Camera tilt and pan controlled by head tilt - done

48 posts in this topic

Posted (edited)

yes it's cool, do you have tried it with a video headset ? Does it have a drift when not moving ?

Do you have seen this video:

ftp://kilrah.dynalias.net/Temp/gyrotest.wmv

Edited by Vrflyer

Share this post


Link to post
Share on other sites

Posted (edited)

Looks pretty good indeed, nice to see another working unit! :)

What's the use of the LCD?

Accelerometers won't drift, but they can't measure yaw. If you check the video above, you'll see that you need to "roll" your head to get a yaw movement. They are sensitive to dynamic acceleration too, so if you move you'll get undesired movements.

Edited by Kilrah

Share this post


Link to post
Share on other sites

Posted (edited)

What a great little project! that's exactly what I've been wanting to do, except that I have Optic 6 and I wanted to use AVR instead of PIC. But your schematics and source code would be of great help in my work (I wouldn't come up with those 10 uH chokes myself).

That's of course clear that accelerometers can not measure yaw, but what you get is a reasonable alternative to expensive commerical products. Kind of a poor-man's head tracker, which does what is supposed to do, although somewhat differently, but it costs way less than commerical products and is drift free. I am sure that it would take only a couple of minutes to get used to the idea that you have to tilt your head sidewise to make the camera pan. And the centering is very natural in this setup.

I guess that with some low-pass filtering or with a sensitivity (expo) curve added it would be possible to eliminate those jerky movements.

Another possibility is to replace accelerometer's one axis with a gyro module (like those by Analog Devices) to get yaw reading. At least this is what I'm thinking now, theoretically.

Edited by mieczotronix

Share this post


Link to post
Share on other sites

Posted

I imagine a guy with headset moving the head trying to position camera, he will look dystrophic :)

Share this post


Link to post
Share on other sites

Posted

I imagine a guy with headset moving the head trying to position camera, he will look dystrophic :)

C'mon. Don't tell me that a guy wearing the VFP gear on its head looks otherwise absolutely normal ;)

Share this post


Link to post
Share on other sites

Posted

Yep, I definitely already got some strange looks from people when wearing mine :lol:

What the heck are these guys doing anyway? ;)

PICT0187.jpg

Share this post


Link to post
Share on other sites

Posted

I've heard of "flying under the hood" but never flying under the cardigin :lol:

Share this post


Link to post
Share on other sites

Posted

Kilrah, you look someone who want to proove even blind people can fly r/c planes :)

Share this post


Link to post
Share on other sites

Posted

Haha! Looks like you need to add a helmet or visor to your video glasses. I constructed one that works pretty good with my I-Glasses. Here's two pictures of it. As you can see, it flips open so you can see the surrounding area if need be. Then flip it closed and almost no light gets in at all.

Works great, although I think I like the looks of VRFlyer's visor better. :)

post-8-1152037904_thumb.jpg

Share this post


Link to post
Share on other sites

Posted

2nd pic...

post-8-1152038026_thumb.jpg

Share this post


Link to post
Share on other sites

Posted

Haha :lol:

Well I actually can fly without adding anything. It's just that it's even better like this. And I'm usually carrying all the gear in a backpack, with R/C TX, laptop, goggles, video RX,... so don't have much more space left ;)

Share this post


Link to post
Share on other sites

Posted

Your glasses must be brighter than mine then. I can't see the image at all in direct sunlight.

Share this post


Link to post
Share on other sites

Posted

They have CCFL backlights, which are indeed pretty powerful. But, you need to turn on the glasses and let them heat up for about one minute before taking off to get full brightness. There's a "power save" feature that turns the backlight off automatically when you're not wearing the goggles (IR sensor on the headset!), that I'm obviously not using for that reason... even if the battery drain is impressive :blink:

Share this post


Link to post
Share on other sites

Posted

Twinturbostang, it's good what you've done. You can easily replace the wood panel with acrilyc sheet. Dark tint acrylic is rare, I found in little store where they made outside sign for store, a lot of retail acrylic from different color, they sold me everything I want for a ridiculous price ;)

Share this post


Link to post
Share on other sites

Posted

seems like no one is interrested in what Prikupets did (well apart from me)

and this topic now seems to drift towards Darth-Vader helmet replica construction

I've analyzed the source code of Pirkupets' (it's pretty well documented and clear even if I know nothing about PIC architecture).

Kilrah asked what the LCD was for in the project. From the source code it seems that it was for debugging, it displayed measured and processed values.

The accelerator processing does some kind of dead band and low-pass filtering (moving average) on measured accelerometer readings. The code also extends servo movements. It does not implement any sensitivity curve, or extreme measurements rejection which I think might improve the device a bit.

Anyway form me this little project if of great value and very inspiring.

Share this post


Link to post
Share on other sites

Posted

I just had a quick look at the code - it seems to be synchronising itself on the TX's PPM output, and to change 3 of the channels before giving it to the PPM in again. That's what I wanted to do as well initially, but I still had to offer full frame generation for the few TX's that only have one pin for both in and out (e.g. Multiplex). Unfortunately I haven't been able to do it because the user interface/configuration routine took so much code space I ran out of memory to add that feature... :(

Share this post


Link to post
Share on other sites

Posted

I'm very interested in the project, I would like to see that one year ago, but now I got the gyro, I will not need it, but I wll read all the post.

Share this post


Link to post
Share on other sites

Posted

C'mon. Don't tell me that a guy wearing the VFP gear on its head looks otherwise absolutely normalĀ  ;)

It does if your name is Geordi LaForge :lol:

I have a simple solution for using my iTheater glasses in direct sunlight, the Duckbill Cap mod:

DuckbillCap.jpg

Share this post


Link to post
Share on other sites

Posted

The project is interesting. I agree that the tilting head movement for yaw control would suffice. But, it would be grand if it used a more natural movement.

BTW, this project is similar in function to the HeadTrack-R design attempt from a few years ago. Pitch and Roll used accelerometers, and yaw sensing was from a digital compass. The compass module turned out to be a bad choice. I did experience that accelerometers are perfect for pitch and roll sensing on your head (just don't jump around while you do that :) ).

In case it offers any inspiration, some info on the old HeadTrack-R project, and some photos, are on the old rc-cam forum resource site: http://groups.msn.com/rccam/general.msnw?a...ID_Message=6106

Share this post


Link to post
Share on other sites

Posted (edited)

I just had a quick look at the code - it seems to be synchronising itself on the TX's PPM output, and to change 3 of the channels before giving it to the PPM in again.

Exactly. This little project is actually three projects in one. It does I ever wanted to do with microcontrollers and servos: extends servo movements, processes the PPM stream and does the headtracking!

I also wonder whether using this technique I could turn my Optic 6 (5 proportional + 1 discrete channels) into 8 discrete channel machine. Just by replacing servo cycles I don't need with what's useful and by extending the 6-channel PPM train with two another servo pulses. I hope this will work since I don't believe the PPM signal that goes to the trainer port is subject to any further processing in the transmitter before it gets to the FM module.

And thanks Mr RC-CAM for the links! I'm quite new to this forum, so it will be a lot of reading for me

Edited by mieczotronix

Share this post


Link to post
Share on other sites

Posted

Unfortunately I haven't been able to do it because the user interface/configuration routine took so much code space I ran out of memory to add that feature... :(

maybe you should scale down and use a DIP-switch/jumper config interface. Why bother with menus and UI for stuff that's only done once.

Share this post


Link to post
Share on other sites

Posted

That's the point, it's not only done once. Depending on the R/C TX, like with mine, I have to redo the configuration each time I change model...

For the settings I've implemented if everything was to be done with hardware switches and stuff I'd need like 12 dips and 2 pots... which would mean 14 extra input pins on the controller plus the actual controls, thus a ~3x bigger unit...

Compromises, as always ;)

Share this post


Link to post
Share on other sites

Posted

Hmmm... I wonder what I'd have if I hacked into Prikupits' schematic and dropped in a dual axis, analog output gyro?

Share this post


Link to post
Share on other sites

Posted (edited)

Something strange :lol:

The accelerometer gives out an information that can be directly treated as a position. With a gyro the first thing you'd have to do would be to integrate the rotational speed info it gives you to get that position (where the drift comes from :) )

Then, you would add a good differential amp between its reference terminals and its outputs, and hope your sensor is really symmetric ;)

Mmh, a post appeared on top of the page..

The Optic 6 hack might work. You should first ensure that there is no further processing as you say, for example by using a transmitter that has a 8-channel output and connecting it to the trainer port. Some TX's are pretty strict regarding the input signal and will automatically switch back to the master if looks too different from what's expected. The Futabas do accept a bit more throw than normal, so I've been able to boost the travel to about 150%. I still use a module in the aircraft to get the maximum mechanical servo throw though. They all transmit 8 channels.

Edited by Kilrah

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now