Jump to content
Vrflyer

camera movement follow headset movement

Recommended Posts

Kudos to Kilrah and VRFLYER. Thanks for sharing this development. Kilrah, make sure to post to this forum a picture of your castle when you buy it.

Cliff

Share this post


Link to post
Share on other sites

Hi Kilrah -

Do you currently offer any discounts on the Gyroncontrol? $400 is just a bit too steep for me, especially after factoring in other costs involved into building home VR flying project...

Thanks!

Share this post


Link to post
Share on other sites

Thanks for your compliments cliffo.

As for a discount, sorry but it's really not the right moment. I'm making the Gyrocontrol myself, and already wonder how I'll be able to satisfy the interest that showed up these days while having another full-time job...

Share this post


Link to post
Share on other sites

Hi all,

I am much interested in this new type of flying, seen for the first time 2 weeks ago. Then i decided to try and and ordered everything on ebay. But I absolutely wanted a headtracking system, and I love the one from kilrah but cost is too much for me.

Si I decided to try to make one based on the one from kilrah. I can tell you Kilrah, you had greats ideas, and I used yours ideas to make mine. It now works fine.

You can see it :

http://forum.modelisme.com/t80457-headtrac...our-du-fpv.html

but it's in french. I haven't tried it in a real plane, we'll see.

Share this post


Link to post
Share on other sites

That's good work. Though, I guess you'll be running into some issues when you get to test it in condition.

I had made my first tries with PG-03 gyros too, but instead of sending a dummy servo signal and reading the output I tapped an ADC at the point where the anplified analog signal could be found.

But, the Murata sensor used in that gyro is quite nonlinear, which makes it pretty unusable. What I see in your video matches pretty much what I've found too. If you look closely at the servo's positions throughout the demo, you'll see that once you've made one back and forth movement on the tilt axis (the PG), the tilt servo doesn't return at all at its center position. That's because the sensor doesn't behave the same in one way than in the other. But, for your integration to come back to the same position you need the exact same info for one side than for he other.

You won't be able to get good results with the PG unless you do some floating-point scaling and certainly calibration of each single sensor you'd want to use.

I had made pretty complex algorithms to try and cancel the sensor's offset variation over time which were working pretty well, but that issue made me throw my PG's in the bin. Not worth it.

That problem is less noticeable on the pan axis (Futaba gyro), but you made less movement with it in the video. From my point of view if you want to do that with heli gyros "as is" you'll have to go at least with pretty good heading-lock models (with heading-lock deactivated of course, but as an integration is also used in that feature that would ensure that the gyro is of sufficient quality).

You'd end spending €150 on each of the 2 gyros...

The problem is that this application is VERY demanding in precision and stability compared to the other common uses. A heading-lock heli gyro uses the same principle, but if it drifts 2°/sec the pilot will simply trim a bit to compensate it. No big deal. Here, your senses will notice any unwanted response very easily, and it will be annoying.

Share this post


Link to post
Share on other sites

Kilrah, have you considered using a 3-axis gyro along with a 3-axis accelerometer to compensate for drift? It's a heck of alot of computation to do between the two, so you'd need a beefy uC, but it should be able to get rid of errors like that. And cost alot more.

Share this post


Link to post
Share on other sites
Kilrah, have you considered using a 3-axis gyro along with a 3-axis accelerometer to compensate for drift? It's a heck of alot of computation to do between the two, so you'd need a beefy uC, but it should be able to get rid of errors like that. And cost alot more.

Eickst,

I think Kilrah is finished. The programming work, integration with the radio and the gyro he is using he has developed a good working model that needs little if any further development.

His problem is he needs to sell it to XYZ, Co., stay on as consultant for $xxx,xxx per yr/month and let the new company put the product out in force. His solution will expand the RC industry dramatically and touch untold others.

Am I right Kilrah?

Cliff

Share this post


Link to post
Share on other sites

Eickst,

I think Kilrah is finished. The programming work, integration with the radio and the gyro he is using he has developed a good working model that needs little if any further development.

His problem is he needs to sell it to XYZ, Co., stay on as consultant for $xxx,xxx per yr/month and let the new company put the product out in force. His solution will expand the RC industry dramatically and touch untold others.

Am I right Kilrah?

Cliff

Oh, I am sure he isn't finished. Guys like Kilrah are never finished. There is always one last tweak, one last little enhancement to make, one more change to the firmware....

So he'll have plenty of product revisions (as most products do) and each one will be better than the last. I see kilrah putting in sensors to determine not only the attitude of the wearer's head, but the direction his eyes are pointing, too, similar to the Apache and Cobra gunship aiming system for the cannon.

Share this post


Link to post
Share on other sites

Actually it's a bit of both. If I follwed cliffo's way I'd maybe have time to improve the existing setup, or even start again from scratch using other solutions.

But for now, I'm already struggling to cope with the sudden interest that is obviously not expected to go on the descending slope. The current unit is a final version. Of course every engineer wants to tackle things until he's perfectly satisfied. But that approach unfortunately doesn't fit our world. There's a moment when you just have to decide it's ready even if you know you could do better. The thing is, you can always do better. You have to be realistic. At some point you have to consider other things like costs, time and opportunity. You want to change something? Add a sensor. That will need a hardware redesign. Which will need a software redesign. It's the best way to end with nothing functional, or... 3 years later than expected. I know someone who has followed that path...

My first goal was to fly FPV. Considering the "realistic" criteria I can say I'm now satisfied with what I have. I have a good quality wireless camera, reliable enough for me to enjoy flying with it without bieng limited by its performance. I have a head tracker that works well enough for me to feel like I'm in my plane. Forced by some friends who found it good too I made a few more units. We now eventually sell stuff all around the world... certainly not what we'd been expecting at any point. But it seems that I'm hard enough to please so that others find it good too ;)

That never excludes other models though, improvements always have their place. And for the moment, I don't see the advantage of throwing a full 6DOF IMU and the related costs and complexity for "consumer" headtracking. Now if the Army hires me, I might think about it :)

The current stuff works. I find it more interesting to expand the system even more rather than to stick to one aspect only.

As a conclusion, cliffo nailed it. Our current organization of 2 students doing that mainly for interest next to a full-time occupation can't stay anymore, some serious reorganization will have to take place.

I don't like the consultant idea though. I like what I do, I manufacture all my stuff myself, and enjoy it. But I'd indeed want to still have some spare time to use what I originally did for myself ;)

Edited by Kilrah

Share this post


Link to post
Share on other sites

yeah, kilrah, it seems like you are doing what makes you happy, which is good. It definitely gets expensive, and I fall into that category where I always try and improve stuff, but I never plan on taking anything to market, its all for personal use so I always just keep going until I get it just where I want it.

Share this post


Link to post
Share on other sites

Kilrah,

Yeah, I can see four orders for trackers putting you behind a week. ha,ha... Perhaps soon that will be how many are sold every 10 seconds.

You have already found solutions for the major nigs and nags. This is like the first car. It would get you down the street and around the corner without breaking down. So it accomplished the major goals, but as more interest develops you will not be able to manufacture enough and can maybe work on improving what is there, without it still being fun.

Congratulations for the major breakthrough.

Cliff

Share this post


Link to post
Share on other sites

Hi there.

Kilrah, is your head tracker's drift problem solved ? Almost all heli headlock gyro have the same (little) problem, and it' s related to temperature. I have to let mine get temp stabilised when "on" for few min. and then slightly adjust the rudder trim so the drift will be minimal.

IF the drift is caused by the bias circuit, could't it be solved by the use of precision resistors ?

Share this post


Link to post
Share on other sites

People (who probably haven't even used the Gyrocontrol) seem to be blowing this drift "issue" way out of proportion. I have a set, and believe me, it's not a problem at all. I initialize my goggles while they are on a bench at the field, and then look at the camera to see how it's behaving while the goggles are stationary. I only have seen the camera drift a few times, and that may have been because I nudged the cables. In use, they may require a re-center as the electronics gets up to temp, but it's never distracting at all.

To be honest, I find myself recentering the device during flight because of my slight body movements more than anything. It'll be interesting to see how this fall's lower temperatures have an effect on the gear, but I don't expect trouble.

Share this post


Link to post
Share on other sites

Has anyone tried this head tracker gear on a non flying situation with a geared servo pan/tilt head. (Like those now sold by servocity?) I'm filming football for our highschool and have a mounted my Canon GL2 on a pole to get some more heigth. On the pole I have a bescor MP101 pan/tilt drive and remote the camera controls as well. The challenge is that it takes three hands to control everything. The bescor is great for the money but the direction controls are simple buttons (under a joystick) and it has a seperate speed control (which is too fast even at the slowest setting) when zoomed in close. I'm messing around modifying the bescor control so I get proportional speed with a stick deflection (somewhat of a brute force approach) but it occurred to me that perhaps the head tracker setup with be a good approach. One of the problems is that heavily geared servos (You need a geared servo setup with heavy camcorders) can't keep up with the desired movement. (I tried Mr. RC's brilliant pancam and the 5:1 geared servo keeps going for seconds after you tell it to stop. However it might be able to keep up with a 2:1 gearing if that gearing can handle the camera weight.) Anyway I guess what my babbling boils down to is the question as to whether this tracking system has been used with geared pan/tilts? I think it would be really great if I could look at the action by turning my head and then with one hand control record/pause and zoom.

-steve-

ps Also if anyone knows of any one that has done surgery on the bescor MP 101 to put true motor control on it I would be very interested.

Share this post


Link to post
Share on other sites

Hi,

I'm new here and I decided to do a system like this. I have already bought two gyro ( IDX300 breakou board) and I have started the board for the PIC. I'm just waiting the PIC to test the software (I already started the software too).

Today I bought 3D glasses , so I should have everything at the same time. if this project works I will put the drawing into my web site to share this.

( the final project is to do a UAV). We have alreay made some testswith the UAV but a friend of mine, ask more capability on it !

if you read french you can see the blog for that project :

http://serge.laforest.free.fr

then go to the "drone" page.

I have also made some videos ( with OSD) the videos are in:

http://serge.laforest.free.fr/videos

Regards

Serge

Share this post


Link to post
Share on other sites

Serge, you have some pretty interesting videos!.

Can you tell us more about the OSD and equipment you use to display altitude, heading, speed and steering cues?

Also, I couldn't decipher the information on the bottom line. What is being shown there?

Share this post


Link to post
Share on other sites

I made the OSD based on the STV5730 ( obsolete, I bought the last one in France but I know that some people found it in Russia or china). This OSD is directly connected to my "autopilot" board. a PIC on this board send the commands to the STV5730 ( the PIC has all the low level function in my board for security reason). The upper processeur is a gumstix ( see the picture on my web page, section drone). This autopilot board has Gumstix inside for high level stuf like navigation and stabilization ( the discussions between the PIC and the gumstix are on I2C bus). All the low level function are into the PIC software ( altimeter, read the servo position, output the desired servo position, check if we are in automode or manual remote control and OSD.....). I got all the information from my board. this board has 3 magnetometer, 3 gyros, 3 accelerator, pressure sensor, tachymeter ( for helico), analyze the frame from the receiver, generate the frame for the receiver in autopilot mode. there is also a modem inside it to send the datas in real time to the audio channel of the video transmitter, so I get the datas on my laptop or PDA.

pressure sensors used is MS5534

tachymeter : only by software on my board

Heading : GPS ( LEA-4P) and 3D honeywell sensor HMC1021 and HMC1022

Gyro ENC-03

Accelerator : ADXL202EB

on the link below you will find the organisation of the system

http://serge.laforest.free.fr/drone/drone.jpg

The information on the bottom line is the number of errors by second received fro m the receiver ( PPM frame error) ( receiver was not connected to the system on the video I made, explain why it is always 0/sec!). It is just for a test

Sorry for my english !! :)

Now a friend of mine ( the pilot) ask me to have the camera mounted on 2 servos to follow his head. That's the reason why I'm going to do this interface.

Serge

Edited by sergio123

Share this post


Link to post
Share on other sites

Really sweet project Sergio. Ya gotta post some vids here for all to see. Sounds like something I'm going to eventually build...I'm doing it in slow(very :( ) steps as time and money and knowledge permits. I sure learn alot from this site and rcgroups and all the projects people post. I love this stuff!

Share this post


Link to post
Share on other sites

Thanks for your comments

But I 'm like you : slow slow ... for the same reason as you money, money !

( but also because I travel a lot for my job).

Tomorrow we should make the first immersion fly ( remote controled) with 4 waypoints. The pilot doesn't know where are these waypoint. When he will arrive to the first waypoint ( 40m) the autopilot will tell the pilot the next one and so on.

The arrows you can see on the top of the video are the turn the pilote should made to arrive on the newt waypoint. ( 1 arrow by 30 degre)

I f the weather will be OK I will post the video on the same directory

( sorry if I'm out of the discussion ) ;)

Share this post


Link to post
Share on other sites

Very impressive work, I would like to be able to make my own OSD, I would like an OSD that can be plug in the Seagull flight recorder, it already have an output that goes in a 900 mhz tx to downlink data. I would like to to use this signal and send it to OSD directly. Perhaps the cie Seagull will eventually make one.

Edited by Vrflyer

Share this post


Link to post
Share on other sites

vrflyer,

if your seagul recorder has an RS232 bus and if on this serial bus you have the datas you want to display on OSD, it should be easy to do a OSD fot that. you just need a PIC ( for example) connected to the STV5730 and to this serial bus.

The main problem will be to find a STV5730(A).

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...