Rotary instead of potmeter for axis

[NAX228]

Any of you guys ever looked into using a rotary encoder for yoke and pedal axis instead of regular potmeters?
And can you think of any pro/cons about using rotarys?

[ian@737ng.co.uk]

hello nax............
first thing that springs to mind is how will you calibrate it?      surely it's the calibration which tells the sim where the axis is on it's travel.     what you are suggesting is indeed possible by writing values (in fsuipc) for each increment/decrement, but i have this nagging thought in my mind about accuracy.     at startup, if your encoder is not in the exact same place as the start situation, then positional accuracy has gone out of the window.   and then of course, you can't adjust the response curve with an encoder.
just my thoughts (somebody shoot me if i'm wrong) 
rgds

ian

[NAX228]

Good one Ian. Havent thought of that.
Main reason to try this would be to get the yoke more accurate, but complexity must not kill the project.

[Boeing Skunk Works]

If you want accuracy, then use linear transducers. The accuracy is generally within +/-5% tolerance and is repeatable.

Be prepared to pay for that accuracy.

[fordgt40]

The Master (Ian) is correct as usual. However, there is a way around this if you are using SIOC or similar scripting I/O interface. Put a micro switch at one end of the yokes travel and when you start up turn the yoke to engage the micro switch - this will then register the position of the yoke and the encoder values will inc/dec from this, assuming you have coded  the SIOC script correctly:)

David

[ETomlin]

I think that it's not worth the time to go this route and to use either a high quality pot to begin with or look into the transducers.

[ian@737ng.co.uk]

hello chaps.....
totally get david's point about the microswitch, but remember you need two because you have two
axes to deal with, roll and pitch.    keep it simple, it's much easier and cheaper 
go with high quality components and do your bit with the calibration.   may take you a while, but get the
calibration right and you're 'good to go' 
rgds

ian

[TomR]

Hi

Don't forget that ANY accuracy depends first of the controller and NOT of the connected Pot(s).
If you e.g. use Opencockpit's Analag Inputs of a Mastercard it does only 256 steps, 128 left from centre and 128 right from centre.
A FDS-FC1 does already 512, 256 to each side.
A Bodnar up to 4096 (12 Bit version), 2048 to each side.

So what ever you control in FS is 16383 to each side (Throttle 16383 + Reverse neg. 4096), that means you have to send e.g. the value up to 16383
-  8-Bit Controller (e.g. OC), you send values with steps of 128
- 10 Bit Controller (e.g. FDS-FC1), you send values with steps of 64  (this is already precise enough)
- 12 Bit Controller (e.g. Bodnar), you send values with steps of 8
This is of course if you use the full range of a Pot. If your pot is a rotary you will get this only with a gear, scaled 3:1 or something.
The slider is the easier way.
If you want it really precise you need to go with a e.g. 12 Bit Controller and a gear, together with a hi quality pot (10 to 20 $ each).
But the experience says that is really not needed if you do your mechanical connection precise enough, a "good" Pot and a "good" controller like FDS-FC1 or Bodnar.

[Maurice]

I may be very dense here, but why on earth would you need that much accuracy? All that is really required is that the centering back to neutral is fairly accurate and that has a lot to do with the mechanical assembly and not the pot.

So what if you need another inch or two of movement on the elevator to raise the nose to the same degree as the last time? Would you even be aware that this is happening?

Just my 2 cents but I'd love to know if I am missing something important here.

Thanks,
Maurice

[jackpilot]

Talking more from experience than electronics knowledge I agree with Mike and Mau.
1) my first TQ used high quality pots but just a nearly straight connection to a USB converter, (meaning no controller)
I was never able to sync the power levers and even the idle position was erratic, Approaches were all S shaped!
2) On my recent TQ , I calibrated everything with FSUIPC and used an FDS FC1 controller  with Phidgets slider pots.
All works impeccably, thrust is easy to adjust to very precise values and sync  between the two levers stays within less than 1%. Flaps are calibrated with a slider too and never miss a beat.

My 2¢.

[NAX228]

Idea hereby terminated