Friction Clutch for Lever Ideas?

[shortbus=]

Could you give some more information? How big is the bracket, length, height, width? Material of bracket and thickness. Better pot dimensions, or a link to data sheet.

How fast would you need them?

 

[dknguyen]

Could you give some more information? How big is the bracket, length, height, width? Material of bracket and thickness. Better pot dimensions, or a link to data sheet.

How fast would you need them?

I don't need them fast at all. Time is not an issue.

The main requirements of the bracket are the following (all approximate with leeway):
-no deeper/taller than 5cm to fit within height of enclosure
-width as narrow as possible so levers can be placed side-by-side close enough to be grabbed simultanouesly by the fingers which is about 3cm max. BUt the potentiomter shaft limits everything to be about 2cm wide.
-around 5cm long to accomodate for the ~3 cm slot I want the lever to slide back and forth in

Potentiometer (the bushing mount version):
https://www.electro-tech-online.com/custompdfs/2010/01/157.pdf
What I have right now for the bracket is in the attached PDF.

 

[shortbus=]

I'll sketch you up something that I came up with and post it to see what you think. I can make the bracket for you too. I was a tool&die maker and built a ton of prototype stuff in my work days. Something like this is what I need to get me off my a$$ and get my home machine shop set-up !

I did some thinking on your AOA/Slip stream sensors. Have you considered a sensor like they use in a "Segway"(I think thats what they're called), a inverted pendulum. That would get your AOA measurement. The AOA doesn't have to be measured from a airfoil does it? Just off of level, right?

The slip stream is harder. Do you really need to measure slip stream? I came up with a way for that using another Segway sensor but its a little more complicated, and needs the vane/airfoil.

 

[dknguyen]

I'll sketch you up something that I came up with and post it to see what you think. I can make the bracket for you too. I was a tool&die maker and built a ton of prototype stuff in my work days. Something like this is what I need to get me off my a$$ and get my home machine shop set-up !

I did some thinking on your AOA/Slip stream sensors. Have you considered a sensor like they use in a "Segway"(I think thats what they're called), a inverted pendulum. That would get your AOA measurement. The AOA doesn't have to be measured from a airfoil does it? Just off of level, right?

The slip stream is harder. Do you really need to measure slip stream? I came up with a way for that using another Segway sensor but its a little more complicated, and needs the vane/airfoil.

That'd be great if it could get it out more cheaply than eMachineShop can (or lower amounts so I spend less money in total. That's one of the reasons I'm getting 4 to use and 4 extra rather than just 4).

The definition of AOA is angle of the slipstream strikes the airfoil relative to the chord line of the airfoil, so yes it has to be measured relative to the air.

Measuring relative to the level is relatively easy and is part of attitude for yaw/pitch/roll measurement for the aircraft -you got accelerometers+gyro sensor fusion or mechanical equivelants like the inverted pendelum and for aircraft you also have light/UV/IR based horizon sensors.

The AOA measurement is not for attitude detection. It's to measure flight characteristics for things like optimal cruise, climb, glide, and descent. It's also used to tell how close you are to stall and how much more lift you can gain from the aircraft (when combined with airspeed measurements) Hopefully the attached diagram explains how AOA is different than the pitch of the airplane.

EDIT: Errr...it seems I may need to get you to modify shaft collars or make them too if you are also willing. I had preferred for them to be stainless steel (the bracket too actually) but that might not work out so well. Oh well, plenty of other materials.

 

[KMoffett]

Even more daunting is how to cheaply construct an vane AOA/sideslip sensor like this:
https://www.electro-tech-online.com/custompdfs/2010/01/ABprobe-1.pdf
how on earth would you simply, reliably, and accurately mount the shafts of both wind vanes to a center shaft like that?! while also providing bearings for rotation of the wind vanes?! and where would you stick the angular sensors for the vanes?!

I'm thinking maybe an optical quadrature decoder with a SMT µC. Also an optical sensor for a 0° reference point. Very low friction. Easy to implement...well, conceptually anyway.

Ken

 

[shortbus=]

I hope my upload worked. first time I've tried since the site change.

I hope you can figure out what I'm trying to show. Instead of a shaft collar I figured to make a "friction block." It will fit inside the bracket and set screw to the pot shaft. There are two small balls with small springs that will apply pressure to the inside of the bracket. Make sense?

As far as using stainless for the bracket, it's hard to form something by hand with stainless. It's springs back a lot.
Hard to make a narrow U shaped part.

I don't know a lot about planes and less about helicopters. I thought that AOA and slip stream were two different things. How about using one of your pots and mounting a vane on it? Make an extension to put the vane out of the side of the airframe and mount the pot inside?

Is this controller for a plane or heli? Why four throttles? Are they mounted two left and two right?

If you want me to make these for you I'll be sure to ship them by UPS

 

[Mr RB]

You don't need to machine the collars!

Just buy 8 knobs for pots, the ones that have a brass insert and a grub screw. Then unscrew the grub screws and replace with a little bolt of the same thread type.

(edit) Or take Shortbus up on his incredibly generous offer.

 

[dknguyen]

I hope my upload worked. first time I've tried since the site change.

I hope you can figure out what I'm trying to show. Instead of a shaft collar I figured to make a "friction block." It will fit inside the bracket and set screw to the pot shaft. There are two small balls with small springs that will apply pressure to the inside of the bracket. Make sense?

As far as using stainless for the bracket, it's hard to form something by hand with stainless. It's springs back a lot.
Hard to make a narrow U shaped part.

I don't know a lot about planes and less about helicopters. I thought that AOA and slip stream were two different things. How about using one of your pots and mounting a vane on it? Make an extension to put the vane out of the side of the airframe and mount the pot inside?

Is this controller for a plane or heli? Why four throttles? Are they mounted two left and two right?

If you want me to make these for you I'll be sure to ship them by UPS

This is for an electric sailplane so 4 throttles is for worst case- throttle, flaps, spoilers, and airbrakes. For a heli you only need two...RPM and collective. But the controller is really for everything and anything. I have a separate 3-axis joystick I will use for yaw/pitch/roll or rudder/elevator/aileron.

THat's a pretty fancy friction disc lol. Far more than what I had in mind Integrated collarshaft, lever mount and friction disc mhmmm. I have very much been leaning towards a clamping method to secure the lever to the pot shaft because I didn't want to damage the pot shaft (it's stainless steel and completely rounded). I realize that the single-piece clamp collars need to flex which really limits what you can do with the collar (shape, threaded holes elsewhere in the collar). But the two-piece clamping collars seem to overcome those problems and does not place the same limits on the shape (for things like threaded holes or bulkier to accomodate its job as a friction disc).

The AOA thing is something I'm just mulling over how to do. Throttle lever is much more immediate. I was hoping to use the pitot tube itself as a boom mount since it will probably be in the best place already as far as airflow is concerned. Mounting anything on the pitot tube seems to sound like more shaft collars...but definately clamp types! I definately don't want to damage the expensive pitot tube with a set screw

 

[shortbus=]

I can make the part with a two piece clamp. The single set screw will actually hold better. As far as messing up the pot shaft, after it is adjusted the first time how often will it be changed?

 

[dknguyen]

I can make the part with a two piece clamp. The single set screw will actually hold better. As far as messing up the pot shaft, after it is adjusted the first time how often will it be changed?

Hopefully never, but I fiddle around with things a lot and previous motor shafts I have have been absolutely mangled by setscrews. And in this case, whenever the setscrew is removed and goes in again, it will usually be very near, but not always on, where it last was. I'd rather not run into complications dealing with screwing in beside the bite marks from setscrews since the collar determines where the lever is.

I actually managed to find a source for the shaft collars I was looking for (you know, the clamping shaft collar that also has a M4 setscrew hole and has a inch bore of the proper thickness). So, if there's any better friction clutch mechanism that can be independent of the shaft collar, I'm all ears! (Why couldn't it just be in the pot?! lol). One of the things I was concerned about was metal dust being formed from a metal-on-metal friction mechanism which could dirty things up...and at worse get into the nearby PCB. DOes that happen with spring-ball friction devices?

 

[shortbus=]

quote, dknguyen " One of the things I was concerned about was metal dust being formed from a metal-on-metal friction mechanism which could dirty things up...and at worse get into the nearby PCB. DOes that happen with spring-ball friction devices? "

No the ball rolls and makes a "silk like" feel to the movement.

I did some searching on RC sail planes. They look like the balsa and tissue planes my uncle had hanging in his garage rafters when I was a kid! They were free flight planes, the tail was held on by a rubber band, it used a piece of string that was soaked in saltpeter and set on fire at launch. The string would burn through the rubber band and drop the tail off to stop the flight! They had 4 or 5 foot wing spans.

Your AOA sensor,You probably now about this; Charles River Radio Controllers - Active Flying Wing Pitch Stabilizer Concept
I thought it was a real interesting site, you might want to check out.

Glad you found your shaft collars. You could get your friction from using these; McMaster-Carr

1. Find a spring that will slip over the outside dia. of the main body(not the flange)
2. Cut the spring a little longer than the main body (have to play with the length)
3. Slide the bushing over the threads of the pot.
4. Put the collar on the pot shaft.

The flange of the bushing will rub against the collar giving the friction. You adjust the spring length to get the friction you want.

 

[dknguyen]

Ahhh, I realized I forgot a critical step. THe potentiometers I was going to use (and really, every potentiometer I can find) has a electrical rotation of 300 degrees or more. I only needed ~40 degrees rotation for the lever. I had known this, but had never realized how small 40 degrees is compared to 300 degrees. That means that my output voltage range is only 13% of my the full-scale voltage. If the ADC has 10-bit resolution, that's only a 136 steps. ANd I can't seem to find any potentiomters that have smaller electrical rotation angles.

SO unless I can, it looks like this lever won't go ahead and I'll have to find some other method. I'm looking at the 1-axis and 2-axis joysticks right now and they are sigifnicantly less expensive than 3-axis joysticks of the same series (30% of the cost!...some single-axis joysticks are 15% of the cost!). They aren't as thin as I would like though so I wouldn't be able to pull off the 4-levers beside each other. I could do two though, though the width is at the maximum limit so that you can grab two levers at the same time.

 

[KMoffett]

Wonder if you could hack the little pots out of one of these: SparkFun Electronics - Thumb Joystick.

Or hack some pots out of old analog joysticks from a thrift store.

Ken

 

[dknguyen]

Wonder if you could hack the little pots out of one of these: SparkFun Electronics - Thumb Joystick.

Or hack some pots out of old analog joysticks from a thrift store.

Ken

The 2-axis joysticks aren't difficult to find (heck, I could just buy a replacement USB RC transmitter for my RC flight simulator by paying $50 and rip the gimbals out of those!). And the cost of new 2-axis ones is very reasonable compared to the 3-axis ones. The price problem comes with the 3-axis gimbal which are rare and less used. It seems to get that extra axis tends to cost a lot (at least if you want an all-metal mechanism).

You can get economy 2-axis for $50 and and 3-axis gimbals $100 (100% more cost for 50% more axis!). But the problem with gimbals is there are so many fundamental possible options that are critical to it's function (like friction hold vs centering, or NWSE guided feel vs unguided feel and those aren't available on cheaper models.). Like these ones are quite cheap by comparison:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=679-2257-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=679-2259-ND

other models cost about 2x or 3x more. I'm wary of the cheaper though because these do seem to be a product that is overpriced while being underbuilt, particularily the 3-axis potentiometer-based sticks that have lots of moving parts (lots of cheap 2-axis sticks seem to be floating around and used with no problems) and I need some level of reliability here. Hall sensors seem to have allowed 3-axis joysticks to be produced that have almost no moving parts). Maybe 3 parts or so to support the control stick. All the sensing is done by a small magnet on the base end of the stick hovering over a 3-axis hall sensing IC that detects the Z-axis just as easily as the X and Y axis. They are still overpriced though...but fortunately they seem to be less expensive than their potentiometer-based counterparts by about 30% which turns out to be quite a bit since they are still over priced lol.

It's quite a neat application of hall sensors actually- the far end of the stick supporting a magnet that just moves around the above the hall sensor IC for the XY and rotates above the IC for the Z-axis.
http://www.melexis.com/Sensor_ICs_H...Triaxis™_3D-Joystick_Position_Sensor_648.aspx

THat little one you linked is the kind of thing I'd use a camera control for. THe sticks are a bit too short and sensitive to fly a model.

 

[Sceadwian]

DK a couple opamps could be used to select a voltage window from your pot and scale it.

 

[dknguyen]

DK a couple opamps could be used to select a voltage window from your pot and scale it.

Yeah, I know. But I'm trying to avoid extra things like that and am willing to take alternative routes to make it happen. Why do something you don't have to?

 

[Sceadwian]

Because it's easier and probably cheaper to use a couple opamps to adjust your signal properly than it will be to find a pot that has the exact physical requirements you desire.

 

[bryan1]

Ahhh, I realized I forgot a critical step. THe potentiometers I was going to use (and really, every potentiometer I can find) has a electrical rotation of 300 degrees or more. I only needed ~40 degrees rotation for the lever. I had known this, but had never realized how small 40 degrees is compared to 300 degrees. That means that my output voltage range is only 13% of my the full-scale voltage. If the ADC has 10-bit resolution, that's only a 136 steps. ANd I can't seem to find any potentiomters that have smaller electrical rotation angles.

Easy fix, just put a small gear on the pot shaft and a larger gear on a lay shaft that holds the friction disk and lever.

 

[shortbus=]

How about using these; Precision Linear Motion Potentiometer Selection Guide

Some thing like the LCP12Y would be ideal. mount one end to the bracket, the other to the lever. Adjust the length of the pivot of lever to attachment point of the pot get the correct stroke.

The same company has a lot of different style joysticks too.


Using the linear pots will allow you to get the levers as close together as you need also. Using the original pots you would get a two levers next to each other, then one inch of space, then the next two levers. Because of the size of the pot bodies( 1/2 inch each)

 

[dknguyen]

How about using these; Precision Linear Motion Potentiometer Selection Guide

Some thing like the LCP12Y would be ideal. mount one end to the bracket, the other to the lever. Adjust the length of the pivot of lever to attachment point of the pot get the correct stroke.

The same company has a lot of different style joysticks too.


Using the linear pots will allow you to get the levers as close together as you need also. Using the original pots you would get a two levers next to each other, then one inch of space, then the next two levers. Because of the size of the pot bodies( 1/2 inch each)

Yeah was searching long and hard for linear pots initially. They were my first practical choice (with the thin multiple throttle levers being the ideal, but expensive route). BUt I ran into huge difficulties finding panel mounted ones (I don't want to have to mount some on a PCB then mount that to the enclosure...more things to go wrong). Only two sources...Alps which are overpriced and unavailable and Bourns which were all dual ganged. Both kinds were only available in carbon composition either way. THat's when I decided it might be worth it to get throttle levers made. If I can't find <60 degree pots, then I may go with two single axis joysticks (not enough space for more) and two sliders.