EFIS For Homebuilt / Experimental Aircraft.

Discussion in 'Instruments / Avionics / Electrical System' started by JMillar, Apr 11, 2008.

Help Support HomeBuiltAirplanes Forum by donating:

1. Mar 22, 2017

donjohnston

Member

Joined:
Oct 9, 2016
Messages:
11
4
Location:
Panama City, FL
If you've got the electronics and programming ability, yes. You could do it and save some money.

Not necessarily. At some point you have to decide would you rather A) build an airplane with some minor personalization and end up with an actual airplane that flies. or B) design and build subsystems forever on an airplane that never leaves the shop.

You just have to decide where that line is.

I spent about 15 hours designing and building my own annunicator panel. Total cost was about $25. I could have easily achieved the exact same results with 1 hour of time and$10 in lamps. I decided it was worth it.

In the case of an EFIS, that's a rather large undertaking... at least for me.

Attached Files:

• 2013-08-14 IMG_20130814_191418_518 (Large).jpg
File size:
32 KB
Views:
44
ninelima, fly2kads and gtae07 like this.
2. Mar 23, 2017

ninelima

Member

Joined:
Jul 26, 2015
Messages:
10
3
Location:
South Africa
I have to agree with donjohnston.

Any homemade EFIS with decent performance will probably require a commercial AHRS. I tried building a few homemade AHRS's using MEMS/Arduino/Sparkfun etc over the years, but I have never managed to get anything comparable to the commercial offerings (Garmin G3X, Dynon, MGL et al).

The display and presentation of the information is also not trivial. Making smooth graphic displays can take some time to achieve. For example Kwik EFIS is the result of a number of prior projects over at least 3-5 years.

For what its worth, the entire codebase for Kwik EFIS is available for anyone to use. It would be easily adaptable to Intel based embedded processors. https://f-droid.org/repository/browse/?fdfilter=kwik&fdid=player.efis.pfd

With the code and depending on your backround, you should probably budget at least 3 months to make the code "your own".

Last edited: Mar 23, 2017
3. Mar 23, 2017

spaschke

Well-Known Member

Joined:
Oct 24, 2012
Messages:
221
48
Location:
Denver, CO
I have started working on an AHRS like what you describe. arduino with an IMU and maybe another sensor or 2. It will be a lot of work. I have also considered using a "PixHawk" or similar R/C controller, modifying or replacing the software. I haven't completely ruled it out yet.
There are several filters you have to run the data through to make what I call "steady" data. It is my opinion that you need to have at least 2 IMUs and compare the data to avoid the "occasional data spikes" and not use the cheapest IMU's. Know your data so you can through out the bad.
The commercial offerings mostly use the same "better" chips. I would guess that part of the reason they are so expensive are the redundant subsystems needed to get "steady" data.
I agree that it will be a long process even if I use an existing open source project.
I am starting with a more simple instrument. A directional indicator, Altimeter, VSI, then on to the EFIS. Eventually I would like to try to use SDR to make a VOR and/or ILS receiver.

My plan is to use Raspberry PIs to do the display. Arduinos will do the data gathering. I have put some thought into attaching the arduino shields to the Raspberry PI via a shield interface board. I just bought one to test with.
The Intel Edison would be a great controller platform also, I just don't know anything about it. And also, there are soooo many supporting Arduino shields and PI daughterboards out there.

4. Mar 24, 2017

spaschke

Well-Known Member

Joined:
Oct 24, 2012
Messages:
221
48
Location:
Denver, CO
So I read about the intel edison and it is not very similar to the Arduino nor the raspberry Pi, it's primary function is an internet connector-processor.

Another option is use a smartphone or tablet with EFIS software, using data fed from an Arduino.

5. Mar 24, 2017

ninelima

Member

Joined:
Jul 26, 2015
Messages:
10
3
Location:
South Africa
Oleg from http://experimentalavionics.com/
has apparently had some success with the Adafruit BNO055 9 axis absolute orientation sensor board.

He sells blank PC bords and some other components for not much money. It might make a nice starting point.

Last edited: Mar 24, 2017
6. Mar 25, 2017

ninelima

Member

Joined:
Jul 26, 2015
Messages:
10
3
Location:
South Africa
This could work. It is the way I am thinking. The tablets are very capable and sophisticated pieces of electronics, considering how ridiculously little they cost.

Having said that, Daylight / Bright sunshine view-ability is a bit of an issue ... see: [video=youtube;-VKt1pFnpbY]https://www.youtube.com/watch?v=-VKt1pFnpbY[/video]

I would try the "Adafruit BNO055 9 axis absolute orientation sensor board" with a WiFi or USB to a tablet. If the BNO055 is good enough, it should be possible to get a decent "el cheapo" EFIS for a couple hundred dollars.

Last edited by a moderator: Mar 20, 2019
7. Mar 25, 2017

cluttonfred

Well-Known Member

Joined:
Feb 13, 2010
Messages:
6,351
2,208
Location:
World traveler
What about taking a different approach and not trying to integrate everything into one device? A simple four-digit segmented LED and added LED bar, arc, circle or matrix would provide bright digital and analog readouts, could be driven by a very small, cheap Arduino board and fit in an ordinary instrument panel hole. You could have a single, standardized display design and only the sensors and software change between instruments. Start with something simple: ALT/VSI, ASI, compass, and eventually work up to the gyro instruments. I think you'd also get a lot more people interested in tackling more modest projects like this than a full EFIS.

8. Mar 25, 2017

Hot Wings

Well-Known MemberHBA Supporter

Joined:
Nov 14, 2009
Messages:
5,987
1,999
Location:
Rocky Mountains
I've been thinking about this kind of thing for a few years now. Maybe it was earlier in this thread...? I've now given up trying to reinvent this particular wheel since there is off the shelf hardware available for what I consider a VERY reasonable price that does pretty much what most of us amateurs can come up with. The data gathering electronics are good enough now to be usable but the display side is still a big problem - IMHO.

I'm in the minority and think there is far too much information being displayed and hate having to use touch screens in moving vehicles. (That's a subject for another thread) Finding cheap daylight readable displays is also seemingly impossible.

What I'd like to see is a standard developed for the transfer of data from one device to another similar to the standards used for things like CAN bus and ADS-b data. Maybe something as simple as a CSV file with a specified format would work? If we had this kind of standardization we could more easily mix and match the hardware and software from the various sources, both professional and amateur.

On the display side of this problem I've been looking at augmented reality glasses recently as maybe being a practical option. They are kind of expensive but seem, to me at least, to be the area we amateurs should be spending our time experimenting?

http://store.reconinstruments.com/Recon-Jet

9. Mar 25, 2017

cluttonfred

Well-Known Member

Joined:
Feb 13, 2010
Messages:
6,351
2,208
Location:
World traveler
Hi, Hot Wings, and yes, we've had a few similar threads including Simplified electronic gauge format and Arduino altimeter & airspeed indicator. I am not so sure that we need someone to create a data standard, rather we need someone to create a successful design and that will become the de facto data standard.

When I see the amount of money that Belite charges for their electronic instruments, I really think there could be a market for something like this. A simple, standardized enclosure in standard instrument hole sizes with a segmented LED or a bright white monochrome graphic OLED for the data, a rotary encoder with button function for settings and adjustment, and maybe a circle of 16 addressable, color-changing LEDs to serve as the dial. That same circle could be used for all the dials to provide at a glance indication of airspeed, vertical speed, engine RPM, compass north, etc. I also like the idea of large white decal icons on the faceplate to distinguish the instruments from each other.

I think the smart thing would be to make the displays/enclosures a standardized component into which you can then put whatever logic board and sensors you need. When I look at the costs of components on something like Adafruit, I think an altimeter or compass could be done for about $50, airspeed a little more, gyros about double. That's pretty competitive and would be very interesting for folks not looking for a full glass panel--the portable GPS or tablet strapped to your leg meets a lot of those needs already. 10. Mar 25, 2017 Hot Wings Hot Wings Well-Known MemberHBA Supporter Joined: Nov 14, 2009 Messages: 5,987 Likes Received: 1,999 Location: Rocky Mountains It would be nice if someone had the time and talent to come up with such a standardized design. I'm not that guy.:ermm: Unfortunately in our capitalist system someone that does have the time, talent, and bankroll to do this is likely to make their system dependent on propriety software or hardware. That seems to be the situation with current EFIS on the market. Can you imagine trying to interface GPS data with devices if the NEMA standard didn't exist? Or trying to sell a device that needed your own proprietary format to transfer that GPS data? The only way a manufacturer could get away with that kind of lock on their product is if it was combined with some other functionality that wasn't available readily from another source, thereby forcing you to buy their proprietary GPS unit. Where have we seen this kind of market manipulation before.....?? The hardware for what we are thinking about really isn't a big problem - at least from the electronics side. The packaging/housing to put it all together in a neat and professional manner - is lacking. With 3D printing evolving so quickly that may not be as much of a problem in the near future. This is where the Belite instruments have made their market. From the electronics side of things there is nothing in their product line that is all that impressive. The prices, compared to what you or I would spend to duplicate a similar unit, seems pretty reasonable. And I'm a cheap CHEAP guy! They have nice neat professional packaging, sunlight readable displays and we don't have to spend a couple of hours assembling a kit and uploading software to an Arduino board.* The failure of the Belight line of products, for me, is that they are all individual units - lectronic replacements for the old standard steam gauges. I'd like one display module (maybe 2) that would display all of the functions being monitored by command, or only when one is approaching some preset limit. I've got other things to do when flying than constantly monitor gauges that are only telling me everything is normal. That's what those little electronic Minions that reside in IC chips are for. :gig: * But that IS part of the fun, much like building your own plane. 11. Mar 26, 2017 Slars Slars Active Member Joined: Oct 26, 2014 Messages: 33 Likes Received: 9 Location: Carlisle, PA / USA I have been looking into making an EFIS using Arduino and either a PC or Raspberry PI for the display. There appears to be drone or RC guys out there that have made considerable progress. The one I found the most interesting and informative was using the MPU-9250 9 DOF chip and a BMP280 pressure chip running on a Teensy 3.x. Both are available from a number of sources. If you do a google search for "mpu 9250 Kris Winer", you will find lots of good information. I have also experimented with sunlight readable screens and have been pleasantly surprised by how good glossy LCD screens work in direct sunlight. They must be glossy screens, not matte or covered with a glossy piece of plastic. Just angle them down slightly so you don't get the direct reflection of the sunlight and you're good to go. The lighter background seems to work the best. Here are a couple pictures: cluttonfred likes this. 12. Mar 27, 2017 spaschke spaschke Well-Known Member Joined: Oct 24, 2012 Messages: 221 Likes Received: 48 Location: Denver, CO Thats what I just purchased. Haven't opened the package yet. 13. Mar 29, 2017 spaschke spaschke Well-Known Member Joined: Oct 24, 2012 Messages: 221 Likes Received: 48 Location: Denver, CO I hooked up the BNO055 last night and played around with it. It gives very steady data. Much better than the first board I had by WaveShare. 14. Apr 1, 2017 ninelima ninelima Member Joined: Jul 26, 2015 Messages: 10 Likes Received: 3 Location: South Africa Sounds good. The acid test if however if it works under acceleration. If you can get some early flight test data on the BNO055, A single coordinated 360 degree turn will tell you all you need to know I'd really like to hear your results. It would be great to have something that finally works. 15. Apr 18, 2017 BJC BJC Well-Known Member Joined: Oct 7, 2013 Messages: 8,879 Likes Received: 5,732 Location: 97FL, Florida, USA Saw this at Sun n Fun, priced at$500.

http://www.talosavionics.com/aeolus-sense-3b/

If / when someone develops a similar box for an engine / fuel / electrical system monitor, a tablet-based instrument panel will be cheap, light, and easy to install. Use solar cells to charge the system battery and avoid the transponder and ADS-B requirements.

BJC

16. Apr 18, 2017

Pops

Well-Known Member

Joined:
Jan 1, 2013
Messages:
6,671
5,552
Location:
USA.
What I am doing on my JMR Special but using mechanical oil pressure and temp gauges with a tiny-tack for rpm for the engine instruments. Two solar panels under the windshield.
My hanger and workshop is on solar except for the air compressor and welder.

BJC likes this.
17. Apr 18, 2017

tspear

Well-Known Member

Joined:
Feb 13, 2014
Messages:
762
165
Location:
Oneida
Going off topic, instead of building the ADHRS system itself; I thought Aspen and GRT published the WiFi connection protocols, and how to get the data. This would make it easier to develop something cool, like a HUD.

Tim

18. Apr 18, 2017

cluttonfred

Well-Known Member

Joined:
Feb 13, 2010
Messages:
6,351
2,208
Location:
World traveler
BJC likes this.
19. Apr 18, 2017

BJC

Well-Known Member

Joined:
Oct 7, 2013
Messages:
8,879
5,732
Location:
97FL, Florida, USA
20. May 15, 2017

Joined:
Nov 12, 2011
Messages:
96