How to modify a rudder dead area?

[Perdu]

Hi guys.

Can anybody knows how I can modify my transmitter (Futaba 6EXA) rudder stick? I need a dead area from the centre to about 30% travel right and left and then normal to the end, if possible adjustable.

After my stroke I'm trying to re-teach myself from mode2 to mode3 (left stick ailerons, elevator, right stick rudder, throttle) but I can't feel my right hand so I find it difficult to sense the centre point when I fly (cause weird aerobatics ;) and lots of crashes :( )

Sorry if the post is difficult to understand, sometimes I cannot find the proper words.

[vintage1]

Try a lot of negative exponential. It wont be a dead area, but it will reduce sensitivity. A LOT.

[Perdu]

I'm using full negative exponential but is not enough, and I maybe need stronger stick springs

[Acetronics]

Hi, Not-too-perdu

the trick would be to use an ear plug ( 2 nd Rx "in ze pocket" + little beeper ) ... Beep low for left and beep hi for right - to learn you to place your hand.

Just a 2nd Rx in the Pocket, plus a simple beeper.


the dead band is not a solution ... you do know it

Also To test : the three axis stick ...

Alain

PS: c'est bien la première fois en 40 ans que j'entends qqun réclamer pour utiliser les "3 axes" ... lol.

[Perdu]

Hi Alian

That's a good idea, I'll try but because I can't feel, I used the full travel stop to fly.
The throttle is OK for me, I'm using a longer stick (4 ins) and grip similar to a car gear stick, I can't use my fingers.


sorry but I can't understand French, I used a translate, I think you're saying used only ailerons, elevator and throttle. I can but I hope to use rudder if I can.

Regards
Tony

[pmackenzie]

sorry but I can't understand French, I used a translate, I think you're saying used only ailerons, elevator and throttle. I can but I hope to use rudder if I can.

Regards
Tony


I think your screen confused him.
Perdu is French for "lost" :)

You could do it mechanicaly in the airplane.
Have the rudder conected to an idler are that is spring loaded to center. The have a lot of slop in the conection between the servo and the idler.

Or just spring load the rudder to centre it and use a sloppy connection to give it some deadzone.


The stick assembly might also be able to be modified to build in some slop.


Electronically, a PIC could be programmed to produce an adjustable deadband. It would go between the receiver and the servo.

Pat MacKenzie

[Acetronics]

Hi, Tony

I had a Google Earth look to your field ... nice.

For the PS: generally, most people ( excluding Sailplanes users ) only use "2 Axles" - Ailerons and elevator - and always "forget" the Rudder ... :rolleyes:

BTW, if using a small µProcessor ... why not use a MMA 7260 or Equ 3 Axis Accelerometer ???

You place it on your Head ( a helmet can ALSO be useful in its normal use ... :D ) and, with a little circuitry, it can drive your rudder ... just a plug to place on your Tx.

Head bent left = rudder Left ... héhé ... :p

Elektor magazine raised some times ago an inexpensive gadget on that principle.

http://www.elektor.com/magazines/2007/april/two-axis-2g-accelerometer-with-spyder-and-a.81064.lynkx

Alain

[Brandano]

You could also try to modify the left stick to have the rudder on a twist motion. This means you'd have to fly pinching the microstick rather than using your thumbs, and might not be easy to make it into a compact enough package. Seems to work well enough on PC joysticks, and was also used in the Comanche prototype.
[edit] googling for "single stick transmitter" brings up a few results. Most are simple 3CH transmitters, but some are custom designs with pitch roll and yaw on a single stick, albeit most of them appear to be set up to be operated with the right hand. I think it should be possible to have the transmitter modified for use with a single hand, perhaps replacing the throttle channel with a slider with a couple of spring loaded detents too.

[kd7ost]

Hi guys.

Can anybody knows how I can modify my transmitter (Futaba 6EXA) rudder stick? I need a dead area from the centre to about 30% travel right and left and then normal to the end, if possible adjustable.

After my stroke I'm trying to re-teach myself from mode2 to mode3 (left stick ailerons, elevator, right stick rudder, throttle) but I can't feel my right hand so I find it difficult to sense the centre point when I fly (cause weird aerobatics ;) and lots of crashes :( )

Sorry if the post is difficult to understand, sometimes I cannot find the proper words.


Hi Perdu,

A few good idea's here and a few bad ones. One idea could induce major flutter.

How much throw do you need and on what kind of plane? I mean are you trying to fly aerobatics with a big rudder ship or is it a trainer type plane? Prehaps something in between? What do you need the rudder to do? Yaw for takeoff and landing correction or hold knife edge flight?

Dan

[alexcmag]

If you have someone near with a PIC programmer, I can make a tiny program for you, it only needs a programed PIC12F629, very cheap, a servo extension, and very few soldering.

Last year I was going to a flight field for the weekend, with the car roof full of model airplanes. I stopped on the local bakery and met a guy who asked if I would like to buy his glow model airplanes, because he had a strike and was not able to flight with only one hand.

Since I have a broken PS2 joystick, I started to build a one-handed transmitter, but due to my work I didn't finished it yet.

It uses one stick of PS2 control for aileron/rudder and elevator, microswitch on the stick (L3/R3) for self-trimming, R1/L1 for throttle up, R2/L2 for throttle down. It is enough for most models.

I also bought a pair of 3-axis accelerometers to limit vertical acceleration (to limit elevator) and automatic rudder coordination.

[Perdu]

Hi guys

Thanks for the help.

Pat: I used to set cryptic crosswords puzzles and I used Perdu (lost from view) for the name I used, now I can't even spell :rolleyes: lol I'm writing from Via Voice dictation as the moment.

I like the mechanicaly solution, simple is my middle name, I'm a electronic dummy. I am thinking about buying a cheap light electric foamy to learn so flutter is not a problem. Maybe spring loaded rudder with a closed loop and slack wires. In the future I'll have to use a electronically solution.

Alain: my field the OK (Google Earth 51 39'20"N, 0 03'44"W) but Electric only. :( but is only 10 minutes away from home :)

maybe I'll have to settle for 2 Axles now but I was just getting hang of the rudder, stalls turns, knife edge etc when my stroke happened.


if using a small µProcessor ... why not use a MMA 7260 or Equ 3 Axis Accelerometer ???

???? sorry, as I say I'm a electronic dummy

I like the idea for a helmet controller :D


Brandado: I read a magazine article years ago about people modifying the transmitter using a wheel from a car transmitter, on the top of their stick. something to think about, similar to try to learn to pat my head and rubber your stomach at the same time. :)


Tony

[Perdu]

Hi Dan

How much throw do you need and on what kind of plane? I mean are you trying to fly aerobatics with a big rudder ship or is it a trainer type plane? Prehaps something in between? What do you need the rudder to do? Yaw for takeoff and landing correction or hold knife edge flight?

really something in between I suppose.

I'm a normal club flyer I suppose use rudder for takeoff, stall turns and I was trying to learn knife edge.

Tony

[Perdu]

Hi Alexcmag

I have to find a PIC programmer but haven't got a clue what a PIC does. :confused:

Last year I was going to a flight field for the weekend, with the car roof full of model airplanes. I stopped on the local bakery and met a guy who asked if I would like to buy his glow model airplanes, because he had a strike and was not able to flight with only one hand.
check out http://uk.youtube.com/watch?v=TiTIO6In-T4 RC Helicopter Flying using one hand only :cool:

Tony

[Brandano]

The problem can see with using a head mounted controller is that it restricts your head movement and may bring the model out of sight. Perhaps a controller made to be held with your teeth and activated with your tongue would give you enough control finesse. Or a foot paddle like the one used to regulate speed on dentist drills. The modifications to the transmitter would be in any case pretty simple ones, just rewiring a few potentiometers. In any case, everything should be tested extensively on the simulator.
Oh, i just thought of another option: big slider attached to the rudder channel on the bottom of the transmitter, and you just keep the transmitter pressed against your waist and move the whole thing sideways. Naturally with a centering spring and a good detent.

[kd7ost]

Hi Dan



really something in between I suppose.

I'm a normal club flyer I suppose use rudder for takeoff, stall turns and I was trying to learn knife edge.

Tony


Well let me throw this out there and you'll have to decide what will work. I'm all about simplicity and it sounds like you need to learn to fly in a new way anyhow.

I don't know if you use a transmitter tray. I don't. I rarely use a neck strap in fact but you should use one of those at least.

When I learned to fly 20+ (maybe just a little longer) years ago, my instructor was a real safety minded old guy that was very thorough. He wouldn't let me adjust my trim tabs in flight without looking at the transmitter. This idea from him came from some early model transmitters where it was easy for a beginner to thumb for the throttle or elevator tab and accidentally turn the transmitter off due to the power switch location.

So he made me keep my eye's on the plane and hold my transmitter up until it was in my field of view and just under the view of the distant plane. That way with just slight glances, without dropping my head, I could look to ensure I had my thumb in the right place. I could quickly rotate my eyeballs just a few degrees to do the job. This has stuck and I still do it frequently to this day even though my power switches are at the bottom of the case front now.

Building off this idea, what if you fastened a short brightly colored stick or small diameter rod onto your right stick. Maybe a brightly colored button on the top to make it look bigger. Then when you are ready to execute a rudder maneuver, just hold the transmitter up so you can see that knob in your lower peripheral vision?

Do you think that might work?

Dan

[Perdu]

Hi Dan

Really I want to fly with the same way, I am OK now flying left-handed but the problem is when I fly rudder I can't feel the rudder stick, I'm OK for the throttle because I can use the ratchet. Try to think flying without the springs from the rudder stick. Recently I was flying and I adjust the throttle and then I was flying nearly full rudder, not nice.

Really and needs a stop from the centre rudder stick and when I used rudder I have to force the stick to use right or left.
I've modify my transmitter using long stick, I can't grip my fingers or thumb so I used a palm grip also I got a long tray.

(I'd tried to add a photo, my modification and grip I don't know how to do it) :confused:

Tony

[Perdu]

I hope it works

http://www.rcgroups.com/forums/showatt.php?attachmentid=2131927&stc=1

[Acetronics]

if using a small µProcessor ... why not use a MMA 7260 or Equ 3 Axis Accelerometer ???


???? sorry, as I say I'm a electronic dummy

I like the idea for a helmet controller :D


Tony

Hi, Tony

Those little chips are able to measure accelerations ( they output voltage proportionnal to absolute acceleration ) ... especially for gravity acceleration.

so, they are able to show you when something is placed horizontally ... ( some electronic levels are now appearing, replacing the old " air bubble" levels ).

very simple to use, inexpensive, sensitive enough to show your head leans one side or the other ... but not too much !!!

let's add a little " bargraph " just over your eyes to give you a "leaning scale" and hop ... et voilà. :)

Alain

[ChrisAttebery]

Why can't you use a 3 axis gimbal with your left hand and leave only the throttle on the right?

I know there are radios out there that were made like this. The stick moves up/down/left/right and then twists for the third axis.

I had a Microsoft joystick that was like this in the mid 90's. It made certain gaming manuevers really easy.

[orraman]

Greetings Perdu,

I have a pic programmer and would be happy to try to programme a pic micro to go between your RX and rudder servo.
The software you have been offered is for an 8 pin chip and I have a spare board ready for aerial photography that could be reprogrammed. The surface mount board measures 8 x 18mm, input is via a servo lead with pins for servo output.

Many years ago I flew with a Flight Link 3 axis 'cuddle box' from choice and can vouch for the viability of this system.

Remap is possibly the most cost efficient charity ever. There are 'panels' (volunteer engineers and technicians) all over the UK with the sole aim of helping disabled people with mechanical and electrical needs that are not available 'off the shelf'. There is every chance that your local panel would be able to fit a 5K pot to the top of your stick for true 3 axis, if not, I could try, but local is almost essential for fitting things to people.

If anyone has a schematic and pic software for a gyro solution I could try to make such a board for you.

Dave

[alexcmag]

Dave,

Since you're near, if you would like to try, attached is a simple C code for servo center dead zone, but the code itself is useful for a lot of servo toys.

It was compiled with CC5X: http://www.bknd.com/cc5x/

I made it for PIC12F675, but PIC12F629 is OK.

Input is on pin 4, output is on pin 5, so it only needs 4 pins of PIC.

The HEX can be written to a PIC12F629 without changes if your programmer don't complain. If so, please change the #pragma chip below and comment the " ANSEL = 0b00000000; // for 12F675" line. Or just tell me for which PIC you need the code.

#pragma chip PIC12F675
#pragma config = 0x2E4C

#pragma bit can1 @ GPIO.3
#pragma bit out1 @ GPIO.2

// servo limits, in microseconds
#define SERVO_DEAD_ZONE 400
#define SERVO_MIN 800
#define SERVO_MAX 2200
#define SERVO_CENTER 1500


The SERVO_DEAD_ZONE constant tells what is the dead zone size. In this case, it is a huge dead zone, but it can be any value between 0 and 600 (full servo travel for one side)

It was not tested in fact, because I'm without PICs here, but the core was tested in a lot of products and I tested the "dead zone" transformation logic in a desktop program.

Unfortunately sending it from my country (Brazil) to UK is very expensive, but if needed I expect to travel to Europe in November (France, actually), I hope freight is cheaper within EU.

[Perdu]

Hi guys

Again thanks you for the help.

I think the pic solution is best for me. I've been two years practice to learn to fly left-handed and the thought to learn 3 Axis stick is frightening.

Dave: Thank you, if the offer still stands I'll be most grateful to pick up (sorry, I know it's the wrong phrase) the offer. Of course I'll pay for your time, materials, expense and p&p.
I'll source the strong springs, the Ripmix, Futaba head office, service center is only 20 minutes away.
I've got a brushless Twin star, suitable to learning.

and alexmag thank-you for help with the code.

Tony

[Acetronics]

Hi, Alex

I'd like you to confirm ...

Signal inside "dead zone" = servo stays at center pos
Signal outside "dead zone" = signal passes to servo ???

Did I miss something ???

Alain

[Acetronics]

Hi, Tony

As I am just on the other shore of the channel ... could you PM me a Postal Address and confirm the servo plug type you use ( Futaba, I wonder ... ??? )

Regards
Alain

[Perdu]

Hi Alain

PM sent

Tony

[orraman]

alexmag,

Thank you for the software, but as I have no C language knowledge I am unable to find where the problem lies but 2 DIL (in different countries!) and one SM 12F675 chips have displayed the same 0.8V DC on pin 5 (GPIO 2) when full stick movement showed 1 to 2ms on the scope on pin 4 (GPIO 3). PICkit 2 showed checksum 004B and config as 004C which sees OK with All Protect showing.
Nowadays I use Basic or Picaxe but am happy to load Pic hex files.

Tony,
I am confident that amongst us all we will be able to help with the flying and for myself, I will help with anything I can.

Acetronics,
Earlier I got the word gyro mixed up with accelerometer but in the photo Tony is using a single point attachment for his neckstrap, would it be possible to get a useful output from a sensor if he is able to tilt the transmitter? Long ago I serviced self levelling tables for an X-ray machines that worked reliably with a pendulum. If such a system (not pendulum) was viable I would do any 'grunt' work that I can to help get it made.

Dave

[Acetronics]

Hi, Dave

Time for me to go to the Kitchen ... yes, a french "chef" ( lol )

I continue the chat tomorrow morning.


Alain

PS: I felt a bigger problem in Alex's soft ... my questions come from that ... What about the transition ???

[ZAGNUT]

just be careful adding stronger springs: http://www.rcgroups.com/forums/showthread.php?t=943178

and single stick radios with a big rudder knob on top used to be quite popular and still are today, especially amongst the handicapped. if you check out the yahoo group "MP8K" you'll find a few fliers that still use these as they can't fly any other way. a couple of guys there are even having a small run of CNC machined "chidgey sticks" made up.

[orraman]

Acetronics,
Should the transition at the edge of the 'dead band' be a step or a linear progression?
The step should be possible on a Picaxe 08M within it's 80 lines of basic and may well be all that is required.

ZAGNUT,
A most impressive photo, I followed your link as far back as http://www.jensenjetmodels.com/ and their modern 3 axis transmitters, braw!

Dave

[alexcmag]

Sorry, guys, there was a problem with my code... Now it is correct.

I set dead zone to +-250us as example, I think it is a good parameter

It works this way:

If input is between 1250us and 1750us, output to servo is always 1500us (center)

If input is less then 1250us, output to servo is increased by 250us:
- For 1200us, output is 1450us
- For 1100us, output is 1350us
- For 1000us, output is 1250us
- For 900us, output is 1150us

If input is more then 1750us, output to servo is decreased by 250us
- For 1800us, output is 1550us
- For 2000us, output is 1750us
- For 2100us, output is 1850us

This way it also decrease servo travel (response.gif shows how it works graphically).

It is quite easy to change the math for "dead zone", but I would like to know what is better.

Some ideas:
- Same as above, but with a wider dead zone

- Same as above, but with full servo travel

- Non-proportional: near endpoint left make servo to do full travel left, near endpoint right make servo to do full travel right, useful for hummerheads, snap rolls

- Keyboard-game style: near endpoint servo starts to turn in selected direction, near center or near other ending servo starts to go back to center, so it is possible to have a near-proportional servo without feeling the stick by "shaking" to the desired side. This might be useful for turn coordination on slower/bigger airplanes.

[orraman]

alexcmag,

Sorry, but not there yet on my set-up.

Futaba Tx, trim centred, surface mount 12F675 on small camera board~~
Stick moving right, immediate restricted servo movement to about half stick travel then no further servo movement.
Stick moving left, no servo response until near limit of stick travel, then very small servo movement.

With trim hard right ~~
Stick moving right, dead band with small servo movement, ending before the end of stick travel.
Stick moving left, small dead band then very small servo movement, about 1/16", then on further servo movement up to the limit of stick travel.

Servo movement to the right is linear (without any step) from the edge of the dead band.
I could offer scope readings if you wish.

Dave

[Perdu]

Hi guys

Zagnut: Thank you for a warning about this springs

When a I started this thread I thought it was easy to resolve.

In my half dead brain ( lol ) I could possible modify the rudder pot itself, maybe cut, remove the coating on the middle section on the pot board and some way connect the now two sections together.

Tony

[Acetronics]

Hi, Tony

I'm a bit embarrassed with the "onboard" solution ...

see :

- the solutions we've looked at ... never deal with the trim setting. : How to allow the trim to work as usual ??? ( using old "trim-servos" , or a servo-on tray system ??? )

- The deadband is Ok, but there will be no "smoothness" at all as we will very kwickly reach end of travel of the stick. Very good for snap rolls ... but not for Taxiing ... or fine rudder use ...

The "pseudo exponential" gizmo is very easy to program ( despite some trails ... lol ) - I will be able to send you one this week.- ... but will that be a real and useful answer to the question ???

I really do not know ...

Alain

[pmackenzie]

I just had a thought.
Instead of making a PIC based device to go in the plane, one could be made to go in the transmitter. There is a similar thing for the Parkzone 2.4 ghz to give reversing, v-tail mixing, etc.

The PIC would read the pot and drive an electronicdigital "pot" that would be connected where the original pot was.
http://www.analog.com/static/imported-files/Data_Sheets/AD5245.pdf

The advantage would be that you would need only one device for all your planes.
Disadvantage would be slight loss of resolution since it is only an 8 bit device. You should be able to do 10 bits with the in-plane device.

Pat MacKenzie

[pmackenzie]

Hi, Tony

I'm a bit embarrassed with the "onboard" solution ...

see :

- the solutions we've looked at ... never deal with the trim setting. : How to allow the trim to work as usual ??? ( using old "trim-servos" , or a servo-on tray system ??? )

- The deadband is Ok, but there will be no "smoothness" at all as we will very kwickly reach end of travel of the stick. Very good for snap rolls ... but not for Taxiing ... or fine rudder use ...

The "pseudo exponential" gizmo is very easy to program ( despite some trails ... lol ) - I will be able to send you one this week.- ... but will that be a real and useful answer to the question ???

I really do not know ...

Alain


As I thought about this, here is how my "gedanken" device would handle trims for an on-board solution:.

The device should also read the gear channel output, much like remote gain gyros do.
With the switch in one direction the device reverts to "straight thru" operation so the trim behaves normally.
Flip the switch the other way and two things happen.
The neutral point for the deadband is taken as the pulse width at the time the switch is thrown. In addition the deadband is adjusted based on how far from neutral the gear switch ATV has been set.

Pat MacKenzie

[Acetronics]

Hi, Pat

Problem ...

will the free hand be able to toggle a sooooo little Gear switch ??? ... not sure at all.

Alain

[pmackenzie]

He could get someone else to trim the rudder for him.
Once the plane is trimmed and he is flying there will be little need to flip the switch.

The idea is to make the device field programmable in as simple a manner as possible.
Both neutral trim and deadband adjustments can be made in the field.

Re-compiling every time you want to make adjustments is not very practical. :)

Pat MacKenzie

[Acetronics]

He could get someone else to trim the rudder for him.

Pat MacKenzie

I understood this is part of the problem ... to be independant. :)

Alain

[pmackenzie]

Flipping the gear switch is no more difficult that turning on the transmitter.
He is flying now, so presumably he can trim the model, or is getting it trimmed.
The gear switch method is not complicated or involved.


Or do you have a better idea?

( I left out one point - neutral would also be learned at turn on time so there is no need to flip the gear switch every flight)

Pat MacKenzie

[Perdu]

Hi guy

If I understand you're trying to do, using the gear switch to trim to the rudder.

I can easily move the gear switch to the position where the trainer switch is, I'm flying with my left thumb so I can use my left fingers. or if it's easier I never used the flap Channel, using Channel six for the trim, its proportional, and again I can easily reposition the flap control.

Tony

[pmackenzie]

Not exactly.

The gear (or flap) switch selects bertween normal and deadband mode for the rudder.

In normal mode the rudder trim and rudder control will work normally.
So you fly or trim on the bench with the rudder trim as usual.

Then you flip the switch and you have the deadband activated. The rudder position during the deadband will be the previously trimmed neutral.

Because of the deadband the trim won't do much of anything.

You can adjust the amount of deadband by adjusting the amount of gear (or flap) travel away from neutral.


Make sense?

I am going to try to work on this a bit today, but I am not as fast at coding as alexcmag :)

Pat MacKenzie

[Perdu]

Hi Pat

I understand.
I'll butt out now. lol

Tony

[Acetronics]

Hi, Pat

The thing difficult to understand in your idea is you speak about a gear SWITCH and further tell this command is adjustable ...

I'd understand the use of the Flap POT to set the Deadband amount ... and the gear SWITCH to memorize it AND toggle mode ...

Ok for the trim memorization, I do agree that's fine.


If Tony thinks it's doable for him ... I code ( PicBasic Pro ...) and weld it.


BTW Tony : Are Flap and Gear following each other in time ( Channel N° 5 and 6 ) ??? - programming way is linked to time between outputting two channels.

Alain

[pmackenzie]

You adjust the gear throws in the radio to adjust the gain.
The same method is used with gyros that have a remote gain function.

Only higher end radios have sliders for auxiliary functions, but almost all have a gear switch that can have its end points adjusted.

The switch is handy because you can pick the OFF position and then go right back to your previous amount of deadband. Not as easy on a slider or knob.

Pat MacKenzie

[Perdu]

Hi Alain

on my transmitter: Gear is Channel 5 and flaps is Channel 6

Tony Howe

[pmackenzie]

To put some numbers to this :

if gear pulse width < 1.5 msec then dead band = 0

for gear pulse width > 1,5 msec then deadband = gear pulse -1.5 msec


Pat MacKenzie

[orraman]

Greetings Tony,

I have sent you Private Message.

Dave

[alexcmag]

Dave,

Looks like timing is different in my setup and yours...

I tested with my own design servo-tester, but since it was not a real test, I tested again using my Futaba T7YF TX with a GWS R4P receiver.

In both cases, with trim centered, servo is centered from half-travel left to half-travel right.

After that servo moves.

If it is not difficult for you, please attache scope readings, but it is possible a problem with trim, subtrim or stick centering. So... Pat's suggestion of using a channel to turn dead-zone on and off is great.


Pat, I liked the idea of using a second channel for activating it.

Actually, this software used to be a mixer before removing code, so my prototype hardware is ready for testing.

I like these kind of brainstorming...


Perdu,

I forgot to ask if you need full servo travel or not, and what length of stick dead-zone you need.

[alexcmag]

BTW.

About modifying the TX...

There is no need to use a digital potentiometer to simulate the real one.

The easier way is to place a micro controller inside a DIN-6 connector, reading PPM out and placing signal on PPM in. Then can replace the trainer switch by a on-off switch, so when you turn on trainer switch with the "dead-zone plugin" in the trainer port, the dead-zone is enabled.

If the connector is not in place, the TX electronics don't try to get the PPM in from trainer port even if you turn trainer switch on, so it is safe against accidentally removing it.

[pmackenzie]

Good point, however:

The main advantage of the digital pot idea is that the trim will still function.
Intercepting the PPM stream will have the same problems in this regard as the on-board solution.

Another advantage would be that any mixes involving the rudder would still work properly, which is not the case for a PPM stream or on-board solution.

The digital pot I linked to is only worth a couple of bucks. You would also need to add some sort of pot to adjust the deadband and a switch to turn the function on and off.

Pat MacKenzie

[alexcmag]

Good point, Pat

In this case, don't worry about 8 bits.

As I said, the program I posted born as a v-tail mixer (actually I use it as a any-mixer).

First versions had only 100 steps, and since a few years ago there was not many suppliers of RC acessories, I made a lot of them for my friends and nobody noted it was only 100 steps... Currently the PIC12F683 version has 1024 steps, but actually it doesn't matters, so 256 steps from a 8-bit eepot is OK for most uses.

[pmackenzie]

It just occured to me that with the 9303 you could program a deadband. :o

Use one Pmix to mix rudder to rudder at -100% to kill the output.
Then use one of the mixes that let you set a curve to get the throw back, but leave it flat in the middle to get deadband.
( I tried using only the Pmix with curves but could not quite get it to work)

Works quite well.
Trim still works.
You can switch deadband on or off.
Total throw is unchanged.


Only downside: you only have two choices for dead band width since there is only a 7 point curve. Either 3 points wide or 5 points.

This should work with any transmitter that has mixing curves.

Pat MacKenzie

[arkcom]

I realize this is a whole different direction than you guys are going but... How about installing the rudder control arm sideways (hole front to back), put a wheel collar on the pushrod 3/4 in front on the control arm, then a spring from an ink pen, then the control arm, then a spring, then another wheel collar. Very simple, and can still be trimmed.

[Perdu]

:mad: :censored: I have a prolbem my voise-text and text-voise prog, talk when it's ok.

[Perdu]

Hi Dave

I sent a reply from your Private Message

Tony

[Perdu]

Hi Alex

I forgot to ask if you need full servo travel or not, and what length of stick dead-zone you need. I don't need full servo travel, I can use push rods and horns to be adjust the throws.
I was thinking about the dead zone last night. On my transmitter, centre stick to full travel its about 30 degrees and I think I need about 10 degrees left and 10 degrees right dead zones.

Tony

[orraman]

alexcmag,

Scoped the 12F675 as you requested and the trim centre of that TX is not 1.5ms, thank you, your programme is fine.

Tony,

That means that it is now on a small SM board with alexcmag's programme and a Futaba lead if you would care to try it.
And I would be delighted to send it to you.

Thank you for the PM.

Dave

[Perdu]

Hi Dave

Thank you very much.
I'm very happy if you can send it to me.
I'll send a PM to you.

And thank you Guys again for the time and brain work you'd used.

Tony

[alexcmag]

Dave,

Thanks for the info. I was concerned about variance in internal RC oscilator.


Perdu,

The code I posted before has a half-travel dead area (15 degrees in your TX), so I modified it for 1/3 (attached code).


To programmers: this change is made by changing the SERVO_DEAD_ZONE constant from
#define SERVO_DEAD_ZONE 250

to

#define SERVO_DEAD_ZONE 166

The math is: travel left or right is 500ms, since Perdu needs near 1/3 dead area, 500/3 = 166

[AndyKunz]

Didn't somebody used to have a gameboy-style RC transmitter? Would maybe pushing buttons for controls be easier to handle? Just thinking...

Andy

[orraman]

Greetings Tony,

Managed to get the board to the Post in a rush so no note inside.
Plug your rudder servo in line with the colours of the heatshrink on the pins above.

Anything else let me know.

alexcmag,

Now you tell me! Nothing lost, I could reprogram the one that's on it's way to Tony once he has some 'hands on' experience.

Tony
I have another board so if I can find another SM 12F675 I could send the new program for you to choose between both programs if you wish.

AndyKunz

As a one time Tapper and Dunter of Reed Radio switches to get the servos to dither off-centre for smoother turns I suggested that with a Servo Slowdown program that may be a way forward for him.
Once he gets some practice with the 'dead band' rudder it could well be a useful technique.


Dave

[alexcmag]

Didn't somebody used to have a gameboy-style RC transmitter? Would maybe pushing buttons for controls be easier to handle? Just thinking...

Andy

I'm doing it, actually, but didn't finished yet.

To keep it light, easy and powerful, I'm doing it with a 2.4Ghz module (XbeePro) with 1 mile range.

This module is light, cheap, reliable and low-power, so I can make all the stuff to work with 1 cell LiIon and a 3V regulator.

[AndyKunz]

Hi Alex,

I was thinking of a commercial tx out there, not a home brew. I was sure I saw one advertised a few years ago.

Andy

[ChrisAttebery]

Wouldn't this be the same as setting Expo to -100% (Futaba)? you don't get any throw until you hit the end of the stick travel.


Didn't somebody used to have a gameboy-style RC transmitter? Would maybe pushing buttons for controls be easier to handle? Just thinking...

Andy

[Perdu]

Hi Dave.

Thanks are received it today.

The dead zone itself its good
but the trims are really weird.
I think the trim it moving the centre of the transmitter dead zone itself.

trim centre: stick towards the left I have about 5 degrees dead, servo movement 32 degrees
stick towards the right I have about 25 degrees dead, servo movement 3 degrees

I move the trim +110: stick deadzone equal 15 degrees servo movement equal 18 degrees
That OK I'll use the push rods to trim my rudder.

Tony
attached a graph for the deadzone numbers

[orraman]

Tony,

Looking at your numbers, what you have is viable in as much as using +110 does give you equal deflection and you still have a wee bit extra trim if needed.
Your results are better than my results but on alexcmac's urging I measured the pulse length on the scope and I found centre was a not the proper 1.5ms and this could be happening with your Tx to a lesser extent. Reprogramming that chip to a better centre would help if needed.

I have loaded alexcmax's new 33% program on a smaller board and checked it with a different Tx and it is much more central, well within the trim range.
I have arranged for it to be posted tomorrow so you will be able to try both to see what you fancy.

The original Twin Star flew well on rudder, for the first few flights with a new beginner I would transfer the other channels to the buddy box and 'smooth' thing out a bit and "fess up" after they were doing well.

Dave

[Perdu]

Hi Dave

My twinstar flying well, I disabled to rudder when I had started flying left-handed, today I connected rudder again so I'm ready to go, weather permitting.
I've set up the rudder with small throws and when I OK with that I'll try larger throws.

Tony

[alexcmag]

Good to know it is working, even with these center problems.

The problem with centering can be solved using an extra channel to activate it. Or... just by considering the initial position on turn-on as center. What do you think about it?

[Perdu]

Hi Alex

Thank you for compiling a code.
I think I'll not use another channel at the moment, I can easily used the push rods for trim. I'm eager to try but the weather.... snowed last night.

Tony

[alexcmag]

New release...

This one has auto-centering.

It will "remember" the initial position and use it as center.

[Perdu]

Hi Dave and Alex:

I received the other one yesterday, thanks.
The new one is near perfect, trim only +10, more travel 25 degrees each way.

Maybe it's different when I fly for real but in my living room, checking the rudder without looking down to the transmitter, I have no unwanted movement from the rudder at all.
I'll let you know how I'm doing when I fly for real.

and Guys again thank you very much for the help you'd given me.

Tony

[alexcmag]

I'm glad to help, please tell how it works in the flying field, and if you need any change.

[Perdu]

Hi guys

Today I had a window for the weather so I went to the field.
It's great, first flight I reduced the rudder throws, got good altitudes then pottering around with circuits and loop using only rudder. Second flight I increase the throws, several untidy stall turns and tried a couple knife edge, I hold it for 2 or 3 seconds ( the Twin Star not good for the knife edge lol )
Dave, Alex and Guys thank you very much for your help.

By the way, third flight, launched and the Bec cut. my fault I forgot to check the battery. anyway it'll be easily repairable.

PS: on my Twin Star when the Bec cut, one motor cut but the other motor gets more power, weird ( matched brushless Motors, matched speed controller and I cut the red receiver cable from one esc )

Dave are sent a PM

Most regards
Tony

[orraman]

Tony
Perhaps?
Each Electronic Speed Control will have an internal Low Voltage Cut-off for its own motor and when the first motor cuts the reduced load will allow the battery voltage to recover to some extent increasing the revs of the other.

If you wish a LVC to act on both motors at the same time you could consider a micro.
The throttle channel of the receiver could be input to a micro, tested, and passed on to the motors.
After a delay (shorter than that of the ESCs) a loss of signal would send 1ms pulses to the motors to shut them down until a solid link is re-established.

A lead from the battery positive could feed a potential divider on the micro which would take a running average of the pack voltage.
At some voltage slightly higher than LVC of the ESCs the micro would shut down the motors with the possibility of a further burst of power if the battery recovers sufficiently.
All of this is within the ESCs, but they act independently and in the case of a twin ............

Such a programme fills a Picaxe but I wonder if there is space and pins on a 12F675?

Glad you have an easy repair job.

Alex I have your new software downloaded.


Dave

[evo62]

Good on you Perdu!

I stumbled on this thread just 'browsing' around and am glad to hear you've got back into the air. Well done!

Cheers
Clovus

[pmackenzie]

Tony
Perhaps?
Each Electronic Speed Control will have an internal Low Voltage Cut-off for its own motor and when the first motor cuts the reduced load will allow the battery voltage to recover to some extent increasing the revs of the other.

If you wish a LVC to act on both motors at the same time you could consider a micro.
The throttle channel of the receiver could be input to a micro, tested, and passed on to the motors.
After a delay (shorter than that of the ESCs) a loss of signal would send 1ms pulses to the motors to shut them down until a solid link is re-established.

A lead from the battery positive could feed a potential divider on the micro which would take a running average of the pack voltage.
At some voltage slightly higher than LVC of the ESCs the micro would shut down the motors with the possibility of a further burst of power if the battery recovers sufficiently.
All of this is within the ESCs, but they act independently and in the case of a twin ............

Such a programme fills a Picaxe but I wonder if there is space and pins on a 12F675?

Glad you have an easy repair job.

Alex I have your new software downloaded.


Dave

There used to be three devices on the market that can do this.

Now there only seems to be one:
http://www.dimensionengineering.com/LipoShield.htm

(FMA and Diversity used to sell similar products)

Pat MacKenzie

[Perdu]

Hi guys

I searched the threads for twins problems.

ALWAYS use a soft cutoff bec for a twin!!!
Tony

[Perdu]

Hi Dave

I've sent a PM

Tony

[orraman]

Tony,

Nice catch that

Quote:
ALWAYS use a soft cutoff bec for a twin!!!

Both boards re-programmed with Alex's Automatic Hex and on their way to you.

Alex,

Works like a charm, thank you.

Dave

[alexcmag]

Tony,

Nice catch that

Quote:
ALWAYS use a soft cutoff bec for a twin!!!

Both boards re-programmed with Alex's Automatic Hex and on their way to you.

Alex,

Works like a charm, thank you.

Dave

You're welcome.

Fortunately you're near to program the PICs, my travel to Europe was postpone for december, so it would a long time for me to post from Germany to UK...

At least I will probably get some snow in Bavaria (in my country to see some snow we must keep bags ready for travel, because it snows only a few days in a small montain city 1000km away from here...).

BTW. If you look into the code, it is not difficult to make any special logic on it, so if you need anything, just ask.

[Perdu]

Hi guys.

I'll take the opportunity to say thank-you very much for the help, ideas and suggestions you had given me especially for Dave and Alex.

The board's and the code are perfect for me and I can't wait to practise again but I think the flying season is over in the UK.

I'll let you know how I'm doing in the new year.

Happy flying!
Tony