Fri Dec 19, 2014 11:50 am
I made a non-permanent AOA over a year ago, and finally got it working well last spring. It uses the SM screw that attaches to the nav light. It is battery powered, and lasts a dozen hours or more. It wirelessly transmits data to the receiver velcroed to the turtle deck, which is about the size of a pack of cigs. The reciever has LED's and an analog meter. It is trained during flight by pressing a button sequentially during a power-off stall, an engine idle target approach speed, and Vy that are calculated for a given flight weight. The AOA is indicated regardless of weight after the training is complete.
I 3D printed the probe body and receiver. I designed, laid out, fabricated, and assembled the PCA's, which include a recharge port for the Li batteries inside. I also wrote the code.
I wanted to make a commercial product out of it because the target retail price would be around 1/2 to 1/8 of the other units out there, and there is no installation cost or fuss aside from a 2 minutes install at the wingtip with a phillips screwdriver, and optionally some velcro for inside the cockpit.
The unit works well, and was interesting to play with for a while. I gave several units to people to evaluate. One was at a flight school, where it was reported to be very cool for their 152 and 172, and they were going to show it to students. The mechanic saw it, moaned, removed it, threw it away, and allegedly wanted to complain to the FAA I was "selling unapproved gear" or some crap like that.Really weird. I have not heard from anyone at the FSDO, and I highly doubt they would have a problem with something like this. Nonetheless, I was out money and effort because of this individual. The unit is less obtrusive than a GoPro by far. It generates interest by anyone I mention it to, but few pilots really believe an AOA will really improve their flying. I tend to agree, but...
Other people I talked to who had either installed an AOA or were definitely going to seemed like they would much rather install the 3x to 8x more expensive units they saw in their glossy magazines plus pay a lot to someone to cut tin and install them than use their own screwdriver to install these units, which are barely visible unless you know what you are looking at. Pilots can be a fickle bunch.
In any case, the unit works on Cessnas and Pipers and older Beeches,or anything that has the nav light screw I talked about for the most part. I decided not to pursue making them commercially at this time because I doubt the orders it would take to break even on parts alone (mold, etc) would materialize quickly enough in the climate I encountered the last go-around. I may get rid of the receiver and write iOS and Android apps instead to reduce costs even more, and appeal to younger pilots since it is clear the rest want their technology to be 30 years old, expensive, and have a bunch of paperwork to go with it like the units they see in the magazines.
In short, I think it works as well as either of the units I tried out during the development phase. Each of them was between 1200 and 2000 to install with labor and all. Mine would have costed a small fraction of that and required no installation costs.
In terms of using a Pi or its knock offs, it would likely be an exercise in overkill. I get data at 3Hz with a $2.08 Atmel, and my code base is roughly 3k because it was initially written in about 6 hours without going back to tidy things up . A lot more time was spent in tinkering with the code, but nothing aimed at cleaning it up. A M0 would save some cost but take more real estate. A Pi would leave 90% of its expensive resources and almost all its clock cycles unused. One use of the Pi would be to use the VGA output to mask to an MFD via the AUX port for glass panels. Those that fly that equipment might be more inclined to pay the Pi price. Unfortunately, many avionics guys swear up and down that the AUX port connections have to be certified somehow, even as I've seen everything from cheap DVD players to play My Little Pony movies to game systems hooked up to the G1000 AUX video input.
The next rev would reduce the size of the Tx board and improve battery life with a different SoC wireless solution, possibly even eliminating the Atmel, and get rid of the dedicated RX and moving it to an app (since all the cool kids seem to find that infinitely more interesting than a dedicated display).
Unless the demand is there, it just hasn't penciled out very well to make it an obviously good way to spend more of my time.