![]() |
iPad EFIS Update
So back at the beginning of the year I posted a tentative panel design and got quite a bit of useful feedback from the VAF community. The design was (is) intended as a showcase of the iPad EFIS that I've been working on for the past three years. Here's what happened next...
Last month I finally got the EFIS hardware and software to the point that it was ready to move to the next stage of testing. For the past year I've been flying with the display for my EIFS (an iPad) on a RAM mount. It was my primary source of flight and nav data, but I still had the Dynon and steam gauges for backup. In almost 80 hours of flying, the numbers had matched up - airspeed sensor is working. Altimeter is working. GPS is working. The time had come to take it to the next level. The next level can be summed up best by an old joke. If you're familiar with a bacon and egg sandwich, you know that it requires contributions from both a chicken and a pig. The saying goes that while the chicken is "involved" the pig is "committed." It was time to move from chicken to pig. Getting to pig stage involved removing the panel. The Dynon SkyView - gone. The Garmin GNS-430 - outta there. The steam gauges - history. The switches came out. All of the wiring came out. In the end it was more-or-less tabula rasa. About the only survivors from the original panel were some Klixon breakers. In the end it was a pretty sad sight... ![]() The next step was to rewire everything from the lights to the starter switch, to the autopilot. All told, the process took about 50 hours of work. Some of this was replacing wire (everything is now Tefzel). Some of this was the learning curve (you have to make space for things like antenna leads, power leads...). In the end, it was actually less painful than I thought. I ended up adding service loops and removable pigtails to make it easier to remove the panel. I also had to work around a sub-panel that had been chopped up to allow various legacy avionics to fit. ![]() Once all the new stuff was installed, I mounted a few other bits - the VAL 2KR Comm radio, the Trig remote transponder (a survivor from the SkyView system). The PS Engineering PMA8000BTi audio panel and the Apollo / Garmin SL-40 COM both moved to the new panel. Once everything was mounted up, I installed the panel (for the first of about 10 times). ![]() While I had the seats, floor, and pretty much everything else out, I installed a TruTrak Vizion autopilot. This replaced the Dynon autopilot, and fortunately Dynon used the same mounting structures as TruTrak, so the installation was relatively painless. The only challenge was building and running the harness, which took about five hours between pitch and roll. Running the pitch harness through the very limited space in the floor tunnel was probably the most challenging bit. (Many thanks to my friend Eugene for helping out with the harness and for holding the fan in place while I worked in the tail cone in 100? heat.) ![]() When all was said and done, the new hardware was working. The autopilot passed its ground check. The engine monitor was showing rational cold values. The iPad chargers were charging and coolers were cooling. The next step as to haul it out of the hangar and see if the engine would start. Fortunately that turned out to be a non-event. The new mag switches and starter button worked. The new engine monitor showed what I expected to see. All was good. ![]() I taxied around for ten or fifteen minutes and nothing exploded. At that point the last step was to re-route the throttle and mixture cables. The builders had opted to install both on a vertical riser that ran from the top of the fuel pump shroud to the bottom of the panel. The riser also held the manual trim control. As you may be able to see in the picture above, I had separated the throttle and mixture section from the trim section. My A&P friend Robert did most of the work of moving the controls to a very nice 9" milled aluminum throttle bracket from SteinAir. (At some point I plan on replacing the manual trim control with electric trim but for now the manual wheel remains.) The next day I ran the entire system through an extensive ground check. I re-tested the autopilot. I ran a basic pitot / static check (no leaks). Once I was sure it was about as good as it was going to get, I took it up for a brief test flight. In about forty minutes a was able to confirm that everything was working. Instruments were all in the expected ranges. Autopilot track and nav modes worked as advertised. Flight characteristics where not noticeably different. Here's a quick video showing the system in action: https://www.youtube.com/watch?time_c...&v=Md6_hLj8YzU Since then I've put another three hours on it, all nominal. So in about 70 hours, over the course of two weeks, N101PR underwent a pretty dramatic transformation. All in all I'm happy with the results. The new panel does a great job of showing off FlightView (the new EFIS). The airplane is roughly 20 pounds lighter. The power draw is minimal. The new layout is more intuitive. The new autopilot is a bit smoother. The new air vents provide a much needed boost in ventilation. Best of all, it's a bit more "my" aircraft. Time well spent. For those who are interested in the details on FlightView, I'll post something tomorrow. You can also check out the overview on my web site here. |
Steve, thanks for the update----I had been wondering about how the project was going.
Congrats on moving up a level with the plane. What is the long skinny unit just below the left iPad? |
Great looking panel. It?s just amazing how quickly EFIS are evolving and the functionality that is available. Love the weight savings too. Im a few years away from completion and imagining how much more changes will happen by then. I use iPad w ForeFlight as backup to my Garmin 430 but this is really tempting for VFR only folks I am sure. Thanks.
|
Very Cool
Looks nice. I'm very interested in learning more about this... mainly how the iPads hold up in extreme heat. I'm sure you have fans behind the panel, blowing on them but I know in the hot summer day that may not be enough.
Very very cool. can't wait to see your report in flight. Best of luck. |
Looks great, and I've lusted after this concept for years. But any idea on whether the Flightbox guys will support the android platform? I moved on from Apple a looong time ago, in favor of more options, more flexibility, brighter screens, higher resolution, less heat sensitivity, less *money*, etc.
Thanks, Charlie |
Quote:
FlightBar is the fix for that. The knobs controls zoom on the MFD and various values (baro, bugs, etc.) on the PFD. The buttons map to the soft buttons at the bottom of the screen. Makes the system even easier to use. Here's a pic that makes it a bit easier to see: ![]() |
Quote:
I can see the advantage - shorter runs from the sensors; only a single data cable running through the firewall; easier access to the box (depending on where you mount it). Anyone know what the average temperature is forward of the firewall? |
Quote:
The other advantage to the iPad is that there's one manufacturer and a limited number of form factors. With Android it is very hard to build something as complex as FlightView and have it operate properly on all tablets. Some Android tablets are great (Samsung). Some are just horrible (cheap stuff from AliBaba). Some use non-standard screen sizes. Some run relatively cool (Samsung) while others would work as a backup cabin heater. The iPad is not a perfect platform, but it's a consistent one. I can reasonably be expected to own one of each supported model and understand the variables associated with it. Also, for whatever its worth, the underlying iOS-ness is largely irrelevant in this application. Once you fire up FlightView you're using an EFIS, not a tablet. |
Quote:
The cooling mount that we're offering (FlightDock) includes a variable speed fan with a thermal control module. It keeps a constant medium-volume flow of air across the back of the iPad. As the panel heats up, the controller speeds up the fan. On a truly hot day (I've tested it up to 109?F) it blasts the iPad. iOS includes and API that allows the app to subscribe to the thermal stress state of the system. FlightView subscribes to this and uses it to monitor the health of the platform. The scale is pretty rough: nominal; fair; serious; critical. The OS starts throttling the CPUs if you tread into "serious" territory. The app includes a little thermometer icon that appears if the stress level gets above fair. Yellow = serious. Red = critical. For whatever it's worth, I've never seen red in flight. I have seen yellow on some iPad models, and the performance definitely degrades a bit. I'm working on some code changes that will being disabling various secondary features which require quite a bit of power. This should help. Finally, I'm working on a different cooling rig for the FlightDock. It uses a thermoelectric cooler module (Peltier) to chill the iPad. The design I have can be easily swapped in if you happen to live in a particularly hot place. It will probably add something like $100 to the total cost and will draw a good bit of power (upwards of 5 amps at 14v) when running. |
This is looking awesome and sounding like a great fit for my future kitfox project :) A couple questions:
- AOA support in development? - Is the code open source? - What kinds of knobs/buttons are you using? I've found that for example the Garmin knobs feel much more solid than the Dynon knobs/buttons. - Any electronic circuit breaker system in the future? :) - I notice the rpm in the video is flickering into the red. Signal smoothing issue or were you just flying close to redline? - Will you be at OSH where we can see it in person? - Synthetic vision? - charts? How do they get updated? - flight plan integration with foreflight/gpilot? Thanks! -Guil |
| All times are GMT -6. The time now is 07:07 AM. |