Saturday, August 23, 2014

Servo projects

No flying this Myday as my luck has finally run out in the realm of good weather - hot, muggy, and threatening thunderstorms in a most convincing tone. I took the opportunity to visit the new local indoor karting track, which was a ton of fun, but at $20 for an eight minute race (well, cheaper than that with a Groupon discount, but still...) it only took an hour - how to fill the rest of the day?

Easy: pull some projects off the shelf.  There's some back story to this, but I will try to keep it brief.

I all started with the airsickness detectors. These are little doors in the side of the airplane, one on each side, that are intended to allow a stream of cooling air into the cockpit which, if you think about it, it nothing more than a greenhouse in the warmer months. I call them 'airsickness detectors' due to the propensity of those passengers who are feeling a little queasy to "need more air."

In any event, once strapped in and settled into my seat, it is difficult to reach the control arm of the left side door and impossible to reach the right side. I started to think about a way to fix that to make it easier for a single person to adjust both doors. Having spent quite a few years of my youth building, flying, and crashing R/C airplanes, my thoughts immediately turned to servos, the little electric motors that move the control surfaces of the airplanes. The difference between a servo and a regular electric motor is one of proportional control - move the joystick in a specific direction, and the servo will track that movement proportionally. That means I would replace the joystick with a knob and use it to control the exact position of the door.

To do so requires some electronics. I spent quite a bit of time searching the internet for a way to create the electronic device required to do it, but eventually came across something better. This is exactly what I needed: the Dual Servo Driver from I'm going to go off subject here for a second to say something about ServoCity: buying stuff off of the internet is typically a impersonal affair, but something about the folks that run ServoCity makes it feel like you were right there in an actual store and waited on with a smile. Maybe it was the little packet of candy that they included with the shipment, or their very personable presence on Facebook, but either way it was a very satisfying experience.

So, with the servo driver on order it was just a matter of finding some servos. They aren't cheap, after all, and I wanted to be sure that this experiment would work before springing for them. Not a problem: I borrowed a few from The Jackson Two, who had also followed the tried & true path from R/C to real-world airplane building.

I got as far as building up a servo with a control arm that would fit well with the handle on the doors, buying a voltage regulator that would drop the 12VDC of the airplane power to the 5VDC demanded by both the servo controller and the servos themselves, and buying a pair of knob controls from China, and then....

I am going to digress for a moment here to introduce the technology that has emerged as the very pinnacle of the Selfie Generation's narcissism: the GoPro Action Camera. This tiny little device can be attached to just about any surface and provides crystal clear point-of-view video from all sorts and adventurous that most of us will never experience in person. Here are a couple of examples:

Things to note:

  - The camera(s) can be mounted to just about anything, but once mounted, the composition of the shot is fixed. This is true even in the karting video; the only reason that it isn't a fixed view is that the camera itself is mounted on a moving platform.
  - My life is unexciting - I feel like I could not live up to the standards of these GoPro users.

This changed my mind:

Having an airplane, I figure I can at least top that last one with no great (or risky) effort.

Although the in-flight portoins of this video were recorded with a competing brand to the GoPro, the concept is still similar:

Having mooted the second issue, I was left with the first issue to address: what to so about the fixed composition? It wasn't that big of an issue with the camera right there next to me in the airplane - if I wanted to bad enough, I could adjust the direction in which it was pointed. But what if the camera is mounted outside the cockpit?

Consider this video:

That required at least three cameras. They cost $400 each.  There has to be a better way to get more variety in the outside segments....

I think there is, and (not) surprisingly enough, it involves servos.

I got to thinking that only two servos are needed: one to pan the camera (rotate it left and right) and one to tilt the camera up and down. With two servos already in hand, I again turned to the internet for help and again found what I was looking for at Servo City, the SPT-100 Pan & Tilt System:

I would still need a servo controller, but rather than knobs I would want a joystick. Incredibly, Servo City had anticipated just that need with the Two Servo Joystick:

I ordered both products, along with a few smaller accessories. If you look at the pan & tilt, you see that the bottom servo serves as the base for the whole shebang. The next question was how to mount that to something solid.  Easy! Servo mounts.

Now it was just a matter of deciding where to mount it on the plane. I consulted with Kyle, Smoke Oil Manager/Technician for The Jackson Two Single-Ship Formation Team, who had previously mounted a GoPro to his plane. 

I decided that I wanted it to be on the bottom of the wing to allow for better viewing of the ground. This would be a problem for the camera operator, though, because he wouldn't be able to see where the camera was pointed (assuming a low-wing plane), but it turns out that a helper app can be installed on a smart phone and the app provides an electronic viewfinder. Perfect!

We decided on two potential mounting spots that would allow the mounting of the pan & tilt system to be installed without drilling holes into the airframe: we would mount it to one of the removable inspection plates under the airplane. I voted in favor of the one on the left wing just below and behind the leading edge, while Kyle favored one down in the belly of the plane. Both of us have expensive paint jobs on those panels that we were reluctant to mar, so Kyle came up with the idea of just buying new ones from Van's. They're dirt cheap, so why not??

With the not-so-great weather we're having this weekend, I finally got around to kludging together a rather sloppy mount. I used .063" thick aluminum to form a doubler to provide more rigidity and bolted the mount directly to that. The cardboard box is a simulated camera:

It only kinda-sorta works. There are a number of problems:

  - I got the orientation of the camera wrong, and it requires the complete disassembly of the mount to change it.
  - The servo attachments are too weak for this usage. There is a very real chance that they would work loose and drop the camera. It's not unheard of.

 - The mount itself if not rigid enough to hold the camera steady - this is also partially due to the servos hanging upside down. 

  - The mount sticks too far out into the slipstream. That could put a lot of pressure on the mounting plate. The .063" doubler is probably string enough, but it would be nice to have as little drag as possible.

This is the state of the project today, but after a little more research, I think I might have a fix for those problems. As it turns out, there is another mount that addresses pretty much all of my concerns: the SPT200 Pan & Tilt System, which mathematically speaking should be twice as capable as the SPT100. It's also twice as expensive, so there is that to consider.  It is designed for a heavier camera and has better rigidity, the servo mounts are far more robust, the center of the pan axis can be adjusted without taking apart the mount, and the bottom server can be installed inside the airplane.


I'm going to contact the folks at Servo City and see if they agree that this mount should be up to the task. If they agree, the next step will be to order one and give it a try.


I received some feedback from others that have experimented with GoPro mounts - it seems that the protective plastic case of the GoPro is too flexible/not rigid enough. This allows the camera to vibrate, and its internal attempts to stabilize the picture cause what they call "jelly rolling."

This makes me wonder if I should be looking into a very rigid mounting system like this:

More specifically, in this configuration:

With the bottom of the servo inside the wing or fuselage, the aluminum mount plate and camera supports would provide a great deal of rigidity, but at the cost of the 'tilt' axis control and the protective box for the camera. I'm wondering if it is worth the trade-off. At a minimum, I think I would have to find a way to allow the tilt to be ground-adjustable.

An example this problem and ways of addressing it in a different modality of flight:

No comments:

Post a Comment