Dynamic Controls Power chair, RC Boat trailer project

[gcebiker]

In preparation for the RC Boat Trailer...

Revisiting the Dynamic Shark Joystick emulator board i built up, took a while to find the right 'arduino sketch' so found it and the board is not working.

...enter the Logic Analyzer, legendary tool..we should all have one.

Turns out i had a ground fault, seeing the data stream on screen i could see that there was no joystick input.
Logic Analyzer led to me to testing with multi meter.

A short time to find the problem and ill rewire and retest in the morning, then onto the next phases.

Ground fault ...I FOUND you...hahahaha


Code for the above.

This image is not technically correct for the code following.
D2 is the on/off
D-D10 Max485 pins
....nearly there but you get the idea.




Code

SR_Emulator_joy_in_shield_software_serial.zip

(4.57 KiB) Downloaded 855 times

[gcebiker]

sorry if pictures to big, i have a 32inch monitor. If i need to resize can someone let me know and ill do that.

[Burgerman]

2x 30 inch dell 2560x1600 desktop monitors here... So the picture is somewhat miniature like a playing card.
Even my laptop with a small 13.3 inch screen has a massive 3200 x 1800 pixels! This is 2016!

[gcebiker]

Fixed the broken wire and then found that a few of them are suspicious.

I will need to order some silicone insulated wire instead of PVC, the insulation's so stiff its causing the wire to break where it solders to the pins.

Minor changes to the code, i will need to add some PID type stuff to this as well.
The joystick is a bit savage when it kicks in.

Code: Select all

 {
      directionPotVal = analogRead(directionPot);
     
      directionPotVal = constrain(directionPotVal, 1, 1023);// Constrain to real numbers. 0 and 1024 are out of range for the controller.
      direction = map(directionPotVal, 0, 1024, 1, 1023); //here in case direction is inverted in wiring
      if (direction > 312 && direction < 712)                   // set dead zone of joystick
        direction = 512;
    }
    {
      speedPotVal = analogRead(speedPot);
      speedPotVal = constrain(speedPotVal, 1, 1023);    // Constrain value to real numbers.
      speed = map(speedPotVal, 0, 1024, 1023, 1 );  // here in case of speed is inverted in wiring
      //speed dead zone mapping
      if (speed > 312 && speed < 712) // set dead zone of joystick
        speed = 512;
    }


[LROBBINS]

Rather than PID it sounds like you need some "filtering" of the JS. First, you really should average a number of analog reads made one right after the next just to get rid of noise. Second, if you are using the Arduino's ADC to read more than 1 pin, it's a good idea to discard the first reading when switching pins as there can be carryover of charge in the condensor used in the ADC. Third, consider reading the joystick through n successive loop cycles and doing a rolling average to smooth out the change in value over time (this is actually equivalent to the I term of a PID). If you want some code snippets for any of these, I'll be glad to send them to you.
Ciao,
Lenny

[LROBBINS]

Also, I don't think that more flexible insulation is going to keep your wires from breaking. It's the soldered wire, not the PVC, that's brittle - it becomes like solid wire and won't handle vibration. What you need is strain relief to keep the soldered connections from moving. What I find most reliable is to drill a hole nearby large enough to pass the insulation, feed the wire up through that to make a fairly large loop, solder the end to the solder pad or pin, and then pull most of the loop back through the hole so some insulated wire lays flat against the board. Where space is not available, or when, as is usually the case, I haven't planned things well so I don't have those holes, an expedient is to route several wires together along the edge of the board and then use goop to hold them together as a bundle and bind that whole bundle of wires to the board. (It will never meet good manufacturing practices rules, but it works.)

[ex-Gooserider]

Agreed, strain relief is the problem, not the insulation as such... Soldering stranded wire makes it into solid, which is prone to bending fatigue, just like one can break a paper clip by bending it repeatedly....

As Lenny said, the solution is to get rid of the potential for vibration at the joint by anchoring the wire close to the joint so that any movement is only in the wire itself, not at the joint....

That said, I really like Teflon insulated wire when I can get it... It's harder to strip, but much nicer to solder, as the insulation does not melt.... It is also pretty resistant to abuse, and the insulation is very slightly thinner making for smaller bundles....

ex-Gooserider

[LROBBINS]

Aviation supply houses will usually have teflon insulated wire, but be prepared to pay "medical equipment"-like prices.

[c500user]

Over the years I have repeatedly bought silver plated stranded copper wire with teflon insulation from Apex Jr (a US surplus store).
Apex Jr has different gauges and colours in stock (stocks vary and website is not always updated so you have to ask which gauges and colours are available). Prices seemed OK and nothing like the inflated "medical" prices we are used to
The owner, Steve Slater, will answer any questions you have and can ship world wide. He will even check to see if things fit in a USPS postage-prepaid box so he can quote you exact shipping costs.

http://apexjr.com/wire.html

[LROBBINS]

Thanks, c500 - that's good to know.

[gcebiker]

It will be ok once it has some physical support and its in its box.

At the moment, i am moving it about with no strain relief, the silicon insulated multi strand is not suffering its the cheap JST Extension leads i am using that are breaking at the solder joint. I have put on a JST plastic cover, they provide strain relief at the solder join to the pins.

Did some testing today before putting the motor on the boat, other than a 'solid' PU tyre that had a flat spot, it all went well.
The Roller M1 is going to need some tweaking, profiles are a bit tame for reversing but forwards works a treat.

I think i will have to move the axle back a bit, its been made that is completely removable for tweaking.



Remote is likely to be the Dynamic Joystick Emulator i am tweaking with a regular Hobby 2.4Ghz + LOS module attached.

[gcebiker]

Rather than PID it sounds like you need some "filtering" of the JS. First, you really should average a number of analog reads made one right after the next just to get rid of noise. Second, if you are using the Arduino's ADC to read more than 1 pin, it's a good idea to discard the first reading when switching pins as there can be carryover of charge in the condensor used in the ADC. Third, consider reading the joystick through n successive loop cycles and doing a rolling average to smooth out the change in value over time (this is actually equivalent to the I term of a PID). If you want some code snippets for any of these, I'll be glad to send them to you.
Ciao,
Lenny

The power base takes care of the smoothing in its code, so for now RAW reads will do. Its not going to be analog input when i switch over the RC gear.
For now i think i will just map the values differently, I'll work on the code and post it up as i make changes.

Something like (literal code not actual code...)
if speed is more than the Dead band values then map as if the dead band did not exist, else map per dead band.

[woodygb]

I'd suggest mapping both sides of the deadband .. so zero/start will be immediately after the deadband for all 4 quadrants.

[gcebiker]

Yeah, i know what you mean, thats how i had it on Sunday

Code: Select all

     {
          directionPotVal = analogRead(directionPot);
         
          directionPotVal = constrain(directionPotVal, 1, 1023);// Constrain to real numbers. 0 and 1024 are out of range for the controller.
          direction = map(directionPotVal, 0, 1024, 1, 1023); //here in case direction is inverted in wiring
          if (direction > 312 && direction < 712)                   // set dead zone of joystick
            direction = 512;
        }
        {
          speedPotVal = analogRead(speedPot);
          speedPotVal = constrain(speedPotVal, 1, 1023);    // Constrain value to real numbers.
          speed = map(speedPotVal, 0, 1024, 1023, 1 );  // here in case of speed is inverted in wiring
          //speed dead zone mapping
          if (speed > 312 && speed < 712) // set dead zone of joystick
            speed = 512;
        }


It may not even be an issue when i go over to PPM code to feed the Speed, Direction.
Certainly the Wii Nunchuck wireless version performed smoother than this wired version.

Probably due to the ADC multiplexing.
At one point i was looking at using a PixiHawk and Mission planner but the setups a pain, especially if i want park the trailer somewhere different and dont have the GPS way points mapped out already.

[woodygb]

Could you not go with something like this?

if (directionPotVal >(512+Deadband))//deadband defined as 200
direction = map(directionPotVal, (512+Deadband), 1024, 511, 1023);//high
else if
(directionPotVal <(512-Deadband))
direction = map(directionPotVal, 0, (512-Deadband), 1, 511);//low
else
direction=512;

[gcebiker]

Your idea is certainly much better, thank you Woody...i was trying to do the same but could not figure it out.

[gcebiker]

Thank you Woody

So smooth now, also fixed the wiring up so that i did not need to have the mapping reversed...that was making it more confusing
Depending on how much time the boat takes up, i hope to get a handle on the making of my own boards for a more robust build.

[gcebiker]

Copy of working code for anyone interested, fritzing wiring diagram is at the start of the thread (i cant remember if the pins are correct for the trigger...i think they are tho)

SR_Joystick_input_working_deadband.zip

(4.72 KiB) Downloaded 745 times

[woodygb]

Your pic now shows two MAX485's ...the earlier diagram used a DG419?!

[gcebiker]

Completely forgot, sorry folks.

The fritzing did not have models and i just used the available ones with the right number of pins as a wiring diagram for myself as a guide.

I have done a bit of MS paint editing to the picture to help clear it up.
The pin headder is also wrong (its too long for an arduino Nano V3.0 by one pin...), i will eventually figure out fritzing and get some boards made up.
Project is in the assembly stage with the boat and its prawn season...in a bit of a rush to get the boat in the water.
Motor is being put on today, then next week i have to do the forward steering and control gear fitting.

[gcebiker]

...boat not going anywhere for a while.

Marine shop told me i needed A 'Long Shaft' motor, which didnt fit, so then he told me 'ah well, you will need to order a new transom' ...three months down the track after it took two months for the transom to swim here from New Zealand...and we fitted it up today and now all the other bits dont fit.

...hangs too far out the back.
Steering wont connect

I would of thought that getting in a Marine Dealer would have been a smart thing to do...
I am hoping that he is the nice guy that he seems and will, with any fuss, exchange the motor for a short shaft variant.
After all he was the one that told me i need a Long Shaft when i clearly don't.

Yes i checked before going to the dealer, i thought i needed a short shaft. I had measured the boat per directions and said so to him.

No No No, he says you need a long shaft, ....
Me : U Sure ? i measured it and it needs a short shaft.
He : No you definitely need a long shaft, the Rep from Zego boats was just in here telling me all about them....
Me : oh ok, here is my life savings for a New motor

Just venting, they are closed for the weekend (why a boat dealer that is right next to a popular boat ramp is shut, on a Saturday, is beyond me)
Lets hope he is able to exchange it Monday, I only took it out of its wrapping today.

[Burgerman]

You made the classic error. You presumed the 'expert' knows what hes talking about.

Definition of presumption:
'Mother of all fuckups'...

[gcebiker]

Wired in the 2.4Ghz RC receiver today.
Zip tied it to a plastic 'bolt' container to hold it all in place.

Analog input has been left wired in ...just in case i need it.... maybe i drop the RC unit in the water by mistake , Ill have to put a lanyard on the RC transmitter because dropping it in the drink is a definite possibility.

[Burgerman]

I might suggest a better rx with multiple diversity, esp at low level and around a metal wheelchair. Many things create null spots and multi-path issues on 2.4. I found that from experience of using RC in a house on my chairs for many years. Its very intermittent where there's cars or buildings or lots of WiFi/Bluetooth in the street. Use hopping system like DSMX...

I use an AR10000 spektrum with 3 satellite receivers as a single antenna/rx alone often causes problems around vehicles etc. With at least one mounted high in the fabric backrest. http://cdn.shopify.com/s/files/1/0393/4 ... 1447597453 2 satt's shown. 3 allowed.

I also use a 2 watt WiFi booster to increase 100mW output by 20x plus... I get total reliable safe control at a fair old distance at street level this way. Because a 1 second fail-safe is a long time near water! Or close to expensive cars or roads.

http://www.wheelchairdriver.com/12x.htm

[gcebiker]

Depending on where you live the analogies might not be accurate... i am only driving a very short distance, less than 20meters at the maximum, similar to driving to your letterbox from the garage door and back.

Its in a small park with only a small kayak ramp.

Picture below if from google maps satellite view, if i was better at crawling i would do that instead...


I dont imagine i will be going on my own...ever, lol. This last week with people actually seeing what i have been trying to describe, they have become very positive about it all...most offering their time to 'drive' the boat as i don't have a license as yet...which i will get.

LOSS of signal the trailer will just stop, and corrupt data will also halt the chair.

Driving it with the chairs own power base gives built in protections, if there is a fault in the serial data stream for any reason...it just stops.
And will not start again till the Joystick is power cycled.

ITS MUCH safer than control via the Analog inputs.

With a walking capable helper i am sure if any thing happens ...which will just shut the chair down, they can walk over and drive it close enough so that i can get in and back the trailer myself.

I have been sailing for 12 years since the brain injury accident and i am very aware of having multi back up systems in place.
I have a very keen bunch of fisher friends. Best back up a guy could ask for !

[Burgerman]

Depending on where you live the analogies might not be accurate... i am only driving a very short distance, less than 20meters at the maximum, similar to driving to your letterbox from the garage door and back.

I am flying planes and quad copters miles away over a town full of Wi-Fi and phones with Bluetooth... Thousands of them! Faultlessly, with DSMX channel hopping, 4 separate receivers in different positions and planes, 2000mW RF power.
But I have had powerchairs running on for 1 second before failsafe kicks in and stops them in the house, and around vehicles because of all the metalwork/reflected signals. At 10 paces away. So I do worry! That 1 second doesn't sound like much but it seems like an age when the chair is heading towards a road of your van... Or climbing up a kitchen unit!

[gcebiker]

Marine shop owner says to me today ; 'Ill see what i can do and call you back after lunch today'
...its midnight and still no call, i wonder when he has lunch.

Took the Roller M1 down to the ramp today to check heights and power.
Needs to be raised 245 mm or it hits the ramp (it has a decline, then flat, then another decline)

Chair is GUTLESS, i didnt change any settings (other than to enable the 100% speed , profile 4)
I will need to look into if the power can be improved upon, it wont even back up or turn if i have the trailer hooked up.
Pulls ok tho on the flat, with the trailer and boat hooked up.

No tests as yet with the motor attached till i hear back from the Marine shop.

Worse case , i have to change from linkage steering on the boat to Cable or one of those RC units i have seen on Kicker motors in the USA.

[ex-Gooserider]

Personally I've never liked working with Fritzing - even when using breadboards....

If you want to get into making your own, I'd look seriously at KiCAD - free software, specifically designed for electronics drawing, starting from a schematic and going to complete multi-layer board layout... A lot of the development work on it is done and / or funded by the atom smashing folks over at CERN.... We have several professional board designers at the Asylum, and they tell me that KiCad is comparable in functionality to some of the multi-thousand dollar / seat commercial packages like Altium....

A lot of people like the 'free' version of Eagle, but if you start on it, you will soon find that it has a lot of limitations on board size, layer count, and so on that are designed to get you to buy the 'upgrade' to the commercial version... KiCAD effectively has no limits - or at least not that are more than what you can get made at the typical board house...

Also once done with testing, I'd be wanting to get the hardware into a more robust box, both for strength and for environmental protection...

ex-Gooserider

[gcebiker]

Personally I've never liked working with Fritzing - even when using breadboards....

If you want to get into making your own, I'd look seriously at KiCAD - free software, specifically designed for electronics drawing, starting from a schematic and going to complete multi-layer board layout... A lot of the development work on it is done and / or funded by the atom smashing folks over at CERN.... We have several professional board designers at the Asylum, and they tell me that KiCad is comparable in functionality to some of the multi-thousand dollar / seat commercial packages like Altium....
ex-Gooserider

Thanks, ex-Gooserider. Ill be definitely checking out KiCAD.

Especially now as i am working on the GPS shield for upgrading a regular trolling motor to GPS 'Anchor' Style and in the last few days built up a little Arduino DigiSpark85 as a reef tank fan controller. I would love to be able to offer people the board house plan/link to make their own for this and other arduino shield projects.


Those of you with excellent eye sight may notice i have some wiring issues in that image...they have been fixed and i added the missing Pull Down resistors and a PWM cap for the fan...i had a very bad day .

Mobility scooter not quite working out at the moment...
A, Jim Slade from Slades Marine at Paradise point insisted i needed a long shaft...even after i measured my boat and I thought i needed a short shaft. The steering is not linking up (its fixed linkage steering), tho i am only finding out the problem now...he had known about it for months and had told me to "just buy a new transom" ...
NOW its Christmas and Aussies want to go to sleep on the job as it is...Christmas its 10 times worse.

B, The mobility scooter with LiFePo4 batteries...is NOT heavy enough...an issue when you are towing a 200kg boat and trailer up a boat ramp !
...so lead bricks going back in eventually but for now, parts have been purchased to turn the Mobility scooter/ Boat Trailer into a....Boat Trailer for a CAR

At the moment this will mean a friend parking the car / trailer for me.

...Later on however, I will remove the seat from the Mobility scooter and relocate the tow ball to over the center of the chair...and give it another go

Merry Christmas to all.

[gcebiker]

Dealer eventually took the motor back , that he had ordered in error.

And he only charged me $320 for pay for his mistake...plus the $300+ on the transom i also got following his advice ..after he had seen that it was the wrong motor
4 months ago !

Today i purchased a NEW 6hp 4 stroke Yamaha Outboard with 12v 6amp charging for the Zego and the trailer is being changed into a full Road Register-able Trailer.

I had to up spec the trailer, Main Roads over here did not know how to register a mobility scooter boat trailer and the insurance companies were even more in a spin.
So now its about spec'ing up the M1 Roller with lead acid batteries (LiFe are to light). and seeing what settings i can change to up its power, its a funny thing to drive.

Plenty of go when going in a straight line up quite steep hills but if i try to reverse or turn (on level ground) it gets stuck on its own front castors, Ive not looked into the settings as yet (apart from turning on profile 4). I am sure it will be straight forward to clear up.

I will put the LiFePo4 into the boat, 2 x 50ah LiFe will come in handy with the trolling motor expected to be in use quite a bit.

Just downloaded KiCad, ill post up a decent schematic if i have any success with it.