Servo "Whiskey Compass"...will this work?

[JBaymore]

So one of the things I really want to have for some added realism is the Whiskey Compass on the windscreen center post turning as the plane comes around.

So I had this idea for creating one.

[TacitBlue]

I honestly dont know, but Im new at this stuff. I would say, get the magnet and compass and try it out w/ out the servo before buying the expensive stuff.

[SilverFox441]

Sadly, no it won't work.

Modification to a servo to make it turn 360 degrees removes the feedback ability...the servo no longer knows where it is pointed.

What might work is to add a rotary encoder to the servo output. The encoder could pass the position back to the Phidget so that the system would know where the output was. I don't think that the programming exists in FS2Phidgets to use that feedback, but the guy who wrote it seems cool and might add the capability if asked nice.

I do like the idea of magnetically coupling the pices together...vary the distance slightly and you can get the typical "Whiskey Compass Hunt" and is lags behind and the oscillates and overswings a little bit.

For my own Whiskey compass I was thinking of gearing a standard servo output by 1:4...the output is then 360, plus a little bit of available over travel. The only hangup is that when rotating past end-of-travel points the servo would have to swing all the other way around.

I can live with that limitation...but if the software can be tweaked to allow your idea, well I'll go another way.

[SAto]

So one of the things I really want to have for some added realism is the Whiskey Compass on the windscreen center post turning as the plane comes around.

Key to this idea is modifying the HS-322 for 360 degree rotation.  This servo is only about $10.00  ;).

I think this sould work, I haven't worked with servos at all though I'm using step motors (because I get them for free ).
You should be able to tell how far the servo has travelled based on how many pulses you've given it. or however they work. Then you need to provide a reference for it to know where it starts up. You could use an optical interuptor and spin the motor until the interruptor is blocked. Then you have a starting point and can start to rotate the motor to the desired position.

As I stated I've never worked with Servos and don't really know how they work and I've never used phidgets either.
But I think Simkits altimeter works this way, maybe their compass does as well. Could be an idea to check out their assembly manuals, they are found online at http://www.simkits.com/brochures.php


This is the way I'm going to do it (using steppers though)

-SAto

[SilverFox441]

You're magnetic coupling concept gave me an idea...one which needs no servo at all...and no stepper motors.

What we need to do is create a large "air core" magnetic instrument. An air core instrument movement is capable of rotating the magnetic field around 360 degrees...and would over-ride the Earth's magnetic field at close range. The compass would think it was reading correctly...but we fool it. Feedback isn't needed...there's no way in the physical universe for the reading to get wrong in the first place once it's aligned!

I have to dig back into my books for EM field creation...but this is doable and would be very easy to have programmed into a FS2Phidget...only you would need the output from three channels of a digital output board or a dedicated controller...I'll have to see hwat I can figure out.

[SAto]

What we need to do is create a large "air core" magnetic instrument. An air core instrument movement is capable of rotating the magnetic field around 360 degrees...and would over-ride the Earth's magnetic field at close range. The compass would think it was reading correctly...but we fool it. Feedback isn't needed...there's no way in the physical universe for the reading to get wrong in the first place once it's aligned!

That's a good idea!
Don't know how strong the magnetic field is but as previously stated a little lag would actually increase realism.

My problem with this is generating the sine cosine signals to the instrument. Know there are some specialty circuits to do this but they are not easy to come by.
Might do it with a PIC though...

[SilverFox441]

I had an idea that is useable in more places...and doesn't require handling the air-core driver circuits.

It also has a better chance of being implemented into FS2Phidgets...the basic function is useful for more than just FS. Thanks to SAto for mentioning optical interrupters...I knew there was an easier way to do what was required that I was forgetting.

Same exact setup as the original...except we add an optical encoder disk, an optical encoder and the encoder input to the mix. Using this system the system would basically use "dead-reckoning" based on known rotation rate and duration of rotation. The encoder disk/interrupter will allow the system to self-calibrate and to re-calibrate continuously as it rotates.



What the Encoder would like like. This has only 4 encoded positions...there is no reason it couldn't have even more:

[SAto]

Same exact setup as the original...except we add an optical encoder disk, an optical encoder and the encoder input to the mix. Using this system the system would

If you've got access to discarded printers many of them have such encoder disks in them (much greater resolution though) combined with DC motors... suppose it's cheaper than step motors.

Maybe you could use them, even the circuitry in the printer to read this could be used...

As I'm working on a  <50$ budget for the entire pit I need to scavange everything, buying the phidget is not an option for me. But luckily I have access to a lot of scrapped
but working computer equipment.

[TacitBlue]

$50!? I spent that much building my rudder pedals!

John, I have the solution! turn this into a full motion pit, then you wont have to do anything to the compass.

[SilverFox441]

Tacit...that doesn't solve the problem.

It just transfers it from the Whiskey Compass to the Cockpit Heading Sensor...the exact same problem will appear there in your plan.

[TacitBlue]

LOL, I know.

[SAto]

$50!? I spent that much building my rudder pedals!

Yeah I know... hopefully my financial situation will pick up...
But I'm pretty lucky getting a lot of needed materials for free.
Old Computers, printers, discarded metal shelves etc. Provide almost everything I need. Oh and free samples from semiconductor companies

[TacitBlue]

Not a thing wrong with free stuff. Ive gotten all of my hard drives for free.

[JBaymore]

So.... thinking about this issue for the "Whiskey Compass"........

If a servo can be set up with gears to rotate for a full 360 degrees even though it cannot fully go that far ....... not full unlimited rotation...... maybe the compass thing is still doable with only a minor "hiccup".

As the aircraft turns, using the magnetic coupling technique between the servo's position (holding the bar magnet) and the actual compass indicator, when the servo reaches the end of its travel to say 360 degrees...... it suddenly and QUICKLY will then go back to see the position for a 1 degree heading.

It is POSSIBLE that the real magnetic compass will not actually "follow" the fast rotation of the bar magnet back, and the magnet will get to the 1 degree position fast enough that there will just be a small heading "bobble" for an instant whenever the plane crosses the 360 degree / 1 degree heading point.

Thoughts on this?

best,

.......................john

[SilverFox441]

I think the "bobble" will be fairly noticeable...but with a fast servo it might not be too bad.

A Hitec 322 servo has a 60 deg travel time of .19 sec a 5V. Geared for 360 deg would load it up a bit...so make it .25/60 deg. You would need to gear it 5/1 (approx 80 degree servo travel would give you 400 deg total travel). Make the end rotation .05 sec/60 deg.

You might just be able to "lose" the compass for a moment, catching it as you come back around .3 seconds later. The real trick will be careful spacing of the magnet to override the compass... but allow it to get lost when wanted.

If that doesn't work there is a way to modify it so that it does:

Replace the bar magnet with a low powered electromagnet (hand wind one...it's easy). Power this magnet at all time except when crossing the 360 deg point. At 360 release the electro and apply power to a second magnet, a "holder". The holder hangs on to the compass until the servo swings back around and then releases it to the original magnet.

This is much more complex...and quite probably unnecessary. You should be able to build it the way you last described.

You can find gears here: http://www.servocity.com/html/gears___sprockets.html