VCA from Denny Rowland on Vimeo.
I know from some contacts in the business that there are going to be some significant improvements in the supply of more suitable brushless gimbal motors. however I have always been of the opinion that a full movement 3 axis gimbal was quite a silly idea when the model itself provides us with a very good 3 axis of stabilized movement that needs only to be corrected for a small degree of unwanted movement. High end mil. spec. stuff works on the principle of having an inner (high sensitivity) and an outer axis that works independently to form a much higher degree of accuracy than we will ever see from the brushless system that we currently use. This inner axis is normally created around the use of voice coil actuators. They are far more efficient, faster, powerful and have a linear output. The name comes from the speaker coil that is used to vibrate a paper coil to create sound waves. It can also attenuate the kind of vibration that we see on a typical MR craft. There are however several different types.
So here's one in action on a development platform for a well known Vision MR that desperately needs it.
This prototype Rotary VCA came from the magnets out of a Hard drive and the coil was hand made to test the principle of controlling a single phase actuator with a standard brushless gimbal controller using just two adjacent output pins.
Shown here with a 7.4v. power supply but it has now moved on to 14.8v It is controlling the yaw function with Follow mode which means that it returns to to models heading after correcting for a disturbance. The yaw movement is not shown as it is going into production but I can tell you that it floats the full 95 deg. tilt axis and the camera assy on a magnetic field.
Comments
I would have to say that the only weakness in the system at present is in the controller itself in that it has too much drift which shows whatever motors you use.
The bench test goes like this. That rotary table projects a laser on the wall about twenty feet away and then a second laser is mounted on the gimbal So when I induce a movement the first laser shows the movement and the second one stays still and then it slowly joins the first one again as the FOLLOW mode takes over. It is very accurate compared to a normal brushless motor.
Rob
Not flying at the moment due to the weather and the flu.. I can pretty much bench test this stuff these days and know what will happen. I would not say by any means that the coil that I made for that test sample is right for the job as it runs too cold. That suggests that it needs less turns or thicker wire before I am happy it is right. The cost of production VCA's is outrageous so I will carry on with the testing. There are several other types that are better than the rotary that I made. The linear straight type is better but I need to get some boards made with a single phase output and channel mixing. Some of my very early videos had a type of VCA that stabilised a flat board to which the camera was mounted. It used big industrial servo boards that did the job another way with feedback.
I have two models that will be using this; one is a back-up Phantom Vision and the other is a new type of tricopter called Persistent Observer which I have estimated from the lift curves will fly for around two hours on li-ion cells. So the gimbal arrangement is tiny. The camera will come out of a Nokia cell phone. Nokia have just announced the reintroduction of their 41.5 megapixel Zoom camera. Plenty to keep me busy at the moment.
I think if you want a good little camera that probably wont need any stab. then look at the new Sony action cam.
Denny, I've added your latest video to the blog post and removed the video that is giving an error.
Denny, I'm pretty interested in this concept. Do you have any videos that show the end result performance?
What we need is for Jordi to design a board dedicated to a single phase output with the ability to mix outputs then some interesting coil layouts can be integrated into fewer components.
I am also testing some new tiny cameras that offer some remarkable performance. The new cell phone from Nokia re-introduces their 41.5 megapixel zoom camera and the new Sony Action Cam has a image stab. system that is matched to the right kind of frequencies that we may encounter.
The coil was entirely guesswork and is 275 turns of .21 wire (6.4 Ohms) nothing gets even warm so I could go a lille more aggressive on that to boost the power which is set to 255 which relates to full battery output.
I hope that this quick video shows that the VCA is being controlled by a standard brushless gimbal controller using two adjacent output pins to form a single phase supply. The set-up is handled in the latest GUI and it is not exactly plug and play but once you understand that the VCA has a limited range then you can adjust that range setting by the use of the pole count control.
Thanks for that HeliStorm, we live and learn.
-
1
-
2
of 2 Next