• There has been a recent cluster of spammers accessing BARFer accounts and posting spam. To safeguard your account, please consider changing your password. It would be even better to take the additional step of enabling 2 Factor Authentication (2FA) on your BARF account. Read more here.

Building a Gyro Cam Mount

Jeff gave me a quizzical look when I mentioned Dorna used GPS in their units.

And I have no fucking clue what I'm looking at in that photo, apart from the MicroSD card (and only cause I can read)! :laughing
 
thats cuz i made it :laughing. the square thing is the GPS antenna

its a mildly useful GPS datalogger thatll also record another analog input or two... say, like suspension position :D

all this talk about cool gadgets and projects made me want to work on it again. :nerd
 
Oh, well in that case, let me look at it again.

































































Nope, still no fucking clue what's what! :laughing
 
I'm still not a believer that their Camera runs off the GPS. I believe the GPS is there because Dorna wanted it there for data logging and track position purposes. I believe their camera runs solely off the gyro.
 
Did you read it? Of course the gyroscopic camera doesn't operate just off of the GPS; it's the GPS, accelerometers, gyros, etc. that make it all happen for them. Pretty sure their system utilizes these components to deliver relatively seamless gyroscopic footage given the environment the camera has to operate in.

But this is just water-cooler digression from your thread.
 
Subscribed. If only I was a little smarter I would totally attempt to help you! My .02 is the gyroscope isn't necessary, but it sounds like it's already implemented in a similar application (for the R/C helicopters) making it an inexpensive off the shelf component.

I think it would be doable with just a 3-axis accelerometer (about $25 on a breakout board from Spark Fun) and some clever coding, as someone already mentioned. This would be needed to filter out the high frequency noise that would make the servo jittery. I believe a Kalman filter algorithm would work.
 
Why would you need a 3-axis accelerometer? Remember, its only moving on a single plane. I'm not sure how expensive accelerometers are, but the gyro is the most expensive component in my setup. You can get cheap $20 gyros, but they just won't cut the mustard. That might be the same about a $20 accelerometer. Since I know nothing about programing, I went the route of using a gyro because it could be done with only hardware. Still waiting for my new servo so I can try it out.
 
llenta said:
Why would you need a 3-axis accelerometer? Remember, its only moving on a single plane. I'm not sure how expensive accelerometers are, but the gyro is the most expensive component in my setup. You can get cheap $20 gyros, but they just won't cut the mustard. That might be the same about a $20 accelerometer. Since I know nothing about programing, I went the route of using a gyro because it could be done with only hardware. Still waiting for my new servo so I can try it out.
Click to expand...

a single plane would mean a 2-axis accelerometer is necessary... but no one makes one because they are impossible to use. the chances of u mounting the accelerometer perfectly in that single plane is impossible, so u use all 3-axis to get info for the 2-axis that u care about. vectors n all...

the funny thing is there is already a small microcontroller in your gyro, enabling it to hold a given heading and manipulate your digital servo. so, if i was programming the microcontroller myself, i could easily use a 20$ (or cheaper) gyro and/or a 3-axis accelerometer. im still of the opinion that a gyro AND at least a 1-axis accelerometer are needed to do this perfectly. gyros tend to get lost w/out a reference point. but i guess we'll find out.
 
stangmx13 said:
a single plane would mean a 2-axis accelerometer is necessary... but no one makes one because they are impossible to use. the chances of u mounting the accelerometer perfectly in that single plane is impossible, so u use all 3-axis to get info for the 2-axis that u care about. vectors n all...
Click to expand...

That makes complete sense, and I totally agree.

stangmx13 said:
the funny thing is there is already a small microcontroller in your gyro, enabling it to hold a given heading and manipulate your digital servo. so, if i was programming the microcontroller myself, i could easily use a 20$ (or cheaper) gyro and/or a 3-axis accelerometer. im still of the opinion that a gyro AND at least a 1-axis accelerometer are needed to do this perfectly. gyros tend to get lost w/out a reference point. but i guess we'll find out.
Click to expand...

Well, I would say 99% of the world (including me) does not know how to re-program the gyro to do only and exactly what they want. So we (me) are left with the option of getting more expensive units and using a control module (like the one I have) to set and hold a specific heading. The control module can be used to change the heading on the fly, or just be used to hold the gyro in a neutral position. From there, the gryo does the rest. But I'm sure if you know how to program, then you could probably do a better job.
 
llenta said:
That makes complete sense, and I totally agree.



Well, I would say 99% of the world (including me) does not know how to re-program the gyro to do only and exactly what they want. So we (me) are left with the option of getting more expensive units and using a control module (like the one I have) to set and hold a specific heading. The control module can be used to change the heading on the fly, or just be used to hold the gyro in a neutral position. From there, the gryo does the rest. But I'm sure if you know how to program, then you could probably do a better job.
Click to expand...

either way you choose to make this, the end result is badass!

PS - I don't know shit when it comes to coding except for the basics. I use other people's code and hack at it till it does what I want. I am sure if you picked up an arduino you could do the same. It's really easy to do if you want to try it out and there are tutorials online that can teach you as you go.
 
Last edited:
stangmx13 said:
having a gyro AND accelerometers make their devices quite a bit more robust.

GPS!!!!
Click to expand...

Dood that is awesome! I love nerding out on your board! Did you code the gyro to work with a servo and if so what servo? Would you mind sharing your code?
 
I agree, a Kalman filter is what is needed to make the camera most stable. There is a reason it is used so frequently in satellites for remote sensing and attitude determination.
 
Entoptic said:
Dood that is awesome! I love nerding out on your board! Did you code the gyro to work with a servo and if so what servo? Would you mind sharing your code?
Click to expand...

the board i posted the pic of is for a GPS datalogger, not for anything involving a gyro or servo. though, i can help you w/ Arduino code if u want.
 
I'm waiting for Contour to catch onto the desire for this technology and add it to their camera line.
 
I received my new servo in the mail on Saturday afternoon. It was too late in the day for me to rig up everything as I was working all day and I had other things I needed to take care of. So I decided to push it off and get things ready for next weekend. Here are some pictures of the old and the new side by side.

img1041i.jpg


img1039lj.jpg


I was able to get to the hardware store and pick up a bolt that I needed so it looks like I have everything that I will need to give this another go this weekend.
 
Right on!! Lets test that bad boy out! Sat or Sun?
 
let ride Saturday as I'll probably be out of town on Sunday.
 
So I had some time last night to put together a new mount from the pieces that I ordered. This time the camera will be mounted much closer to the servo, so there should be a lot less stress on it. Oh, and I should add that this new servo is 180 degree, so it can rotate a full 90 degrees each direction (hopefully it never will :teeth). I took a drill and drilled through the center of the RAM ball mount, then threaded the hole with a die. I then took a bolt that has the same pitch as your standard tripod screw and put it through the hole in the RAM ball mount. With this setup you can put any camera that has a universal mount on here. Its much more stable then the old setup and the bottom part of the screw can act as a support.

img1045rt.jpg


If this works better I will coat the bottom of the bolt with rubber or some non-abrasive material. Last thing I want is for it to chew threw the plastics :laughing. For now, I just took some tape and foam and put it under the bolt. It should hold up for at least one ride. I'll see how it does and make adjustments as needed.
 
So its the middle of January and look at the weather, beautiful day to go riding. As if riding in general isn't fun, riding on a new set of rubber that I put on 2 days ago is even better :teeth. I put on a new set of Pirelli Rosso 2 and went to scrub them in, oh and I took my new gyro set-up along with :cool. This time the gyro is mounted right on the servo, and the camera is mounted almost directly to the servo as well. Less parts mean the whole setup is a lot more stable, and it definitely was.

img1049h.jpg


img1051y.jpg


It turns out that I was more excited to get the new rubber on the bike and get the new servo connected and running that I forgot to completely charge my battery for my camera :rofl Oh well, got most of it. I went up 9-Bear Creek-Summit-then Old San Jose up and back, then 17 back home. The camera crapped out when I got up to Old San Jose when I stopped at the store.

It looks like I nailed it with this setup. I think I need to dial back the gyro just a bit, but it recorded the whole time and remained stable. I didn't actually push it that hard considering I was riding by myself and these are brand new tires (and I ride like a girl anyway). But I think this package will do the job well.

Here is a clip from the back side of 9
[youtube]v_0YpL3GSP0[/youtube]

The next step I think is to put this together in a portable package. Then find a volunteer that really knows how to :ride

mmm, can't wait for my next trackday.
 
You must log in or register to reply here.
Back
Top