While construction slowly (very slowly) goes on with my big hexacopter, I’m still working on my little bird for aerial photos and video.  The little bird being a DJI Phantom 1.2 with an Arris CM3000 3-axis gimbal.

I did two test flights yesterday, both with the heavier MadDog 2700mAh battery.  That battery is bigger and more heavy than the stock battery, and seems to make a big enough difference in weight that the Phantom drifts vertically and horizontally.  I also have heavy wobbling and shaking, making video unusable.  That’s the exact description of yesterday’s first flight.  Also as part of the test in the first flight were four hard rubber isolation grommets, as opposed to the very soft ones I’ve had decent results with lately.  Once again, bad vibes.

I’ve finally gotten my hands on a decent blade balancing rod for the capped Phantom 2 propellers I put on the bird.  I switched to these props because the craft has more lift and efficiency with them, as well as double the flight time per battery or more.

I did a rough propeller balance after yesterday’s first flight, and also changed two of the four hard grommets back to soft.  This flight was massively better with only a small fraction of the flight producing vibrations or wobbles.  The problem, and the lesson, is that I changed TWO different things on the bird.  So I don’t know if the change in grommets or the blade balancing was the source of the improvement.  It is also possible that both combined made the difference.

Today’s first test flight (if there is one) will have to be blades balanced with four hard grommets and the heavier battery, since I don’t want to unbalance the props.  Then based on those results I may modify the grommets for the next test.

Following, perhaps a test with the lighter, original DJI batteries.

I had good results with the soft grommets, but they’re so soft I’m afraid the whole gimbal may fall off during flight.  It has come loose during transport in its case.

Frustrating few days.  In attempting to increase my functions and capabilities to include tilt/roll of my gimbal (meaning it could point forward or I could rotate it 90 degrees to do straight down shooting) I screwed up the settings it came with.  First time I plugged it in it nearly bounced off the table top having seizures which would make the worst epileptic look healthy.

I’ve been learning the software which controls the gimbal, SimpleBGC, and slowly working the unit back to functionality.  So far I have two of the three axes working, but the yaw axis (where the camera follows the direction the bird is pointing) is reacting very slowly.

Very frustrating, but I’m learning so much each time I go through these experiences.

Last week I managed to roll some video with the new gimbal down in southern Utah.  The results were fantastic.

I’m limited by the air time I have though, at a safe time of about four minutes with the reduced setup (3-axis gimbal, small mount, 3D printed landing gear extensions, stock DJI Phantom 1 battery).

Some friends critiqued the video and made some great suggestions and observations.  One observation is that there’s still a tiny bit of “jello” effect at some points.  The jello effect comes from the drone vibrating the camera.  The video shakes and looks like jello shaking.

Two directions to go with the jello are to balance the propellers and to work on the mount for the camera.  I did both, and tried some new batteries.  There were good and bad results.

The new batteries are 2700mAh compared to the stock DJI batteries which are 2200mAh.  The net result was over seven minutes in the air, a great improvement.  Since the battery is bigger though, it is much harder to get into the craft.  Good.  More flight time.

I thought perhaps that the rigged setup I had last week was part of the reason for the slight jello issues.  The new gimbal came with eight jello reducing rubber grommets and I flew with four.  That’s the only way this rigged setup would work.  First test today I put back on the small mount with eight jello reducers.  The problem there is that the mount blocked the battery door from opening all the way.  So I got some throw-offs at a local electronics store and lowered the mount half an inch.  See image below:

The result was horrible.  The camera and craft were jiggling terribly.

At that point I decided to go back to the big mount and leg extensions which I purchased with the gimbal.  This time though I went with the shorter of the two possible setups (image below).

I ran into some jello issues again.  I’m thinking perhaps the reason is that the eight grommets are too solid and don’t isolate the camera from the body of the copter.  So my first test tomorrow will be to take four of the eight grommets out and see what effect that has on the jello.

Another small issue too, is that the blades are sometimes visible in the video frame.  This issue may lead me to putting the longer leg extension setup on, as it moves the camera farther from the blades.

In Short

To make a long story longer, I’ve increased flight times which is great.  I’m not satisfied with the shake reduction yet and still working to improve that issue.

On the jello effect, I understand running the video at 60 frames per second instead of 30 helps.  I may try that too, but unfortunately I think the GoPro will not allow the timelapse photo mode to run simultaneously with the 60 frame setup.  Right now I’m shooting 30FPS video and photos every two seconds, which is great.