Dynamic Controls Power chair, RC Boat trailer project

[gcebiker]

These Solid PU tyres did not work out, not strong enough for the weight...its a lot lighter than it looks.

Chair was able to tow well with me in it on flat ground, at the very least it will be usefull for putting the boat away if i do end up using the car to tow it.

[gcebiker]

Esky/live well pods made for the back, need to clean off the weld splatter and paint.



9.9HP Long Shaft Tohatsu has been replaced with a 6HP Yamaha, fitting forward controls to it later in the week and the charging kit.

[ex-Gooserider]

Possible solution to the motor wrapping the wire around itself...

For this part, it doesn't matter what direction the boat is headed in, all you care about is the position of the motor with respect to the mounting brackets or some other fixed point on the boat....

Presumably there must be some sort of input to tell the computer which way it is pointed with respect to the mounting brackets, and a way to get a value of 0-360 out of it, or even better plus and minus 360.... Assuming the position number goes positive turning clockwise as looking down on it from above, and that straight ahead is zero, store a value for which direction the motor is turning. (0=clockwise, 1=counter-clockwise) and have some maximum value (say 300*) that is the most you want the motor to be able to wrap the lead around.... Twice the value should be more than 360, and be in a range that you don't typically run with the motor pointed at...

Allow the program to turn the motor as needed in either direction to +/- 300* but if it wants the motor to turn more than the value, require it to spin in the OPPOSITE direction and approach it from the other side...

As each turn command comes in, add or subtract it from the present position to get a new position, and / or read the position sensor to track the current value...

I.e - motor turns back and forth starting clockwise from zero to any position less than 300*, not a problem. Then command comes in that would tell motor to exceed 300* - say go to 320*.... The shortest rotation would be to continue clockwise, but instead require the motor to spin COUNTER-Clockwise to approach it from the opposite side....

Even if working back and forth near the limit value in one direction, there wouldn't really be a problem as the first time over the limit would make the motor spin back the other direction, and then the working would be in a range that wouldn't hit the limit....

Since I'd expect that most of the action would be with the motor pointing either mostly forward or mostly backwards, picking values that would trip when the motor was pointing sideways seems best....

Hope this makes sense....

ex-Gooserider

[gcebiker]

Yeah it has the data feed from the GPS/ NMEA GPRMC sentences and a digital compass.
Apparently i will also need an accelerometer as the swell is going to mess up the compass headings a bit.

This is the code that is working , like you say...most of the time i am going to be pointing right at where i want to go / stay so it should never be an issue or at least not a big issue. I know the DigitalWrite High/Low look back to front...this relay shield works back to front.

Code: Select all

void steering() {

  // calculate relative direction to destination

  int steer = gps.gprmc_course_to(storedLat, storedLong) - (currentCourse) ;

   // check if negative number, turn into positive.
 if (steer < 0) {
          steer += 360;
        }
        if (steer > 180) {
          steer -= 360;
        }



 Serial.print("CTD used for relay control -  ");
  Serial.print(steer);
  Serial.println();
 
  //Turn left
  if (steer <= -20 &&  steer >= -180)
  {
    digitalWrite(left, LOW);
    digitalWrite(right, HIGH);
  }
  else if (steer >= 20 && steer <= 180)
  {
    digitalWrite(left, HIGH);
    digitalWrite(right, LOW);
  }
  else
  {
    digitalWrite(left, HIGH);
    digitalWrite(right, HIGH);
  }
}


The big thing i had completely forgotten was that the motor is up at the moment...the cord with not much extra for getting wrapped around.
...once the Motor is DOWN...i will have heaps of wire and its all springy wire too.

All proceeding well, it should be in the water next week....car launch, its squid season and unfortunately, I don't have the time at the moment to mess about with the Electric wheelchair/launch and retrieve.

...i am a day behind at the moment, shop ran out of Locktite ! for the Remote control plate i put on the F6C Yamaha 4 stroke.

[Burgerman]

You need a 4x scaled up one of these. http://www.wheelchairdriver.com/gopro/test-control.mp4 With current sensors and correctly working motor compensation it drives exactly the same with a 12lb breeze block on top. It just sucks up extra amps. 3mm of stick, it turns.

[gcebiker]

Have you tuned the motor compensation? If too low, it will have trouble with "high load" situations, including turn-in-place. Careful though, too high and the chair can become unmanageable, even to the point of runaway behavior.

It was set to 108ohms...or was the mOhms, i forget. Anyway, i lowered it to 55 and tinkered with the other settings.

Improved the driving a lot...until i attached the boat.
The chair is limited to 55amps and thats where i think the problem lies, the limitation with the motors.

If i get out of the chair and use it for balance, i can slowly get it back up the driveway (which is a slight incline).
It would go faster but walking and driving the chair is difficult.

----
Ended up taking the boat out for its first run on Monday with the help of a friend.
He drove the boat to near where the car was parked and i was able to get from the car to the waters edge, its not something i think would be possible as an ongoing solution.

It had some slight hickups
- The DrawBar bent when the winch jammed and the boat slid off the trailer.
Easy to fix.
- The tow ball did not quite fit.
Already fixed.
...and i had to many people on board, thought it would be ok with 3...and it kind of was but at one point we were all on the one side and it got a bit hairy for a few seconds.

Other than that , very very happy.

Looking for ways (that i can afford) to fix the lack of power in the wheelchair.
I purchased it broken and its all fixed up now so its possible i could sell it back to the store and get a different one, first researching the limits of the motors installed in the donor chair.

They had another one there that looked to have beefier motors and it came with a Dynamic Controller, Ill be calling the store in the morning to see if they still have it.

Ill also bump UP the resistance , trouble is the setting is global...so it will probably make the chair very difficult to drive when the boat is not on the trailer.

More testing is needed !

Tony

[Burgerman]

It makes it easier to drive. The load compensation when set correctly, automatically compensates for load. So the chair will respond the same, either loaded or not, slope or not, grass or tarmac, hill or no hill.

When its not set high enough you must add more power to overcome a load or weight yourself, and then it runs away after you do it. Or it refuses to turn. When set even fractionally too high it over compensates and can run away.

[gcebiker]

It makes it easier to drive. The load compensation when set correctly, automatically compensates for load. So the chair will respond the same, either loaded or not, slope or not, grass or tarmac, hill or no hill.

When its not set high enough you must add more power to overcome a load or weight yourself, and then it runs away after you do it. Or it refuses to turn. When set even fractionally too high it over compensates and can run away.

Ah, i see. Now it makes sense, i will raise the value and see if it helps with the boat...it probably is under powered tho @ only 55amps.

[Burgerman]

You need to increase a little at a time, and test. At the point where it gets too jumpy, you either have turn rate or forward acceleration set too high. Or compensation set too high. In an ideal setup, with a heavily loaded chair, you will get full Amps that the controller can deliver at around 15% turn on the joystick. So turn rate doesn't need to be set high. 40 is a good max value, and 15 for low speed turn.

[gcebiker]

Made the changes to the resistance and its heaps better but still ME + Boat....not working out.

The Chair is to light and its not looking safe (because when i get off the chair lifts off the ground), so for now the wheelchair has become an awesome tool for moving it.
Our street is quite narrow with all the cars parked and its difficult to boat drive out and reverse in.

I park on the street and use the wheelchair to tow the boat down the driveway, along the footpath and then attach to the car.

Here is a video i took today while out squidding. https://www.youtube.com/watch?v=sNlmWI_ ... e=youtu.be

[woodygb]

because when i get off the chair lifts off the ground

... This may seem like a daft question but ...
Why does the chair lift up? .... I always thought that you set the nose weight of the trailer hitch at 10-15 percent of the trailer weight... thus it should be pressing down on the ball hitch attached to the rear of the chair.

OR ...Do you mean that the casters are lifted off the ground without your weight on the chair?

[Burgerman]

I suspect he means the tow hitch is too high. Needs to be about level with the chairs axle it the front will lift as it pulls the boat.

[gcebiker]

Ill try again, not sure if that link was right - Video of the boat on a little island in the broadwater on the Gold Coast of Queensland, Australia.
https://www.youtube.com/watch?v=sNlmWI_rXhU&feature=youtu.behttps%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DsNlmWI_

The ramp is quite steep, the wheelchair is doing wheelies when i try to drive forward (even with the light alloy trailer and no boat).

Perhaps the easiest thing to do would be to make the Roller M1 into a 'Front wheel drive' , in stead of its current Rear Wheel drive configuration.
It would mean making a new tow hitch but i am also thinking i dont have enough power with the current module/motor combo.

I have someone local that is thinking of donating their old Jazzy600 to me, they have not used it in over two years.

I have done testing in the driveway and the 6 wheel drive Q6Edge i am sitting in handles the load of the combine boat/trailer much easier than the Roller M1.
...the Q6 Edge i have has the drive electronics from my old Jazzy600 in it, so i am comparing like with like hopefully, tho on different platforms.

Regardless, i dont want to be sitting in the chair as i try to drive up the boat ramp, it has to be stable with out me in it..
6 Wheels should help a lot and i will be able to remove the seat and put the Tow Ball over the center of the axles.

[gcebiker]

because when i get off the chair lifts off the ground

... This may seem like a daft question but ...
Why does the chair lift up? .... I always thought that you set the nose weight of the trailer hitch at 10-15 percent of the trailer weight... thus it should be pressing down on the ball hitch attached to the rear of the chair.

OR ...Do you mean that the casters are lifted off the ground without your weight on the chair?

When i try to drive forward , while on the boat ramp , it wheelstands.
I probably need to turn the chair around.

Last nights effort from the footpath to where i can cast into the water.

[gcebiker]

Put new mid wheel and front caster wheels on the Q6 Edge today, after running around on flat topped PU for over a year....its turning a lot easier.

Completely unexpected and something i had not thought of for the M1 Prowler.
First up i am going to try increasing the tyre pressure, and then i will take off the air castors and put on the old PU castors i took off the Q6Edge today.

Its proving its worth just as a tow vehicle in the front yard.

As soon as i find a place to mount the trolling motor i will be able to test out the code (for the GPS spot lock) in the real world and start tweaking it.

[Burgerman]

If doing that helped it means you dont have enough load compensation.

[Burgerman]

DYNAMIC CONTROLS:

Load Compensation automatically compensates for changes in motor speed when the chair drives over loads such as sidewalks, curbs or slopes.

Note: The Load Compensation and Speed Demand Scalar parameters affect the performance of all other speed and acceleration parameters and as such it is important to set these parameters correctly before programming anything else. If these parameters are changed after the chair has been set up, the complete speed/acceleration programming and testing procedure must be repeated.

If Load Compensation is set too low, the chair
- Drives like it is going through thick mud
- Slows down when it goes up a sidewalk edge or up a ramp
- Slows down with heavier users
- Changes direction when it drives over a bump
- Changes direction when the weight of the user shifts.

If Load Compensation is set correctly, the chair
- Drives smoothly
- Keeps the speed reasonably constant. Only slightly slows down on a slope or on thick carpet.
- Keeps the direction constant. Only slightly changes direction when it drives over a bump.

If Load Compensation is set too high, the chair
- Drives very rough
- Is hard to steer or control, vibrates or surges
- Swerves when it drives over a bump
- Motor becomes hotter than normal very easily, decreased motor life

Set Load Compensation to the correct motor resistance value of the used motor. To determine the correct load compensation:

DX2 - Use the Wizard Load Compensation plug-in (Tools -> Plug Ins -> DX2 Power Module Load Compensation).
DX - See the Load Compensation description in the DX System Manual for 3 step-by-step procedures to determine the correct motor resistance.

Note:
Set this parameter to twice the actual value for a PM-S single channel module (used on servo steer wheelchairs).

To test if Load Compensation has the correct value, perform a series of wheelchair tests (drive on a slope, over a bump, up a sidewalk edge, and over thick carpet) and check if the chair behaviour is similar to the correct behaviour described above.

[gcebiker]

So glad i have been posting the working code as i iron it out.

Long intervals between coding sessions has me forgetting which is the current build file.

- Hopefully i will be able to make a video in the near future with the Q6Edge chair - with Dynamic Shark Controller being driven via Remote Control (via the RS485 data bus)

Ive also applied for funding for a Jazzy 623
Mid wheel drive with Dynamic Controls, solid tires. It should easily pull the trailer around the house for me.

[gcebiker]

Ive converted the code over to Remote Control.

Benefits of longer range and a fail safe, its also able to use off the shelf RC units and not tied to one brand.

This is a copy of the working code,
Dynamic Shark power wheelchair Remote - Arduino Emulator, with 2.4ghz Remote Control , large dead band and fail safe.

[gcebiker]

About a month ago, I did a bit of work for a guy on two of his chairs , with payment to be an 8 year Jazzy 600....
After the work was done, he changed his mind and said he would keep it, i was disappointed he had changed his mind....the Jazzys not even going, one motor is apparently blown and the batteries are long dead but what can you do right ?

This afternoon i get a text from him asking me how much i would like to PAY for it...with no offer to pay for the work i did on his 2 power chairs...grrr

Anyway, while that texting was going on. I did a short video of the new RC emulator driving about in my driveway.
When it finishes uploading this is the link to it, https://www.youtube.com/watch?v=6x7QHqm ... e=youtu.be

If anyone has any questions could they please post here so that the community may benefit (rather than the PM's please)

If anyone has the experience with making circuit boards could you please get in contact with me.

I would really appreciate someone drawing up the boards ready to go to a board house. I have run out of the stimulant that i have been using for 8 years (it was a banned supplement years ago but i stocked up) and working new information into my tiny head is increasingly difficult.

In return i would gladly purchase the boards and build/test one and post to you for your own use.

Cheers
Tony

[woodygb]

Tony is this interface instead of a joystick/pod ...or will it work in conjunction with an existing pod ?.

[gcebiker]

Its a complete replacement but just add a DPDT relay/switch and you could use it inline with the regular remote.

[woodygb]

A question about the deadband.
Why so big.... and doesn't the controller also take care of deadband?

[gcebiker]

A question about the deadband.
Why so big.... and doesn't the controller also take care of deadband?

I see other people using it other than myself from time to time and persons not used to power chairs tend to be a bit heavy handed.

Its got both a large deadband and a constrained max.

This is the controller emulator, the main controller is disconnected from the Power Module.
Well its now an aftermarket controller i guess.

[Burgerman]

A big deadband makes it horrible to drive. 5% is OK. 10% starts to feel a little odd. Total.

[gcebiker]

Ive posted the code so anyone will be able to change the values of anything they wish to.

To modify the code download the free Arduino IDE and tinker away (for people wondering how to edit the code, ie people new to this)

Entry level Robot Wars people will be able to use the donor wheelchairs Power Module, should cut costs a bit when starting up / tinkering.

[gcebiker]

Tony is this interface instead of a joystick/pod ...or will it work in conjunction with an existing pod ?.

The code could also be altered to make this remote work as an Attendant joystick, both could be attached to the Shark Bus at the same time.

[gcebiker]

Video finally finish processing, https://www.youtube.com/watch?v=6x7QHqmzDK8

[woodygb]

It would be simple to add in a few lines of code and have a pin read the 0v-5v output of a 20 turn trim pot, this reading could be used to allow the user to vary the deadband say from 10 to 260.

0v = stipulate 10 as the set coded MINIMUM so 10+0
5v = 10+250

[gcebiker]

M1 Roller wont talk to my Shark Emulator , not such a big surprise but i had my fingers crossed.

Pic of the test rig for the Arduino GPS controller, unfortunately the Brushless trolling motor turned up today and it wont fit, much larger shaft size.

I am still hoping i can make use of it with a pulley or push rod.

I am going to look into getting a mount made on the chair i use daily to tow the boat, as in interim to hopefully getting a Jazzy 623 on government grant funding i had had approved.

Or perhaps the local fella with the dead Jazzy 600 has a change of heart / agrees to a trade.
I am not keen on buying it given its not working for him and has been looked at by the 'Pros' , they said it had a broken motor and if that bears true they are not cheap to replace....