DIY gauges??

Discussion in 'Instruments / Avionics / Electrical System' started by Aerowerx, Oct 6, 2014.

1. Oct 6, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
It will be quite a while before I need them, but I have been thinking about what type of instruments I will need. Of course, for an experimental aircraft you don't legally need any instruments. My current 'want' list consists of Tach, Airspeed, Altimeter, Oil Temp, Oil Pressure, Fuel, Turn Coordinator and/or slip indicator, and a magnetic compass. My smart phone can be used for navigation, and a handheld for communication.

There is no need to pay for TSO'd gauges. Non-TSO'd, from what I have seen, are $150 to$200 (USA) each. What are the alternatives?

There are of course, the digital instruments that use some kind of LED/LCD display. Also apps that run on a smartphone or notepad.

But for those who would prefer "real" steam gauges, there is an alternative.

I guess there are a class of people even more crazed than us HBA'rs. There are those who build their own flight simulators, and some of them use quite realistic looking simulated instruments. One supplier is Sea Gull Simulation. And Mike's Flight Deck tells you how to build your own, which can use servos, synchro motors, or automotive type air-core gauges. Mike's web sight even shows how to build your own air-core gauges from scratch.

What I have in the back of my mind is one or more microchip computers with some of the analog sensor chips. The microchips would process the analog inputs and display the results on the gauges. Most of the Flight Sim gauges I have looked at use some kind of digital protocol (such as CAN bus). I guess you could do that if you wished, but I think that would add unnecessary complexity. Just process the inputs and control the gauges directly. Most all microchips currently available would have more than enough power for the task, and many of them already have the necessary analog outputs and inputs.

I haven't looked into it any further than this, but my guess is that, for someone conversant in microchips, it should be much less expensive than even a set of non-TSO gauges.

Has anyone else thought of this idea?

2. Oct 6, 2014

gtae07

Well-Known Member

Joined:
Dec 13, 2012
Messages:
1,795
1,063
Location:
Georgia
For a day-VFR airplane, it might suffice. Me personally, just the time spent to try and figure out how to make it work makes it more economical just to buy some non-TSO gauges (and yes, I realize we're already talking about building airplanes vs. buying them). Do a little looking and you might be able to pick up some used ones for cheap off an RV guy upgrading to glass.

Matt G. and dcstrng like this.
3. Oct 6, 2014

Jan Olieslagers

Well-Known Member

Joined:
May 28, 2014
Messages:
134
19
Location:
Haacht, Belgium
If once you decide to use electricity as the medium to _get_ the data, there is not a single reason to not use electricity to _display_ those same data. I think there are potentially five categories:

1) engine data. These all come in the form of electric values or at least can easily be made to. No discussion here: use standard automotive sensors, then display the data either on "steam" electrical displays or in some electronic device, like the Flydat that is available as a standard accessory for Rotaxen.
2) static/pitot data, i.e. ASI, altimeter, vario. Here you have a choice between (real!) steam gauges or fairly commonly available sensors and a bit of electronics. I have not seen this as a standard electronics kit, but haven't done much a search either... It cannot but be available in the market, and at reasonable cost.
3) information that can be derived from gps data. Thinking primarily of a HSI here, or a simpler form of gyroscopic compass. If are as lucky as I am, to live and fly in an area with next to no magnetic variation, this is peanuts for an able programmer. If not, a way will have to be found to reliably determine and interpret magvar. Should not be too hard either.
4) gyroscopic stuff like artificial horizon and turn coordinator. This is the tricky part: it can be done, doubtlessly, but I see too much evolution as yet. But do you really require these?
5) fuel flow: should be equally easy to implement on some microchip, but reliable flow sensors still seem ridiculously expensive. And here again, do you really require this functionality?

PS I fully agree that passing through CAN only adds complexity. OTOH nothing stops one from adding a CAN interface to whatever is created, though I can't really see the advantage.

Last edited: Oct 6, 2014
4. Oct 6, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
What I was referring to, in regard to the CAN bus, was that the simulated instruments I linked to are used for flight Sims. They would get their data from a PC running Flight Sim, where it would be an advantage.

For our application, though, I don't think it is needed. It adds unnecessary complexity and more possible failure modes.

One concern I have is if the DIY electronics can stand the vibration during flight.

5. Oct 9, 2014

rdj

Well-Known Member

Joined:
Sep 10, 2009
Messages:
288
185
Location:
Northern California
I doubt it would be less expensive. First, it's more complicated than just putting analog inputs into a microchip. Most of those analog inputs are going to need outboard conditioning circuits to enable the microchip to make any sense of the input. Tachometer pulses and temperature, level and pressure sensors are notoriously noisy, and will need both filtering and level conversion before the microchip can use the inputs. You'll also almost certainly need additional filtering and calibration in the microchip software as well.

7. Oct 9, 2014

crytes

Well-Known Member

Joined:
May 16, 2014
Messages:
140
12
Location:
Clarksville, TN / USA
look at what some of the rc guys are with andrino stuff you could probably adapt it to make a complete glass cockpit for cheaper than a set of steam gauges.

8. Oct 9, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
First, I have the skill and experience to undertake such a project, and it seems to be much simpler than some things I have done in the past.

Now, if you are going to have a 'glass cockpit', you are going to need some type of processing unit and the input sensors (with signal conditioning) anyway. Then you are going to have to process the analog data into a format suitable for the graphics screen. Either format it for the interface protocol (a display module), or drive the display directly from the firmware.

What I am proposing is to use air-core motors (or stepping motors) as analog gauges in place of the graphics display. The processing firmware would be much simpler, I think. Essentially requiring only scaling and/or translating. Yes, maybe a bit of algebra, but nothing too difficult.

The mechanical part also does not seem too difficult. After building my own airplane!

I haven't looked into it, but after buying used instruments and having them rebuilt, are you going to save anything over buying non-TSO instruments form Aircraft Spruce?

Another concern is that my panel space is quite limited, so I don't think I can use 3.25 inch gauges, which is all I have seen on eBay.

As for the time required---that is free. This is a hobby, so the only expense is the parts and materials required.

9. Oct 9, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
10. Oct 10, 2014

Jan Olieslagers

Well-Known Member

Joined:
May 28, 2014
Messages:
134
19
Location:
Haacht, Belgium
No simpler way to implement your idea than using servos as used in R/C models - you could glue the dial directly onto the servo wheel! Or add some fancy gears for a two-dial altimeter.
Easy to interface from a microcontroller, too. I think your idea would be easiest to implement by creating a standard module containing a microcontroller, a servo, a dial. Per instrument you would then only have to devise input circuitry and the display scala.

rdj likes this.
11. Oct 10, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
The problem with RC type servos is that they can not rotate 360 degrees. That would be a problem for something like an altimeter or digital gyrocompass.

And why build 7 micro boards when a single off-the-shelf (Arduino) could handle all the functions. Each instrument would have an air core motor (eBay link) and voltage scaling perhaps. Of course each instrument would require a pair of quadrature signals---really no problem at all.

12. Oct 13, 2014

Jan Olieslagers

Well-Known Member

Joined:
May 28, 2014
Messages:
134
19
Location:
Haacht, Belgium
The lack of 360 degrees rotation is easily countered with some gears, as already stated. They get neither speed nor load to bear so can be very basic. If your own words are to be believed, you have achieved success with projects far more complicated.

As for using the Arduino: one of them will be more expensive than 7 dedicated boards, if home-built. And I for one consider the Arduino a toy, though a very nice toy with good educational value - still a toy it is and remains. YMMV.

13. Oct 13, 2014

nerobro

Well-Known Member

Joined:
Nov 6, 2011
Messages:
1,109
312
Location:
Northern Illinois
Arduinos are available for \$10-20. *shrugs* For one off stuff? They're cheap enough. what's important, is that it's a cheap, easily available breakout board for the AVR series microcontrollers. This has to be a prototype at some point...

I've been looking at solutions to this for a while. Gears easily give you full sweep with a servo, and allow concentric drives. From a hobbiest perspective, the built in POT is nice, because it eliminates the need for quadtrature encoding, or however else you'd want to encode the pointer position.

That said, the guages in your car, are steppers.

I bought some air pressure guages.. I really need to hook those up and see if I can get reliable altitude and airspeed indications.

Along similar lines, belites guage line, are built on the TI MSP430 Microcontrollers. So.. this isn't breaking new ground. Just his displays are entirely LCD or more commonly LED, instead of having needles. the LED's i'm pretty good with. I don't like numbers for "I need info NOW" guages.

14. Oct 13, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
I very much disagree.

It has several orders of magnitude more power than the onboard computers used for the Apollo program.

I would classify it as a quite capable general purpose development system. Just because some make toys out of it does not mean it can't be used for more nobile purposes.

clanon likes this.
15. Oct 13, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
I meant far more complicated from the software and electronics viewpoint.

16. Oct 13, 2014

nerobro

Well-Known Member

Joined:
Nov 6, 2011
Messages:
1,109
312
Location:
Northern Illinois
The IDE makes a lot of old timers twitchy. The chip is comercial goods all the way. But the bootloader... We could talk propellor and Parllax's stuff too. That'll make more old timers twitch too.

My only trouble with AVR, is that Amtel isn't making money. If you want to take a product to market, I'd rather look at PIC who is in the black, or TI who's been at it for decades.

17. Oct 14, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
Interesting.

I used the word "arduino", as an example of a microboard that had all the required capabilities, just once in my posts and everyone has latched on to it.

My purpose in starting this thread was to discuss the gauges/readouts. The air-core motors that I mentioned in my OP are my primary concern. What drives them is secondary. It is just that, with a quick look at the available options, the Arduino was the one that stood out as the most capable.

The air-core motors have no need for a potentiometer for position feedback, as in a RC servo. Since they require a pair of quadrature signals, what ever is the phase angle of the quadrature signals defines the motor position---there is no other possibility.

As far as gearing for a multi-pointer gauge, the only place that would be needed would be the altimeter. Even there it is not essential. AS lists several single pointer altimeters that are limited to 10,000 feet. That would be enough for me anyway.

But if you insist on 2 pointers on your altimeter (I admit it would be nice even if limited to 10,000 feet), I would put the smaller (higher increment) pointer directly on the shaft so it would turn only once. The longer pointer would be geared off of it. )Just like the hands on a clock. The hour hand goes around once while the minute hand goes around 12 times.) In this way, there would still be no need for a potentiometer.

18. Oct 16, 2014

Blue Chips

Well-Known Member

Joined:
Oct 15, 2014
Messages:
370
76
Location:
Using a mini microcontroller for engine/flight data display on a LCD screen has been on my mind here lately as well. It's being done but at high cost, still it does seem that these controllers are more then up to the job of doing it and doing it at relatively low cost, today, it makes a lot of sense to go this route, if..... if you have the know how. Unfortunately, my experience in electronics is in purchasing a BeagleBoard Black for this very purpose as time permits. The plan was to pick one element of engine data to display so as to not make the task so overwhelming and then build upon what was learned from that to add additional engine/flight data. The more I think on the use of the small Micro-stepper motors Aerowerks mentioned to display the data the more I like it over the glass display. I've seen examples of just that on Youtube already using the Arduino, operation appeared very smooth and precise.

Ken

19. Oct 16, 2014

Aerowerx

Well-Known Member

Joined:
Dec 1, 2011
Messages:
4,908
1,320
Location:
Marion, Ohio
The air-core motors I linked to in previous posts aren't stepper motors. They are closer to a synchro/resolver design, except those typically have 3 phases while these air-core motors have only 2 phases with a permanent magnet rotor. They are capable of infinite resolution while steppers go in some discrete step size, such as 0.20 degrees.

I have plenty of experience in both the programming aspects and the electronics. Less experience with the mechanical parts. That is why I like to minimize the mechanics as much as possible.

The last couple of days I have been trying to figure out how to come up with a 10:1 (or 100:10:1) gear reduction on a concentric shaft for a two-pointer (or 3 pointer) altimeter.

20. Oct 16, 2014

Joined:
Nov 6, 2011
Messages:
1,109