Burgerman wrote:Put axis back to stock.
1st you must set the power cables to the motors to both go forwards when you push the stick forwards. Physically. Change polarity to the motor that goes the wrong way. Then, if it turns left when you push right, either swap motor connectors left to right on the module or choose "swap motors" in programming.
Burgerman wrote:Well you are going to have to. Or cut the cables and swap polarity and solder and heatshrink. Which is what I would have done from the start. Because thats the only way.
Burgerman wrote:Yes that will do it. Forgot that existed. What do you have motor load compensation set to?
In fact where did you get the current profile in use?
Send me, upload here, a copy.
So I can take a look.
Question, do you have normal hands, good joystick skills?
Burgerman wrote:Why do you have joystick throw set to 80%? In all directions. That will mean you are at max stick or 100% at part stick. Wasting 20% of the resolution.
Also it will never go where you tel it with those settings. Are you happy with it like that?
Burgerman wrote:OK. That will increase stick sensitivity. You also have super sluggish turn accel, turn decel, min turn accel, min turn decel. Mine are all set to 100.
Because otherwise the controls cease to be direct and a delay or acceleration is introduced. Meaning you cant hit a doorway. Like driving a hovercraft.
Set these 4 to 100. If the chair is then too sensitive for you then you should instead reduce the turn speed, and min turn speed. to match the feel of the chair at its minimum speed and at it maximum speed. Do that while stationary, zero turn. Left and right.
Then it will go left in the way a biro does. When you tell it. By the AMOUNT you decide on a directly proportional amount. When you decide that you have turned enough While heading for a gap like a doorway you want to be able to ai the chair by turning the right amount. And then centre the stick. If turn min and max turn decelleration are set to 100 then it stops turning when you centre the stick. If set to even 90% there a delay that makes it keep changing direction after you topl it to stop turning...
Also if a profile is correctly set up to work accurately from very slow speeds, to full speed, and the turn rates etc are all balanced you only need and wil only use 1 profile. All the rest of mine are now disabled. You simply select a lower speed rather than change profile.
I can make a best guess and do a profile called TEST on your file to try if you want. Then try that and see if your joystick capability is adequate so that it works well for you. You will need a proper joystick technique, not just arm on the top of the chairs arm and holding the moving part of the joystick.
You need to cradle the joystick side. And the joystick will need to be located inboard of the arm, with a gap to make this possible. If you cannot do this, then you will forever live in what I call hovercraft mode.
Hand like this: Ignore wheelies. Thats because forward acceleration is set to 100 as well. Watch how easy it is to hit a doorway 3 inches wider than the rear drive old chair, with a 20 stone 60 year old driver... The confidence is because its accurate. But less accurate than when set to 100 for these 4 settings.
https://www.wheelchairdriver.com/gopro/control.mp4 MOSTLY here look at hand position and joystick position. Without this you will never have enough control to use good steer settings. DOWNLOAD first before playing.
Burgerman wrote:You set all 4 to 100 already? Presumably you already found out then that to do so you need to remove the walls in the OEM factory section by setting every thing to 100 or 0...
Now you just need someone to check what the V6 or X5 motor compensation should be set to. 45 is safe. But it may want to be higher.
Burgerman wrote:We should have a database for motor compensation for various chairs.
Such as:
EMD 2 pole 6mph 100mOhm. Run away! Terrible things.
EMD 4 pole 6mph 45mOhm. Bomb proof, lots of torque, mine set to 50mOhm.
AMT 4 pole 8mph 45mOhm.
AMT 4 pole 6mph 45mOhm.
Linix 2 pole (Permobil 7mph?) 70mOhm from memory?
Linix 2 pole (Sunrise various speeds ? Never owned one so unknown?
Linix 4 pole 6mph 45mOhm (mines set to 50) lots of torque, good reliability.
Linix 4 pole 8mph (also called 13kph, 8.5mph etc) 35mOhm. Low torque, step too far. May be better if set to 40mOhm.
X5 and V6 inline motors 60mOhm, 6.2mph.
ANY other chairs stock figure, can let me know!
Will move this to the programming bit. I have many more to add soon.
Bounder ST4 motors 95mOhms 6mph. Yikes That's the stock setting with R-net
Burgerman wrote:Bounder ST4 motors 95mOhms 6mph. Yikes That's the stock setting with R-net
Yes they look physically big. But dont confuse this with power. That impedance level puts them at around the same level of torque x speed as a 2 pole linix or similar right angle motor. The chain drive is a bit antique but chains are pretty efficient compared to a gearbox, esp right angle ones. So it makes up for it a little.
Remember though that the ultimate limit of torque at stall is the 120A from the power module. Many motors would draw two or three thimes this much if they were directly connected to the battery. So for stall torque and low speeds the only reason a typical low impedance 4 pole motor performs better is that the motor load compensation actually works properly.
Why? Imagine a high impedance motor only draws say 100A at stall. And it needs this current to actually initiate the turn. If you decide to do a turn in place your joystick gets pushed FULLY left... That may be set in programming to be 25 or 40% (or less) as the turn speed. So now the controller sends a 25% pulse width. Or if you prefer a quarter of your battery voltage. At full left stick. Since the high impedance motor is only drawing 100A at 24V it now only draws 25A. And the chair wont turn... So now you rely on the motor load compensation with sees the current, and adds to it. By increasing pulsewidth behind the scenes. But it will if not set really high, only add maybe 50% so you end up with 50A to the motor, where 100 is required to start the turn. So the motor appears to lack torque and control.
If the motor naturally wants to pull high currents, because its low impedance, such as in the case of a 4 pole linix/AMT etc, which have stall currents between 300 and 380A then when you decide to send the same 25% turn signal (at full stick) the motor naturally wants to pull 25% of that 380A. So around 95A... And the turn starts straiht away as you wanted. In reality theres also cable resistance thats similar to motor values, and voltage drop due to battery resistance and mosfet resistance and so you still need motor load compensation to help. But thats how it all works.
Burgerman wrote:Yes its amost plausible! Maybe not quite. But close. It may be that the most important thing here is that they will draw as much current as the controller allows regardless. But thats now only the 90A unit. So the static stall torque will be less and the lower gear losses maybe increase that again by maybe 10% or so. A 120A module would mean that they could give greater torque than a 90 degree motor. If the motor load compensation worked as well and was configured properly.
They POSSIBLY had issues with the accountant. OR they found that a 120A module overheated or damaged their motors in some way when under heavy load off road and so decided the 90A cheaper module was enough and saved on warranty costs and purchase cost too.
Return to Everything Powerchair
Users browsing this forum: Bing [Bot] and 85 guests