RC Groups

RC Groups
    FPV Talk
        Discussion DIY head tracker from cheap heli gyro (seems to work!)

#1 zitron Dec 05, 2008 08:49 PM

DIY head tracker from cheap heli gyro (seems to work!)
 
3 Attachment(s)
It looks like I might have enough money to get a cheap pair of goggles :) , so I've started looking for cheap HT solutions. I've previously tried to build a HT using only an accelerometer, it worked, but was not great: http://www.rcgroups.com/forums/showthread.php?t=886268

Since good quality gyros from Analog Devices and such are $60-100 (which explains way HTs are so expensive), I've decided to pull apart a RC rate gyro ($15?) to get the el cheapo analog components and interface it with my own micro-controller.

Well, after messing with it a bit, I got it working! It does drift a bit and doesn't re-centre perfectly, but I'm quite surprised by how well it works for a cheap ass piezo gyro and analog circuit without thermal compensation...

Bill of material:
~$15 gyro
~$12 micro-controller
------------------------
~$30 HT

Potentially for dual axis I would like to use an accelerometer for pitch since it's much simpler and drift free. So I'm looking at a total cost for a full dual axis HT of $45-50. It probably will drift more than the one's currently sold, but what's the fun in just buying everything :D.

I have not yet test it through my RC Tx and Rx, right now it's only connected directly to a servo, here's video:

Single axis gyro head tracker testing (0 min 26 sec)


Hopefully this will inspire more people to build their own HTs!

Cheers,
-Z-

#2 massimo Dec 06, 2008 02:27 AM

zitron nice nice work .

#3 G-unit Dec 06, 2008 04:03 AM

Thats cool

#4 damian123 Dec 06, 2008 12:17 PM

orrrrrrr a mechanical HT :)

nice work tho!!

#5 scrtsqrl Dec 06, 2008 02:19 PM

I like your style...More brains than money...

#6 zitron Dec 07, 2008 12:13 PM

Quote:

Originally Posted by scrtsqrl
I like your style...More brains than money...

Actually, now that my uni paid up the money they owed to me, it's the other way around :D.


damian,

I've actually been using a mechanical tracker, the elastic band to tx antenna type. It works ok, apart from the fact that it's a bit annoying, and well, I didn't get much satisfaction building it :D !


I'll see how far I can go with the cheap gyro. When I get the whole thing working I'll put together some documentation in case anyone else wants to try building one...

-Z-

#7 Daemon Dec 07, 2008 12:34 PM

Quote:

Originally Posted by zitron
I've actually been using a mechanical tracker, the elastic band to tx antenna type. It works ok, apart from the fact that it's a bit annoying, and well, I didn't get much satisfaction building it :D !

It'd probably be less annoying if you used the badge reel as originally
suggested, since it has a near constant light pull instead of getting
quickly tighter as you turn your head with the elastic band.


Quote:

I'll see how far I can go with the cheap gyro. When I get the whole thing working I'll put together some documentation in case anyone else wants to try building one...
Looks to me like you've got a ways to go yet. You'll need to round up
the individual servo PPM pulses from the gyro, and make a proper
combined pulse to feed to the trainer port on the radio on selected
channels. In the end you'll still have a 2 axis gyro HT with inherent
geometric drift problem requiring periodic re-centering. You may
find yourself going back to the mechanical HT because it just works
and you never have to calibrate or recenter it.

ian

#8 zitron Dec 07, 2008 12:57 PM

Quote:

Originally Posted by Daemon
It'd probably be less annoying if you used the badge reel as originally
suggested, since it has a near constant light pull instead of getting
quickly tighter as you turn your head with the elastic band.

Agreed. If I had a badge reel thing it would be better, even without it, the mechanic tracker worked fine. But my original goal was to build a cheap general purpose HT (some what related to my research at uni), not just to use it for FPV, but also for things like walking around in VR n'stuff, where you don't have a convenient post in front of you to clip to...

For FPV alone, a mechanical tracker is cheaper, simpler, and in some ways more effective. I think there is probably a market opportunity for a nicely packaged mechanical tracker.


Quote:

Originally Posted by Daemon
Looks to me like you've got a ways to go yet. You'll need to round up
the individual servo PPM pulses from the gyro, and make a proper
combined pulse to feed to the trainer port on the radio on selected
channels. In the end you'll still have a 2 axis gyro HT with inherent
geometric drift problem requiring periodic re-centering. You may
find yourself going back to the mechanical HT because it just works
and you never have to calibrate or recenter it.

Well integrating it in to the Tx will be simple. I have already modified my cheap RC radio, removed the original digital board, and replaced it with my own (actually I did it because I destroyed the old board by accidentally soldering 12V to 5V :D , but I digress...). So all I need to do is to link my HT with my Tx board with a simple serial connection.

The drift problem is more of a concern. I'm planning to use an 3 axis accelerometer for pitch and roll, so there will be no drift in pitch, and I can attempt to correct the gyro drift using a kalman filter in IMU style, as long as your head is not exactly straight up all the time. This is the part that will take the most work.

-Z-

#9 damian123 Dec 07, 2008 02:03 PM

Quote:

Originally Posted by zitron

damian,

I've actually been using a mechanical tracker, the elastic band to tx antenna type. It works ok, apart from the fact that it's a bit annoying, and well, I didn't get much satisfaction building it :D !



-Z-

hehehe :) it is a bit a annoying I agree :) but hey its cheap and it works!

good luck on your HT!

#10 zitron Dec 21, 2008 10:51 AM

1 Attachment(s)
Update:

I've pretty much got the thing working. Still need to add some drift correction stuff, but otherwise it works great!

It's too cold to fly, so I've been testing it in FSX. Currently there is quite a bit of drift in the panning, I've not yet added any button to re-zero the angle, I've simply programmed it to reset the angle when you lower your head more than 60 dgrs. You can see me do that in the video. Because the roll and pitch axes are accelerometer based, there are no drift in those. Here's a video of me doing the helicopter oil rig inspection mission:

Testing DIY 3DOF head tracker in FSX (8 min 39 sec)


The first couple of times I tried flying with goggles, I was looking around a lot, but I've since stopped doing that because it really gives me motion sickness! :eek:

Cheers,
-Z-

#11 tychoc Dec 21, 2008 11:19 AM

Great job! That looks really good.

-tychoc

#12 mmormota Dec 21, 2008 11:24 AM

What about using a magnetic sensor in order to avoid drift? Even the price is better then gyro's. (Honeywell SSEC - HMC1002-RC etc.)

#13 zitron Dec 23, 2008 05:42 PM

Mmormota, do you know anywhere that sells the HMC1002 magnetometers (or similar) on easily usable boards? It looks interesting, but it needs external amplification circuits to get it working with a micro-controller. Since I only know how to do strip boards, even a simple OpAmp circuit will be quite big, I'd really like my HT to be small...

I've actually been looking at this:
http://www.sparkfun.com/commerce/pro...roducts_id=761

It's a bit more than a cheap RC gyro, but is the cheapest I could find.

Cheers,
-Z-

#14 mmormota Dec 23, 2008 05:55 PM

Good find, I don't know any better for an evaluation kit.

I proposed the magnetic sensor because the drift is inherent feature of the piezo gyro's and practically impossible to avoid it. The smm gyro's are better but more expensive, and hard to find smm sensor's in small quantities.

#15 damian123 Dec 23, 2008 07:19 PM

zitron, I liked that mission as I have done it myself :) I wish I would have a ht for the red bull race mission..
RED BULL RACE - FSX (3 min 25 sec)
:D I got my time down by alot since that video ;)


All times are GMT -5. The time now is 08:58 AM.