail.:(

I have the tracker plugged directly into a usb port with a X52, headset and microphone in an extended usb powered hub.

The HOTAS and the headset work fine.

The input selected is custom.

Headlock axis mode is set to direct.

Headlock on / off is set to H and to toggle, tried that but no luck.

** Update **

I have just checked my key bindings file and there is no device listed for the tracker.

Can anyone suggest what should be in that file for the tracker please?

I hope you can give me some advice.

Cmdr Arnak
 
Last edited:

nosscar

nosscar
Arnak said:
** Update **

The problem is solved.


Hi Folks,

Could I have a little help please?

The tracker is working fine in the calibration software but does not work with ED.

It is a pre-built version.

Should the mouse headlock be set to on or off?

I have tried both settings to no avail.:(

I have the tracker plugged directly into a usb port with a X52, headset and microphone in an extended usb powered hub.

The HOTAS and the headset work fine.

The input selected is custom.

Headlock axis mode is set to direct.

Headlock on / off is set to H and to toggle, tried that but no luck.

** Update **

I have just checked my key bindings file and there is no device listed for the tracker.

Can anyone suggest what should be in that file for the tracker please?

I hope you can give me some advice.

Cmdr Arnak
Click to expand...

you won't see the headtracker listed in the bindings yet. If you can see it in your devices, and moving your head makes it move ok there, then go into keybindings in the game, scroll down to head look section, find the axis for look left/right, click to bind and move your head left/right. same with up/down axis.
You should then see the edtracker axis bound to them.
 
Last edited:

AmroBank

A
^^What nosscar said ^
.
.
.
ED Controls:
.
Headlook mode
.
Look up and down axis
Look Left and Right axis
.
Move EDtracker the appropiate way
.
Headlook F4(or what you like)
/Button Mode Toggle
.
.
 

brumster

brumster
Just an FYI, for anyone who has an MPU-9250 kicking about - I've put a development build up on the UI that *should* work with MPU-9250 breakout boards/sensors. Not massively popular but we have had the occassional person ask about it. The breakouts are typically larger, 10-pin boards that mask over the button on our PCB - but you can build an EDTracker with one as they are pin compatible with the 9150. They are basically a better, less noisy version of the MPU-9150 (allegedly). We've had a background task to get the firmware working on them, and I finally got round to sorting it. So....

If you have one, and want to give it a whirl/give some feedback, then if you put your GUI into DEV mode (underneath the About... menu), you will see an EDTracker2_9250 v4.0.3 firmware in the list which you can flash in. Let us know how it works for you. Functionally it's identical to the 9150 - doesn't do anything different. A minor tweak removed the temperature read-out as I couldn't get it to compile in size, but this will probably return given time (it's not critical)!

Cheers,
Dan & Rob
 

Chann3l

Chann3l
Just got my 9150 in the mail yesterday, tried strictly as joystick input and was a little disappointed as movement felt sluggish and obviously treated as a joystick the camera continues to move after you stop, then I followed the instructions for use with opentrack and it was beyond belief. Movement was instantly reflected on screen. I can now easily track enemies and look around the cockpit as if I'm actually there. The level of immersion has increased 10 fold. Thank you guys, this thing is beyond awesome, cheers.
 
Last edited:

LeoKesler

L
brumster said:
Just an FYI, for anyone who has an MPU-9250 kicking about - I've put a development build up on the UI that *should* work with MPU-9250 breakout boards/sensors. Not massively popular but we have had the occassional person ask about it. The breakouts are typically larger, 10-pin boards that mask over the button on our PCB - but you can build an EDTracker with one as they are pin compatible with the 9150. They are basically a better, less noisy version of the MPU-9150 (allegedly). We've had a background task to get the firmware working on them, and I finally got round to sorting it. So....

If you have one, and want to give it a whirl/give some feedback, then if you put your GUI into DEV mode (underneath the About... menu), you will see an EDTracker2_9250 v4.0.3 firmware in the list which you can flash in. Let us know how it works for you. Functionally it's identical to the 9150 - doesn't do anything different. A minor tweak removed the temperature read-out as I couldn't get it to compile in size, but this will probably return given time (it's not critical)!

Cheers,
Dan & Rob
Click to expand...
I would like to ask if the MPU-9250 will be used officially in future revisions ?
(Excuse me my bad english)
 

brumster

brumster
LeoKesler said:
I would like to ask if the MPU-9250 will be used officially in future revisions ?
(Excuse me my bad english)
Click to expand...

Depends what you mean - from a software perspective, yes, if the feedback is OK then we'll keep it up there and merge changes into all streams - 6050, 9150 and 9250.

From a hardware perspective, we'll continue to use the MPU-9150 in the built-up devices we sell; the 9250 is a bigger breakout board so removes the ability to use a button in our pre-done PCB and enclosure. It's marginally more expensive and while on paper it's a better chip, in practise it doesn't offer anything extra for the end user over the 9150. I think if it was to become massively cheaper then fair enough (losing the button isn't the be all and end all, as you can do the re-center with the software hot-key).
 
You must log in or register to reply here.
Back
Top Bottom