Home-brew flight instrument ideas

[Guest MikePowell]

I've been tinkering with ideas for making simulated flight instruments. Most recently I've been focussing on air-core meter movements. These form the basis for many automotive and marine tachometers, speedometers and gauges. Rather than buying the finished units, I have been trying to build them from inexpensive, readily available materials. I've had some modest success and have posted some pictures and information at www.mikesflightdeck.com/aircore_instruments_1.htm. I haven't packaged the movement to look like a flight instrument yet, but that will come. In the meantime, perhaps the info and pix will be of help to some of you pursuing similar goals.Mikewww.mikesflightdeck.com

[Guest Simpit]

Hmmm... the URL didn't work... but this one http://www.mikesflightdeck.com/aircore_instruments_1.htm did. I think the period at the end messes it up.By the way - those are really neat! What kind of resolution (arc-wise) do you think you can get out of these? And can you get continuous 360 degree motion out of these (for things like altimeters) ?Derek :-yellow1

[Guest MikePowell]

Whoops! Sorry about the URL. Here I thought I was properly dotting my i's and crossing my t's.The basic movement can turn through a continuous 360 degree arc, and keep on going. The limits to resolution seem to be the drive circuitry. The early automotive chips (National LM1819, OnSemi CS289) used piecewise linear approximations to sine and cosine that worked amazingly well. Turns out the actual sine and cosine values are less important than their ratio. If the Tangent comes out okay, you've got it licked. I built a spreadsheet based on the published approximations and found that angular errors were less than a half degree anywhere through a full rotation.The newer chips (Phillips SA5778, etc.) are based on a microcontroller driving onboard DACs. These chips can drive three or four movements, and as a result, the older chips are being obsoleted. I don't know the DAC resolution on the new chips.Right now I'm just driving the prototypes with variable power supplies. I'm not certain how I will ultimately drive them. I may base it on the automotive chips, but perhaps not. The older chips only work through some 300 degrees. The newer microcontroller based ones use an automotive bit serial interface bus, useable, but not where I want to spend my time right now. I have two home-brew approaches blocked out that should handle a full 360 degrees.Right now I'm just starting to look at the major limitations to accuracy. Friction seems to top the list, but rotor weight imbalance, asymmetric field windings, cogging and external magnetic fields contribute errors as well.I think this approach will work well for single pointer type indicators. I had initially thought to use them for concentric pointer style instruments as well, but the friction issue makes that a non-starter without really expensive bearings or perhaps position feedback.Thanks for your comments, Derek.Mikewww.mikesflightdeck.com

[Guest Simpit]

I looked through your website... a lot of good information there! Do you have a background in something related?Derek

[Guest MikePowell]

I worked in R&D electronics for a number of years. I also hold a private pilot's license, though I am far from current.Mikewww.mikesflightdeck.com

[LawnDart]

Mike~..about your friction issue on the bearings. YOu might try what's called a 'Dither' method.Dither is nothing more than an A/C signal riding on top of the DC voltage level. The intent is to vibrate the shaft within the bearing to significantly reduce the break-free friction coefficient. THere are tradeoffs, too complex to go into here, but do some homework to see if you can modulate the DC signal enough to overcome the quiescent friction. Hope this helps!Ray Sotkiewiczwww.avsim.com/bluesideup

[Guest MikePowell]

Ray,Thanks for the suggestion. Dithering would certainly seem to be a possibility. You used Express PCBoards, didn't you? Happy with them?Mikewww.mikesflightdeck.com

[LawnDart]

I used them and I am very happy with the quality. They do an outstanding job and sometimes ship earlier than agreed. All holes, cuts, and runs line up perfectly and the holes are plated and tinned through.The silkscreen job is professional, and even the smallest runs (.08?) maintain their integrity.Just be sure you obey all the little rules about designing circuit boards! (ie: for very small runs, never use a 90-degree corner... use 2 45-degree corners when possible)Cheers~Ray

[Guest MikePowell]

Excellent. Thank you.Mikewww.mikesflightdeck.com

[Guest jsippy]

Hi Mike,Instead of building your own air-core movements, why didn't you acquire automotive instruments and modify their faceplates to suit your simming needs?---John R. SippyChicago, IL, USAPowerMacG4 -Dual 5001024MB RAMnVidiaGF2 - 32MB

[Guest MikePowell]

Wow, John, you must have been looking over my shoulder for the past few hours. I was just at Pep Boys, a local auto parts and service chain asking about after market tachs and gauges. And just a few minutes ago I was checking for same on line (www.sunpro.com , www.jcwhitney.com ). And you are absolutely correct. Using an auto tach would be a fine approach to take. Quite workable, and depending on the size of your project (and budget), probably not too outrageous cost-wise.Basically, there are two reasons why I pursue the build-it-from-scratch approach. First, I really, really like building things, especially things that are uncommon or are regarded as difficult. I enjoy the challenge and get much satisfaction for my successes. (I guess the flip side is that my wife really doesn't care to be around me when I'm grumping about my failures.) And second, I'm really, really cheap. The least costly aircore tach I found tonight was about $23 (on sale) plus shipping. I could have bought a similar unit locally for about $35 plus tax. (D'Arsonval based units are a few dollars cheaper, but are less robust and generally most accurate over only about a 90 degree swing.) Really not bad prices if you're planning on only two or three instruments, but it adds up quickly with more. In some areas, junkyards might be an option. Where I'm at, they are few and far between, as well as being expensive (roughly half the new price!!!). Right now I have the materials cost of the homebrewed, basic aircore movement at about three dollars. Even so, a purchased auto unit would likely be less expensive overall if I simply stopped with the basic gauge as I still have electronics and a case to build. But I plan on adding other features like fail flags, and who knows what else.Thank you for your comment, John. You've brought up a good point, that using auto instruments is not a bad approach for the simpit builder. Until tonight, I hadn't realized such instruments were available so inexpensively. I had only been aware of units on the high side of $100.Regards,Mikewww.mikesflightdeck.com

[Guest jsippy]

Thanks for the reply...I can certainly sympathize with the desire to build things that "others" have deemed too difficult to build from scratch.Ive been messing around with various components with the intent to mate a USB-based, Jet-style (Throttles/Spoiler/Reversers/Flap Lever) quadrant with Fly! Not an easy task considering that Fly! will not accept an axis-based input for flaps position, nor does it have a way to interface any more than on/off control of the spoilers. So in addition to tinkering with rotary encoders-to-keypress emulators, I have been trying my hand at developing add-on drivers using the Fly! SDK.I will start work on porting the EPIC over to the Macintosh shortly after finishing the quadrant... Then come gauges; )---John R. SippyChicago, IL, USAPowerMacG4 -Dual 5001024MB RAMnVidiaGF2 - 32MB

[Guest MikePowell]

I look forward to seeing your work. I hope you will post some pictures. Mikewww.mikesflightdeck.com

[Guest jsippy]

I most certainly can post some pics, that is if my "day job" allows me the time to work on the project some time this century; )---John R. SippyChicago, IL, USAPowerMacG4 -Dual 5001024MB RAMnVidiaGF2 - 32MB

[Guest AirPanther]

Hey Mike... E-mail me if you don't mind. I've got some stuff you may be able to use :-)Robert PratherThe 777 [email protected]