DIY Throttle Quadrant Help

[cakequest 4]

9 hours ago, flyforever said:

the pot must rotate 290 degrees in order to provide the 0-100% electrical input. The lever usually moves 80 degrees. The lever is then equipped with either a pulley or gear so that the ratio is 3 to 1 or so. There's no other way to do this, unless you can find pots that rotate only 90 degrees, like the ones found in a Saitek quadrant.

Correct me if I'm wrong, but I'm 99.99% sure that when programming it, using FSUIPC, you can tell the sim where the 0 and 100 positions are so the pot doesn't need to rotate the full 290 degrees...

[Chock 17,769]

19 minutes ago, cakequest said:

Correct me if I'm wrong, but I'm 99.99% sure that when programming it, using FSUIPC, you can tell the sim where the 0 and 100 positions are so the pot doesn't need to rotate the full 290 degrees...

True that certainly is the case. but it is also true that if you do gear down a larger rotational movement, you will end up with more precision owing to the increased leverage of more movement of the potentiometer matching a smaller movement of the control you operate. Whether you want to do that depends on how precise you want your controller to be.

[oemlegoem 29]

11 hours ago, w6kd said:

If you're going to the trouble of building your own quadrant, don't use pots...use Hall sensors.  Pots--all of them--wear out over time because they involve a wiper sliding over a resistive surface.  A Hall sensor measures the strength of a magnetic field from a nearby magnet...they only wear out when the bearings in them do, typically after tens of millions of cycles.  They also generally provide more accurate, precise and repeatable measurements.  I use both Hall Effect IC packages (typically about $2 each) with a rare earth magnet , and sometimes a Honeywell HE rotary transducer (about $45) to replace rotary pots.  The HE replacements I did on my 15-year-old PFC yoke and throttle quad have not only pretty much indefinitely extended their lives, but they are far smoother and low-maintenance--in the 6 years since I replaced the electronics I have yet to have to touch one of the HE sensors, as opposed to yearly cleaning and/or replacements of the pots that came installed in them.

The Leo Bodnar boards handle 3-wire Hall modules just as easily as pots.

Regards

 

Hallo w6kd

Could you explain a bit more about the "Hall Effect IC packages (typically about $2 each) with a rare earth magnet"? Are they separate components put together in some way?

Thanks for the help

[flyforever 267]

3 hours ago, oemlegoem said:

Could you explain a bit more about the "Hall Effect IC packages (typically about $2 each) with a rare earth magnet"? Are they separate components put together in some way?

That chip is indeed a few dollars. But you need a pcb board, some electronics, and a magnet to build the complete assembly. The chip alone wont do it.  They sell already made assemblies, I believe, for appx 10. One then needs to play with the magnet location and distance. It's still a diy and requires patience. If you're building your own quadrant, you can design the pots arrangement so that they are easy to replace when they fail. This, at least for now, keeps things simple and cheap.

[awreaper 14]

Take a look at this guys website.  He has lots of stuff that plugs in to his boards he sells and those in turn are USB and can be calibrated using Windows calibration.

 

http://www.desktopaviator.com/

 

[Lionel Mandrake 7]

Hello flyforever, you said that 'building a multilever quadrant that is functional is a bit more difficult than simply putting a few parts together...In theory, it can be done. In practice, after all the money and time has been spent, the Saitek looks like a bargain.'

As far as I know the Saitek quadrant does not reproduce the mechanical interlocking found in the majority of throttle gear for turboprops.

For example, in King Airs the levers must be lifted over stops to ground idle, feather and reverse positions.

Using controls like the Saitek without this positive interlocking, one is expected to move the levers through positions at which their action will start to be interpreted differently by the software.

I believe that this is insufficiently realistic to develop any useful skill as during landing the quadrant levers must be operated rapidly and positively, relying on their stops to prevent error, and there is no opportunity to look closely at their markings.

I once found a commercial replica of the King Air quadrant, complete with these essential liftover stops (though perhaps not with realistic clanking noises), but it was about a thousand dollars US and by the time it got to me it would have been at least a thousand English pounds, the price here of a reasonably good secondhand car.

Therefore despite the difficulty I still think my only chance is to persevere with trying to design something which can be made with modest resources.

[cakequest 4]

9 hours ago, Chock said:

True that certainly is the case. but it is also true that if you do gear down a larger rotational movement, you will end up with more precision owing to the increased leverage of more movement of the potentiometer matching a smaller movement of the control you operate. Whether you want to do that depends on how precise you want your controller to be.

So what would you recommend doing?

[Chock 17,769]

1 minute ago, cakequest said:

So what would you recommend doing?

If it was a joystick you were making, I'd say gear it for the increased precision, but since it's a throttle, where a lot of the time you'll be just setting it and forgetting it, I'd be inclined to go with the drawing I did and have it directly linked and calibrate it in the sim, since it is easier and cheaper.

[Bob Scott 7,720]

3 hours ago, flyforever said:

That chip is indeed a few dollars. But you need a pcb board, some electronics, and a magnet to build the complete assembly. The chip alone wont do it.  They sell already made assemblies, I believe, for appx 10. One then needs to play with the magnet location and distance. It's still a diy and requires patience. If you're building your own quadrant, you can design the pots arrangement so that they are easy to replace when they fail. This, at least for now, keeps things simple and cheap.

You need a magnet (I use 1/2" square neodymium N42 magnets in several of my homemade transducers), the HE sensor chip (I use the Allegro A1302), and an analog-to-digital interface board that takes the sensor's output voltage and converts it to a USB digital signal--I use the popular Leo Bodnar BU0836X USB boards.  You have to rig the sensor so that it is in close proximity to the magnet, which is attached to a part of the mechanism that turns--the HE sensor varies its output voltage as the magnet is turned.  There are some YouTube videos out there that show how it works.  Ready-made rotary HE transducers like the Honeywells that I used to replace the roll/pitch axis transducers in my PFC yoke ran about $45 each, and were direct replacements for the stock linear pots with 1/4" shafts.

It is a nontrivial DIY project, but then again so is any project using a pot.  But if you're going to go to all that trouble, why design-in future reliability and maintenance problems?  Most pots are not intended for continuous movement--they're typically used as controls on a device operated with a knob, intended for occasional changes, not hours of non-stop back-and-forth movement that wears out the wiper and/or resistive surface it moves across.  Some of the el-cheapo quadrants like the Saitek stuff actually use trim pots that are designed to be set once in a blue moon during device calibration--as a result they wear out quickly and cause the infamous spiking and reliability problem that Saitek is reviled for.

Regards

 

[cakequest 4]

So one last question (for now) ;)

Should I be selecting a higher resistance potentiometer for a more accurate measurement?

[Lionel Mandrake 7]

If you use a Bodnar interface they recommend using any value between 1K and 100K, nominally 10K. The resistance doesn't alter the accuracy. The instructions for using their boards are on their website as PDFs (e.g. http://www.leobodnar.com/products/BU0836A/BU0836A.pdf).

[flyforever 267]

choosing the right pot resistance is a matter of "noise". Too high resistance( 100k)= high noise( spiky behavior in the sim) ; low resistance= low noise, high current draw. A good compromise is 10k.

[himmelhorse 1,276]

hello guys,

I feel a trifle foolish here because I recommend a Honeytech quadrant (not yet released) in this thread page 1) I feel foolish because I cannot find it now and I feel obliged to apologise to the OP

The yoke was featured on the front page here about a month or less ago. It features a slightly rounded top and a red facing place.  The name is obviously not Honeytech because google finds something completely different. 

I am wondering if any one can either remember the name of this yoke and if not, can someone tell me if is there any way to go back to previous pages of the front page where it is featured. I am pretty sure it is Honey something and it is/will be a yoke throttle combination. The quadrant features an annunciation panel as well.

Thanks so much

Tony

[BuddyDog 36]

35 minutes ago, himmelhorse said:

hello guys,

I feel a trifle foolish here because I recommend a Honeytech quadrant (not yet released) in this thread page 1) I feel foolish because I cannot find it now and I feel obliged to apologise to the OP

The yoke was featured on the front page here about a month or less ago. It features a slightly rounded top and a red facing place.  The name is obviously not Honeytech because google finds something completely different. 

I am wondering if any one can either remember the name of this yoke and if not, can someone tell me if is there any way to go back to previous pages of the front page where it is featured. I am pretty sure it is Honey something and it is/will be a yoke throttle combination. The quadrant features an annunciation panel as well.

Thanks so much

Tony

Honyecomb Products?

 

[cakequest 4]

4 hours ago, himmelhorse said:

hello guys,

I feel a trifle foolish here because I recommend a Honeytech quadrant (not yet released) in this thread page 1) I feel foolish because I cannot find it now and I feel obliged to apologise to the OP

The yoke was featured on the front page here about a month or less ago. It features a slightly rounded top and a red facing place.  The name is obviously not Honeytech because google finds something completely different. 

I am wondering if any one can either remember the name of this yoke and if not, can someone tell me if is there any way to go back to previous pages of the front page where it is featured. I am pretty sure it is Honey something and it is/will be a yoke throttle combination. The quadrant features an annunciation panel as well.

Thanks so much

Tony

http://flyhoneycomb.com/alpha-flight-control/

http://flyhoneycomb.com/bravo-throttle-quadrant-2/

These what you were thinking of?