Arduino driven VOR

[hyamesto]

Well, i must admit - if it is a galvanometer, all that make sense.
Thank you for the good explanation provided. So - no more sound cards driving . Arduino plus some current limiting resistors
should work nicely by themselves.

PS:
I am a little sceptic about implementation of an absolute encoder - most of them don't have 360 deg. /resolution . They are more or less like a codding switch with 8 wires and that would result in 8* 8 = 64 deg. For 360 deg. would need 19 wires on the back ( i suppose ) and some port to read them all. And that makes such a fancy encoder really expensive . Analogue resolvers are expensive too. I don't know - i am not that good with electronics.

PSS: 5 * Original price, wow , that is a lot. Why not coming to live and work in Europe - not much talented people here, mostly ingenious bankers and euro-bureaucrats. And some 20% on top of all - VAT - to gather money and feed all those ( doing nothing ) people. But not much engineers and talented people, willing to do business and produce something useful.

Zlatko: I just saw your post modification....

First:
The absolute encoder to which I referred and you can see properties on the link:

http://www.usdigital.com/products/en...tary/shaft/ma3

Is a magnetic absolute encoder, with 10 bits analog output resolution, that´s 1024 values in 360º, or 0,35º resolution, and use only 3 pins.
Is like having a potenciometer without detents with endless 360º rotation.
Yes, at u$s 60.- its no cheap, but not expensive like analog resolvers.
If you want more resolution, this absolute encoder can be orderer with 12 bits PWM output resolution (4096 values in 360º or 0,087º resolution !!!).

And about "Why not coming to live and work in Europa?"
Yes, right now, wait for me at the airport....I arrive in twelve hours ...
Seriously, it´s too late for me. I am old, have family, properties,... you know.......

And for you and Simsupervisor (why not use your first real name?)...

You can find Vu meters for audio, sometimes cheaper than voltimeters or ammeters (at least in my country are half price).
If the voltimeter, ammeter or VU meter are big or small, just disassembling and use the air-core, and cut the large needle if it is bigger, or suplement with paper or other light material if it is short...

I cannot upload the video tonight, sorry, but i want to make a very good description of the process for calibrate the flag TO-OFF-FROM and the CDI needle. And i want to try another way, for better resolution..
Be patient.
Regards.
Horacio. (Real Name......)

[SimSupervisor]

You can call me Yan.

My main issue is with the glass protection, that's why I say I will use bigger ones. Just haven't found a suitable cheap ones yet.
Vu-meters are harder to find here. But there is no rush to find these. Still lots to do on other instruments.

I can't wait to see what you have in mind to increase resolution. No rush however, take your time. I am just real curious.

[Zlatko]

Couldn't get my hands on such boutique and rare parts as encoder in question, in Europe either. Unless i am not buying 1000+ parts from the local distributor. When i am looking at all those shiny producer's web sites , i feel what the meaning of the word frustration is . To make long story short - if it is not listed on Ebay and Alibaba, etc, it simply doesn't exist.

[xplanematt]

I just skimmed through this thread and thought I would offer a few thoughts....

First, kudos to SimSupervisor for figuring out how to read the OBS scale directly. That's the kind of avionics reverse-engineering that makes this stuff fun. As you probably know, the outputs you are reading off the OBS scale actually drive a phase shifter. The output of the phase shifter drives the CDI needle (which is just a big D'Arsonval meter...aka a volt meter). The signal coming into the OBS gauge is the decoded VOR composite signal that is output from the NAV receiver (which demodulates this composite signal from the 108-117.95mhz carrier signal that the VOR ground station puts out).

I had some help from a ham radio guy who had written software to generate the VOR composite signal and output it to the soundcard. I made some slight modifications to make his software compile on my system, and used it to generate WAV files for each radial from 0 to 359. Then I wrote a little interface software to play the appropriate WAV based on X-Plane's VOR status. Been using this solution for over two years now I think, works a treat. It also means I had to do zero reverse-engineering on the gauge itself. I did learn a lot about the VOR signal (still can't say I fully grasp how to generate it). There is an offset between two signals in the composite signal....the phase shifter you control with the OBS scale shifts the offset signal back into phase with the reference signal when the OBS is tuned to the radial the aircraft is currently on, neutralizing the phase shifter's output (and thus centering the CDI needle). Deviation between the signals generates a voltage on the CDI needle, causing it to deflect from center. Simple, slick.

From what I am reading here, it sounds like you guys are saying the CDI needle requires more than +/- 5vDC to get full deflection, no? Just use an H-bridge to control the needle. You can reverse polarity easily, and you can control the enable pins on the H-bridge chip with PWM to get your "variable" input voltage. I use the L298N a lot. The board I'm using officially supports 12vDC, although I've run 28vDC in it as well (it just makes the 5vDC voltage regulator on this particular board unusable, which doesn't matter if you are powering your 5vDC stuff from somewhere else).

If you are needing absolute rotary encoders with continuous rotation, look no further than the AS5040/AS5045 chips:

http://www.madscientisthut.com/Shopp...0Encoder&xm=on

These are small chips with embedded Hall effect sensors and a bunch of neat logic and noise filtering, as well as several different types of output. I am using the AS5040 breakout board to read the position of a syncro on an altimeter, and it is FLAWLESS. Takes a little more work than a simple potentiometer or bare Hall sensor, for instance, but I get amazing precision, and the altimeter has been absolutely perfect for the year or two since I interfaced it. And, it's $20USD, not $60. I assume they would ship internationally.

With all this being said, there is some variation in how OBS gauges are designed. It is my impression that some systems do ALL the demodulating in the NAV receiver, and the OBS gauge in this instance gets some other type of signal. Additionally, I have two OBS gauges that I think may have been off a commercial simulator, and they use syncros for the OBS scale. I could easily read them with the AS5040 boards, but I am very close to being able to fully and precisely read syncro outputs with an Arduino, which would be a cleaner solution....will probably go that route with these particular gauges.

Matt

[SimSupervisor]

In fact its pretty much the opposite. The D'arsonval needle indicator reacts in less than one volt side-to-side I'd say (without measurments)

I am curious to know how you connected the NAV unit - power and sound card. I never could manage to get mine to work (although, I need to mention it was sold as non-operational...)
That being said, I can confirm your assumption that not all OBS reader are identical. I had 3 old VOR indicator - all Narco. one being the CS-3A and I assume it gets its signal from an additionnal box, as there's no electronics inside, only links to the various D'arsonval and OBS pins. The two others were VOA-4 units - and they use CS-3A internally, but have electronics in there.
I never managed to power any up. One was missing its transistors, and the other seemed OK; it was one of the best-looking instrument I got from the supplier I bought my indicators from (some were rather destroyed by corrosion)

[hyamesto]

Ok: we have a full resolution needle with this connection:



Sketch to test: (Complete sweep of the needle, from left to right maximum scale)

1 int n=1;
2 int out=127;
3 //------------------
4 void setup()
5 {
6 pinMode(6, OUTPUT);
7 }
8 //---------------------
9 void loop() {
10 out+=n; if (out==255 || out==0) n=-n;
11 analogWrite (6, out);
12 analogWrite (13, 0);
13 delay (1);
14 }



How to setup:
1.- Put limiter trimpot to max resistance. Put centering Trimpot at medium.
Comment line 10 of sketch with //
Set in out=127 in line 2
Upload the sketch.

2.- Now, Arduino send value of 127 from PWM output, and now set centering trimpot, to obtain 0 desviation in the needle (center).

3.- Set in out=255 or 0, in line 2, left commented line 10 of sketch.
Upload the sketch again.
Now, Arduino send value of 0 or 255 from PWM output, and now the limiter trimpot to obtain full deflection of needle (left or right, depends of value of int out).
4.- Carefully setting the potentiometer until the needle just touches the physical limiter, no more.)
5.- Send again value of 127 from PWM output, and retouch centering if needed.
6.- Upload the original sketch, with in out=127. (You can change the delay value from 1 to 1000 to see fast response (delay=1) or the "steps" (delay=1000).
7.- If you reset your Arduino, the needle goes to centering position.

Now you have a full resolution needle (256 steps).
It is the maximum resolution provided by FSUIPC Offset 0C48 (NAV1 Localizer needle -127 left to +127 right).

Connect Arduino with Simulator interface software, make your own sketch, some tweaks and enjoy....

FAQ:
Why 22 Kohms more than earlier test?

To reduce the current inside the D'arsonval coil (you need only 150 microamper).
With only 10k trimpot, i see the power led indicator of the arduino changing intensity, signal of to much current flowing inside the coil.
And i have a better resolution (256 steps vs. 194 steps earlier setup)

Why the needle don´t go beyond 3,5 points in scale on each side?

Because the physical stops of the instrument are bad calibrated.. (work to do......)



A video of test:

https://youtu.be/qbpv9LOR-uo

Next test, TO-OFF-FROM Flag with only 2 digital outputs (NOT PWM)
I just need to open and regulate the physical stops, because the Flag locked in TO position.
OFF and FROM working ok...

Now i know why the instrument was descarted.......

Regards.
Horacio.

[hyamesto]

Matt:

I am interested in your soft. Can you post in some free host (Mediafire, Mega) ,or send via mail?
I was trying to do something similar with a sin-cos signals out of phase by 90º, very large to explain here...
but you can see something in this post:

http://www.mycockpit.org/forums/showthread.php?t=25328

With Arduino now, seems to be and easy solution to generate the signals and injecting on the instruments.

As you can see, with only 1 PWM output, i can drive the CDI needle (no need of h-bridge, or more outputs).
The D'ansorval needle indicator reacts with only 150 mv (yes, milivolts) and need only 150µa (microamper)

I don´t think there is a easier or more simple way solution like this.

And for the FROM-OFF-TO flag, only 2 digital outputs (not PWM).

About the VOR internals, the new instruments have a resolver to set the OBS.
Newest maybe use and magnetic absolute encoder (I heard of this from Honeywell, owner of Bendix/King and manufacturer of electronical components)
(And in a Glass Cockpit who knows?)
The older instruments, like Slimsupervisor and I have, has a sort of Wheatstone Bridge (as said, " a sort").....
This is a picture of the internals, and squematics of connection:


Now, they are descarted because the wiper is mechanics (photo 2), and subject to wear, Flags replaced with led lights, etc.
Mine have broken wire coil.....

The AS5040 boards are amazing, but i can´t buy from my country right now (import laws, dollar restrictions, etc).
I was thinking of using the HMC5883L magnetic sensor with a magnet, available in my country, but i read it is very delicate with strong magnetic fields.
Maybe others in the rest of the world will appreciate this information (Simsupervisor, Zlatko .....)

And for all, thanks for collaborating.. As Mahatma Gandhi says: "Satisfaction lies in the effort, not in the attainment"......

Regards.
Horacio.

[hyamesto]

Couldn't get my hands on such boutique and rare parts as encoder in question, in Europe either. Unless i am not buying 1000+ parts from the local distributor. When i am looking at all those shiny producer's web sites , i feel what the meaning of the word frustration is . To make long story short - if it is not listed on Ebay and Alibaba, etc, it simply doesn't exist.

Zlatko:

Frustation is what happened to me friday in the three local electronics shop.....

"Hello Sr, do you have 10K Potenciometers?"
----NO
"Trimpots?"
----NO
" 20 Mhz crystal?"
----NO
"a 0,5 amper short type fuse?"
----NO


The same in every shop.

As you can see, if you can not get a simple fuse to replace, forget something like a simple chip.......

Regards.
Horacio.

PS.: Don´t wait me at the airport, flight delayed without explanation...

[SimSupervisor]

Yeah, electronic shops are not what they used to. That is why I generally buy online. When I find locally I tend to buy moire then I need, just in case.
Horacio, what values resistor do you get for full maximum deflection? I didn't have that many potentiometer and didn't do the whole exercise of tweaking it to the max, 3k worked ok.

I am planing to use the obs ring of the broken unit as the feedback of a servo to scratch build a heading indicator. These are quite useful. ll post up my progress on a new thread as soon as I get an H-bridge

[hyamesto]

As the new graphic says, i use a 22k resistor in serie with a 10k trimpot (to limit current), and a 10k trimpot to centering needle.
With 3k like you setup, worked ok too, but too much current flow inside the D'ansorval, and make the needle least sensitive.



Regards.
Horacio.