Throttle axis now connected - DIY Throttle Pointers Needed

[TheCoyoteHunter]

Hello,

I am building a throttle controller (pics below) and I need some pointers on some of the things I will need.

 

Here's what I'll need,

min 35 digital inputs (on/off, momentary on etc.)

1 Hall Effect input for the actual throttle axis 

5 potentiometer inputs (2 for TDC, 2 for wheel brakes, 1 for zoom control) 

 

Which control board would you guys recommend, currently I'm thinking of getting a Teensy 2.0++. Is there something better out there? 

 

Is there something specific I need to check on the potentiometers? Or am I good to go with 10k pots.

 

This is the hall effect sensor I am thinking of getting.

https://www.mouser.ca/ProductDetail/BI-Technologies-TT-Electronics/6127V1A60L5?qs=6gM%2Fzov0KphmIK%2BwSRnApw%3D%3D

 

 

1 more picture.

 

Edited September 16, 2021 by TheCoyoteHunter

[crash test pilot]

What software do you plan to use? For mmjoy2 a cheap but adequate replacement would be a pro micro clone (around 5 bucks). It has less input ports but for your planned 35 inputs i would use shift registers which you can daisy-chain for up to 132 inputs (8 inputs per sr chip; 8 in a chain, mmjoy supports 2 chains). For the atmega chip based control boards 10k pots are good, just make shure you get linear (b10k) and not logaritmic (a10k) pots. Nice looking prototype btw!

[No1sonuk]

That Hall sensor is very expensive.
Have a look at Tom Stanton's DIY flight controls video.  He used a pot for his throttle, but the section about the rudder pedals could give you an idea for building a custom Hall sensor arrangement.
This link is specifically to the rudder Hall sensor part, but I recommend looking at the earlier parts too.

As for the processor:
Anything with a 32u4 can be made to act as a native USB HID.  That includes Pro Micro (like a 32u4 Nano) and Leonardo (Same form factor as an Uno).
With 35 digital inputs, you may need to get creative with pin multiplexing, though.

Hang on, though...
If 4 of those 35 can be treated as a 4-way "hat" switch, have a look at the Leo Bodnar BU0836A
http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=94&products_id=204&zenid=c772b831809f02c493ede451411fdd4f

That takes 32 buttons (via a matrix) + 4-way digital "hat" + 8 analogue.

If you add this matrix breakout board, you get all the digital inputs broken out into 2-pin headers, and it includes, built-in, the diodes required to make the matrix work properly with multiple input switches closed.  The analogue inputs are already on discrete 3-pin headers.
http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=94&products_id=274&zenid=c772b831809f02c493ede451411fdd4f

The Pros for the Bodnar card include no programming required at all for it to work and higher resolution A to D conversion for the analogue channels than Arduinos.
The Cons include that you can't make special functions like you can in an Arduino. e.g. have two separate L and R magneto switches generate a third "BOTH" output when they're both on, and a discrete "OFF" output when they're both off.  Though I suppose that sort of thing could possibly be done by Joy2Key or somesuch PC software.

[agrasyuk]

For input-only device Bodnar boards are great. 

 

For control as important as throttle I wouldn't go with anything other then hall sensors. Yes they are more expensive then 1$ a piece.  I attempted to build a mechanical assembly with cheap sensors and a rotating magnet but I couldn't accomplish this reliably 

Edited August 26, 2021 by agrasyuk

[TheCoyoteHunter]

Thanks guys,

 

So the Bodnar boards, I cannot find one in Canada. Here is a link to 2 different Teensy boards I have looked at, these are readily available in Canada.

Teensy 3.2  https://www.pjrc.com/store/teensy32.html

Teensy 2.0++  https://www.pjrc.com/store/teensypp.html

 

I would use the Teensy or the Arduino software. I have zero experience with this, only watched a few YouTube videos, and it seems easy enough.

 

This is my 1st attempt at building a diy controller of any kind. I have never programmed a controller before, but I have programmed enough PLCs and the like, so I doubt that I'll have too many issues with programming. Initially I took apart a X-Box controller like the Logitech F310, the board in it has 14 digital inputs, and five 10k pots, but there is no way to get to most of the digital inputs, they are inside the circuit board with no solder points. With 14 digital inputs I could have used shift multipliers to make as many inputs as I would need. Diodes needed of course as well as initial setup of the controller would have been a little tricky, to get the shift keys set. 

 

I don't have a problem spending a more for a good sensor on the throttle axis, the one I listed above should be really easy to mount.

 

This is what I'm thinking of using as the TDC controller https://www.digikey.ca/en/products/detail/2765/1528-1730-ND/6193582?itemSeq=372689342

 

 

 

[agrasyuk]

Bodnar boards are only sold  directly from Leo's website as he is in UK . No programming is necessary, it will be presented to your PC as complete joystick Plug and Play device.

[TheCoyoteHunter]

I'll have to see what it takes to get a Bodnar board to my side of the globe, I really like that matrix breakout board in No1sonuk's post, diodes built in makes things easy and ready to go.

 

Thanks guys

Coyote

[TheCoyoteHunter]

EDIT (This one doesn't have any analog inputs)

Hey guys what about this one?

It has all the bells and whistles, it just says that "not all software supports more than 32 buttons" I read somewhere that DCS is good for 128 buttons.

Shipping to Canada is $10-$12.

http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=94&products_id=300

 

This one should be more than I need. 

Coyote

 

 

 

Edited August 27, 2021 by TheCoyoteHunter

[agrasyuk]

never had an issue ordering from bodnar , ordered several boards for custom projects, always gets here to the states in about 7-10 days.

Yes, BBI-64 is just a button board with no analog axis. works perfectly, I'm driving pair of my MFDs with it - has the exact number of inputs I need for the pair. 
though there is a somewhat strange glitch once i got into the 63 and 64th buttons for the held position of the off-night-day switch.

Spoiler

, if you use that board you will need something for analog axis as well.
for your device I'd go with the BU0836X 

 

good luck!

[TheCoyoteHunter]

Wow this is a nice version of the BU0836.

This model of the BU0836 would allow me to connect, reconnect and add things as I go along, all without soldering.

I'll order a few hall sensors as well, they are cheap enough to play around with.

 

Thanks 

Coyote

[Auntystatic]

this works really well too

https://forums.eagle.ru/topic/97434-an-arduino-usb-hid-controller-composite-usb-controller/

[TheCoyoteHunter]

Wow! Auntystatic that's pretty cool, it looks complicated, but once you study it a bit its actually not bad.

 

I ordered a BU0836X along with a few hall sensors, all from Leo Bodnar's web site. I will try to do the throttle axis with the hall sensor that Leo sells, if it wont be satisfactory I'll get a better one like listed below. So far I have really enjoyed doing this project with my 7 year son, I fear if I spend too much time programming I'll loose his interest. With the BU0836X I wont need any programming, and being that this is my 1st diy input device, I think the BU0836X's option to connect, remove and reconnect all without soldering, will be pretty nice.

 

I do not have any paddles. I control rudder via stick twist, but what I really need is individual analog wheel brakes. What I plan to do is build 2 separate analog push buttons on the lower front of the throttle handle. 

 

Hall Sensor from Leo Bodnar  http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=103_82&products_id=195

BU0836X    http://www.leobodnar.com/shop/index.php?main_page=product_info&products_id=180

Other hall sensor  https://www.mouser.ca/ProductDetail/BI-Technologies-TT-Electronics/6127V1A60L5?qs=6gM%2Fzov0KphmIK%2BwSRnApw%3D%3D

 

Thanks 

Coyote

[TheCoyoteHunter]

Good Day guys,

I have the throttle axis working, as well as an additional analog lever.

The additional analog lever is to be used for the engine rpm control of the WW2 planes, or really anything else. Both use a hall sensor, but different magnets.

 

Both hall sensors are the ones from Leo Bodnar, part #: Allegro A1302

The magnet on the main throttle axis is a rare earth rod magnet from Lee Valley, part #: 99K3602, 1/4" × 1/2" Rod Magnet

The 2 magnets on the lever axis are also from Lee Valley, part #: 99K3203, 3/8" × 1/10" Rare-Earth Magnet. The magnets are facing the hall sensor with 1 north pole and 1 south pole.

 

I am coming for using the slider thingy on my Thrustmaster TCA(160000) stick as the throttle. Going from 3/4 inch throttle movement to the 5.9 inch is well appreciated. Now I know hoy you guys keep the E bracket in the hornet so well.

 

For the input card I have the Leo Bodnar BU0836X, and so far I have to say I am very pleased. I have not installed any software, just plugged it in and used axis tune within DCS to set it up. 

 

On the throttle axis I am using approx 3/4 of the hall sensor's range, I found that if I went any closer to the high/low limits, I lose the straight linearity. For example, if I tried to use the full or near full range of the sensor, the throttle would move really fast at both high and low ends, and go much slower along the middle. I found that with the magnet closer to the sensor I got more range out of the sensor, but lost linearity.

Testing the throttle axis with the free version of the P51 which has lots of travel in its throttle, when I zoom in all the way and watch the throttle move, it does not move in steps, its just so fluid, so smooth. 

The lever axis, uses less than 1/2 of the sensors range, it is also nicely linear. When I assign it to the throttle of the P51, it also moves very smooth, really don't notice a difference in smoothness with the throttle vs. lever axis. Its just that the throttle has 5.9 inch movement and the lever only has 2 inches. 

 

Now, when I assign either the throttle or the lever axis to zoom in DCS, the zoom becomes slightly jittery at a few zoom levels. When assigned to the throttle you barely notice it. The jitter is not very much, my screen recording software (Shadow Play) barely picks it up. Looking around in the forums I notice that zoom jitter is a problem more people are facing. I guess I could install some software to tune the axis a bit but, so far I haven't had to.

I plan to put a rotary zoom controller on the left side of the throttle handle, like the rotary thingy the TWCS has. I will try it with a unipolar linear hall sensor (only reacts to south pole of magnet) and mount the magnet on threads, when you turn the knob the magnet will move closer/further from the sensor via threads. Part number for the unipolar sensor I'm getting is: DRV5056A4QLPG.

 

I have all the other buttons, switches and sensors on order, should have them by mid next week.

 

Here is a video and some pics.

 

 

 

[No1sonuk]

IIRC, most "good" setups as well as Arduino ADCs have 1024 steps.  The Bodnar board analogue axes have 2048 steps.  That could be why it's so smooth.

As for the sensitivity and linearity:  If you're using any magnetic fittings, such as steel screws, plates, etc. they will affect the magnetic field.  It looks like you're holding two magnets in place with a metal screw and washer.  That could be adversely affecting the field between the two magnets.

[TheCoyoteHunter]

1 hour ago, No1sonuk said:

IIRC, most "good" setups as well as Arduino ADCs have 1024 steps.  The Bodnar board analogue axes have 2048 steps.  That could be why it's so smooth.

As for the sensitivity and linearity:  If you're using any magnetic fittings, such as steel screws, plates, etc. they will affect the magnetic field.  It looks like you're holding two magnets in place with a metal screw and washer.  That could be adversely affecting the field between the two magnets.

 

The screws, nuts and washers are stainless steel. Stainless will not react to the magnets, however it might affect the magnetic field. 

I did the testing by taping the magnets and sensors, once I was happy with the results I was getting, I took measurements and then fabricated some plastic to hold the magnets and fastened it with screws. On the smaller lever axis, you will see I have a washer on the opposite side of the 2 magnets, this washer is not stainless and is actually held in place by the magnets. It increases sensitivity by approx. 10%.

 

Testing the sensor vs magnets is so interesting, that tiny hall sensor is truly amazing.

 

I found this document by Honeywell very interesting.

https://sensing.honeywell.com/hallbook.pdf