This is the thread where I document my attempts for this round of T3. Gary mentioned that documentation would be important, so I decided to make this thread to document my entire process. My goal is that anyone who reads through this, once it is finished, will be able to duplicate my process exactly. I will disclose everything I do, no "secrets".
So here's my first bit of progress:
I've been having trouble with my Deans plugs, so I decided to make an XT60 adapter for my octo. The problem with the deans was that, in the past, sometimes one or two of the batteries would disconnect, leaving the others to power the copter. This resulted in one close call and one crash, so I'm replacing the deans and switching to XT60. And yes, I'm aware of the current limitation on XT60's. I am well within it.
I have some more progress to share with you! I'm using an Olympus E-PL2 camera. This is a fairly expensive camera, for my lab anyway. I decided I was not comfortable flying it unless I have some reasonable assurance that it will survive a crash. I'm also not super interested in using a gimble: I don't use them normally and I'd like to keep things simple. Enter: the camera box. I actually used the Olympus' own box as a case. Inside is packing foam that 3DR sent with my Udrones Hexas (whenever I say my, I'm referring to the lab I work in. I'm the main copter guy here.) The packing foam was easy to cut cleanly with a new exacto blade. So here it is:
Good for you Stephen, marks already ;-)
I finished my first test flight carrying the Olympus camera.
The camera box now has mounting gear on it. More on that in a bit. I also brought out a camera calibration card, but unfortunately I couldn't find a way to set the Olympus to infinite focus. If I'd calibrated the first shot with the card, it would have held its focus too close. If I set the camera to autofocus on each shot, it slows down the picture taking speed considerably. So I just took the first shot of a far away building and resolved to look into the focus problem later.
Here's my high tech solution for keeping the camera triggered. The rubber band was a little finicky, I think I'll try velcro next time. The camera can continuously shoot 2 pictures per second. Which is great, because my method uses as many photos as possible.
Here's how the camera box mounts to the copter. I had a spare Mikrokopter center plate, so I bolted it to the camera box and added standoffs. The plate is anchored by some broken 3DR landing gear plates which I cut to size and placed on the inside of the box to distribute the load onto the cardboard. I also have a velcro strap to keep the box closed.
Here it is mounted to the copter. It barely clears the ground. There was still some headroom between the battery and the camera box mount, so I'm going to shorten the standoffs. This copter's all up weight came in at 4.94 kg, the heaviest copter I've flown!
It flew! Very stable due to how heavy the copter was. The only thing I was worried about was the standoffs. I had used rubber shock mounts on them, and the shock mounts turned out to be much too flexible. The payload was wobbling wildly back and forth the entire flight. I'm going to change the standoffs to use a different shock mount setup. The test flight was only a short hover and some travelling around the quad.
Even with the payload shimmying back and forth like it was at a rave, all the pictures turned out crystal clear! No blur, not from the payload mount and not from the prop vibrations. This picture is representative of the others. I flew over this grate because I knew it would be easy to tell if the picture was blurry by looking at the grate. Click to zoom in. Next step: pictures from higher altitudes and autonomous flight!
Time for another update. I decided to do an endurance test to see how long my new setup would stay in the air. I also wanted to know if the camera would have enough room on the card and enough battery life. So I set up the camera in its box and set the copter to loiter with a fresh set of batteries.
It flew for 20 minutes before Mission planner read 13% battery left and the in-the-air voltage was 13.5, so I decided to set it down there. It probably had a few more minutes left in it until the ESCs cut out, but I didn't want to push my luck. So, 20 minutes of safe flight time maximum. Should be plenty of time for my purposes.
I did notice that some of the pictures came out motion blurred. I think that it is because this time I shortened up the standoffs for the camera box until it was flush with the batteries. However, I think the camera box might need a little more freedom so it can dampen out those vibrations. So I'm going to lengthen the standoffs until the box just clears the batteries.