How do chairs know their speed?

[slomobile]

For Rnet systems specifically, or any other system that knows actual wheel speed and/or odometry, How does it know?

There aren't wheel encoders as far as I know.

Is it just estimated from voltage and current?

My best guess is that the power module senses the changing impedance at commutation breaks, and they know the number of commutation segments per rotation. But I'm not sure.

[Burgerman]

Basically joystick position...

Actually Joystick position plus programming as a voltage (a pulsewith). Voltage = rpm. Amps dont enter into it. In fact its the pulsewidth the system generates from your programming and the joystick position. So includes delays for acceleration etc. Does not include motor load compensation. So not actual motor final voltage.
And YOU tell the R-Net the chairs design speed. Say 6.2mph. Then full stick = 6.2mph. And the display ark top number you set to say 6.5mph.

So full stick = whatever you tell it is the design speed... You dont even need a motor attached. Just a light bulb or resistor.

If you drive up a hill the chairs speed is slower than the display. If you drive downhill its faster than shown.
If you drive up to a wall and try to push the wall. The speed simply shows your joystick position. 6MPH CHAIR? It will show 6mph even though you are not moving.

[slomobile]

What a dirty trick! My chair tops out at 6.4MPH indicated and I was impressed that it held that even going up hills. I assumed it was a measured speed (not commanded speed) because it wouldn't reach 6.4MPH indicated when the batteries are low. By using measured battery voltage in the algorithm they maintain the illusion. If only I had tried pushing a wall I would have figured it out. Each motor has a no load current at a given voltage, so if that voltage is applied, the excess current above no load is split between propulsion and heating. So if we had an accurate thermometer attached to a single winding, could we use that that to subtract the heating current and determine the speed? Edit: no, forgot about iron losses

[Jayde1976]

I figured they used voltage..as mine will do 7mph downhill and 6.2uphill and same on the flat until the b atteries are halfway flat then its 5mph uphill

[Burgerman]

That is actually slower uphill when the battery is flat not because of state of charge/battery voltage specifically but the voltage UNDER LOAD. So a lower impedance battery like the Odyssey, or lithium shows this drop to a lesser degree than say a cheap chinese AGM. Its also why a lower Ah battery with a low impedance still gives the same or better range.

If you tell it the chairs its design speed it say 20mph, thats what full stick will display... Its basically bullshit. If it makes you feel faster! If you want true speed use GPS and go at least a 50 metres as that will be way way more accurate and true.

[Fred005]

If you drive up a hill the chairs speed is slower than the display. If you drive downhill its faster than shown.
If you drive up to a wall and try to push the wall. The speed simply shows your joystick position. 6MPH CHAIR? It will show 6mph even though you are not moving.

I don't know for all make of power wheelchair, but for the X8 on a same ''speed gear'' selected on the joystick, with the js in a same position, my speed is exactly the same I can read on my Garmin. So when climbing, speed is true speed, not designed speed.
Once again, on the X8.

Good evening.

[Burgerman]

The X8 is a chair. The CONTROL system is R-Net right? It doesent know what "chair" its on. Works exactly as I described. Try pushing a wall. It will show you moving at X mph. Try turning in place, and same thing... You GPS will show you are not (obviously). And difference between loaded speed and free running speed, or as calibrated in programming gives a percentage of error.

[Fred005]

My, and many, GPS, is really very acurate, for the speed. And why ? I've a sensor on a wheel counting wheel rotation.
And even if the GPS will have a few % error, it'll not be enough to don't match with the X8 speed.
For exemple, today with my handbike I climbed Mont Ventoux.
21.2 km on the map, on the milestone, on every where. 21.45 on my GPS. Is it accurate enough for U ?

And, when with X8 I climb a more than 45° hill, be 100% sure my js is full forword. And the indicated speed 2.5/3 km/h.

Also, if I lift the 4 wheels, and push the JS full forword, the indicated speed is 13/14 km/h, as with no weight wheels rotation increase.

That's all.

[Burgerman]

Try the wall test and you will then understand. Then explain how a display of xx mph while the chair isnt actually moving works!

Yes your r-net must be different to all the rest. It can actually have a motor/wheel speed sensor input added. Does yours have this?

As for GPS it can be extremely accurate. Unless like expresso you live in a big city where few satelites can be seen. Then its not. Also try driving in circles for 30 mins. The GPS will read low distance compared to actual distance. Its averaging your position over time.

[Fred005]

''Try the wall...''
And I brake my 2 fuc..ing feet ...
Not necessary to do with a wall, as I said U, try a 45° climb, and if you can, you'll see your JS full forwoard but your speed not a lot.
If you try with this 45° climb, your feet will stay safe.

[Burgerman]

Heres how accurate GPS is.
Right now because of terrain, rf noise, my TWO GPS usints both think I am in my neibours garden. I am in fact in bed...

So over long distances in straight lines especially they are extremely accurate. Where you path isnt straight, where there are reflections and multipath distortions (changes the time it took to arrive) or where theres a lot of RF noise from say computers or whatever it gets very innacurate for positioning accuracy. And it uses that position to calculate distance covered over time.

Here its got the averaged longditude and lattitude within only 8 meters. And altitude miles out.

Look for yourself.

[Burgerman]

''Try the wall...''
And I brake my 2 fuc..ing feet ...
Not necessary to do with a wall, as I said U, try a 45° climb, and if you can, you'll see your JS full forwoard but your speed not a lot.
If you try with this 45° climb, your feet will stay safe.

Yes it is nessassary. As it will prove that I am correct and that your claim makes no sense. REVERSE into one. A steep slope may show that the controller rolled back power to protect itself instead.

Or just attempt a zero speed turn in place. and then as you move the joystick forwards a little the thing will display speed while you are not moving. It doesent know you are not moving, or what speed you are actually doing. It only cares about how far you push the joystick. Not load. Not actual movement. Go to a curb. Push the stick a little but not enough to climb. It will show at least some MPH. And you are not moving. Not exactly hard to see this. It has no other way to determine speed anyway.

Anyway you can do this with an empty chair. If I want on my bed I would do a vid of this whithout even getting into the chair.

[Fred005]

Ok, ok you're right, I'm wrong, a wrong idiot.
Did U read me BTW ?
NO
''My, and many, GPS, is really very acurate, for the speed. And why ? I've a sensor on a wheel counting wheel rotation.''

This morning, as I said, Mont Ventoux with handbike, from Malaucene: Every body knows it's 21.2 km
I warmed up 2.8km.
So, 21.2 x 2 (climb and back by exactely the same way) + 2.8= 45.2 km.
Garmin say me 45.61km, Delta less than 1%. (A tip, change your GPS or update it...).
So, when I climb with X8, full forwoard joystick, and see my indicated speed @ 2.5 or 3 km, it's ... Far more than 1%.

But as usual, your ''right'', no problem.

[Burgerman]

Did U read me BTW ?
NO
''My, and many, GPS, is really very acurate, for the speed. And why ? I've a sensor on a wheel counting wheel rotation.''

Yes I read it several times. It made no sense and I had no idea what it meant or refered to. GPS has nothing to do with wheel sensors. And is the wheel sensor enything to do with R-Net? Is it connected? (it can be) You didnt say so. I didnt know what to make of it and still dont. I know you are french so presumed a translation thing.

And yes GPS can be very accurate for speed. But not always. Many things can throw it out. Such as multipath signa errors as I just demonstrated. And yes my handheld garmin and my PC with built in Ublox GPS are new and fully up to date. And both agree I am somewhere else!

However the fact that this place you refer to came out with the r-net matching the GPS is not relavant is it. I could guess the distance and be wrong or correct. Next time on a different destination it probably wont match. Because we know the r-net is not and cannot measure distance. It has no way to do so. It matched this destination. So what? Its plus and minus errors on this one destination cancelled each other out about right. Again thats not the point is it. It simply CANNOT determine distance by any other means other than joystick position plus some maths and presumptions. Which cannot ever be accurate. At least not repeatably or in different situations. It does not measure distance at all. It measures the stick position and presumes a speed based on the vehicles max that you program in. This approach "works" in a fashion. Thats why they do it. But has serious issues. And I explained how you can prove this. But you go on about your feet.

[biscuit]

Fred do you have an RPM counter on a wheel shaft? I assumed that would not work for a wheelchair because it has left and right wheels.
I can't see any way for it to deal with anything other than going straight forwards.

[Burgerman]

You can have one, or two, and average the data. But I still have no idea what this has to do with the way the r-net knows the chairs speed. Which was the original posters question.

Re: How do chairs know their speed?

To which I gave the correct answer already. Its just joystick position and includes your programming inc acceleration and speed settings. So if set to profile 3 which is a 3.5mph max, thats what it displays at full stick. Even if you are not moving. Half stick is shown as half that speed. Everything else is not relevant.

[Fred005]

OMG ...
''GPS has nothing to do with wheel sensors''

Well ... today, I mean in '21, GPS means both the Global Positioning System, AND the device.
So when I'm speaking of my GPS it's my Garmin, as I wrote many time.
My Garmin can work, for the speed not position or altitude but I thought you know that, with GPS satellites or, if too obstacles are providing a bad reception, with a speed sensor (visit Garmin, for exemple, web site....).

@biscuit:Fred do you have an RPM counter on a wheel shaft? I assumed that would not work for a wheelchair because it has left and right wheels.
I can't see any way for it to deal with anything other than going straight forwards.
Yes, I've this:
https://buy.garmin.com/fr-FR/FR/p/pn/010-12843-00
And yes it's working fine.

You could be ''right'' if I was only doing ''donuts'' during many hours.
But during my hikes, one time I go straight, an other I turn left, or right.
Width of X8 is 70cm, so even doing donuts will give a small error.

In any cases all what I'm saying have been checked with, for exemple a friend hiking with me (but ... He's wearing is GPS watch always on the left arm, so ... ''I can't see any way for it to deal with anything other than going straight forwards.'' Lol !
Or other friends with a mountain bike, or cheking with topo maps.

In 1 word I don't have tons of electronic device at home, as some here, but at home I'm not often.
On the other hand, on the field, yes I'm. And when, again and again, when I'm climbing a hill, whatever it's slope, when my JS is full forwoard, I never see the same speed, depending of the slope.

After, you can think what you want, no problem.
I was just giving my ''knowledge''.

[terry2]

My speedo is in the back rims and it connected to the R-Net.

The GPS is spot on when using my Note 10+ 5G and my Galaxy watch 4 classic.

[Burgerman]

R-net has the ability to use an encoder. Which is implemented in the myra chairs. Albeit by a non standard myra part I think. But you can use a sensor or shaft encoder. In the case of the myra its used to help with the steer control with the weird servo steering and odd power module they use, I think. Not sure if its also used in speed display or distance covered though. Easy check, drive against a wall, see if it tells you some x mph.

[Burgerman]

OMG ...
''GPS has nothing to do with wheel sensors''

Well ... today, I mean in '21, GPS means both the Global Positioning System, AND the device.
So when I'm speaking of my GPS it's my Garmin, as I wrote many time.
My Garmin can work, for the speed not position or altitude but I thought you know that, with GPS satellites or, if too obstacles are providing a bad reception, with a speed sensor (visit Garmin, for exemple, web site....).

@biscuit:Fred do you have an RPM counter on a wheel shaft? I assumed that would not work for a wheelchair because it has left and right wheels.
I can't see any way for it to deal with anything other than going straight forwards.
Yes, I've this:
https://buy.garmin.com/fr-FR/FR/p/pn/010-12843-00
And yes it's working fine.

You could be ''right'' if I was only doing ''donuts'' during many hours.
But during my hikes, one time I go straight, an other I turn left, or right.
Width of X8 is 70cm, so even doing donuts will give a small error.

In any cases all what I'm saying have been checked with, for exemple a friend hiking with me (but ... He's wearing is GPS watch always on the left arm, so ... ''I can't see any way for it to deal with anything other than going straight forwards.'' Lol !
Or other friends with a mountain bike, or cheking with topo maps.

In 1 word I don't have tons of electronic device at home, as some here, but at home I'm not often.
On the other hand, on the field, yes I'm. And when, again and again, when I'm climbing a hill, whatever it's slope, when my JS is full forwoard, I never see the same speed, depending of the slope.

After, you can think what you want, no problem.
I was just giving my ''knowledge''.

NON of that has anything to do with the question!

What your aftermarket GPS does is irrelivant. What aftermarket thing is fitted is irrelivant. Nobody asked what watch or wheel adapter you fitted.

The question was, as per title, "Re: How do chairs know their speed?" and "For Rnet specifically" which I answered correctly. They really dont. Its just joystick position.

Unless you have some other encoder or aftermarket garmin sensor or encoder like the myra version or whatever thing fitted. Which was not asked.

[Fred005]

NON of that has a thing to do with the question! What your aftermarket GPS does is irrelivant. The question was, as per title, "Re: How do chairs know their speed?" which I answered correctly. They really dont. Its just joystick position in almost all cares unless you have some other thing fitted.

I referred to GPS to show you you're wrong.
But, as usal, you ''can't'' be wrong.

So ... think what you want from your bed, I've better things to do.

You always want to be right so badly that you don't listen to anyone anymore, not even those who talk to you about concrete things.

In this case, I'm not to the right forum, even if some people (even you) helped me.

I ask you to erase all my post from this forum and to remove all my personnal account information, erasing too my account.

Kind regards.

[Burgerman]

OH. But I am NOT wrong. And no GPS data could show that. I can guess a figure and it be right according to GPS too. So the fact that you claim it is the same means nothing. And is NOT what the original post asked.

I have investigated this in detail. And I know that you can do as I suggested to prove that its just joystick position too. But you refuse. I wonder why. Its a simple test. However anyone that chooses can test this super easily for themselves. I can be wrong. But I am not. If ever unsure about anything I say so. Or shut up.

Its not about me never being wrong. Its about misinformation. And possibly your lack of logic. The information about how r-net tells the speed, which was what the original poster specifically asked, is 100% correct. That is how r-net works. Its trivial to test. I can push the joystick slightly forwards or at 45 degrees in my chair, not enough to move - say on grass or against the step on my staircase, and watch it read 2 or 3 mph. While I sit not moving. And so can you. But you refuse test this simple because it makes you look silly. You likely already did this but refuse to say. That adds to the distance being wrong as well as incorrect speed indication.

And no I will not delete all your posts. Because then all the replies and every thread you posted in makes no sense. Its not my fault you cant follow logic. Or refuse to check via a trivial test. Was it you that said you had brain injuries? Maybe this is one of the consequences. But you are unable to see that what your GPS says is irrelivant and proves nothing. Even a stopped clock is right twice a day. But my simple test shows that you are (logically) incorrect.

[Fred005]

So, what you mean is, whatever the slope, it's the joystick position who will indicate the speed.
Ok. If you say that ....
So for the distance it's the same .... Logical ... for you.

My God ... how stupid is your reasoning.
When I hike a ... 15km, 15 wich is indicated by every body, topo maps, guys who knows the area, every one, my total hike distance will be wrong because of my speed is wrong, as you like to say.

You've still a lot to learn about power wheelchair ...

[Burgerman]

So, what you mean is, whatever the slope, it's the joystick position who will indicate the speed.
Ok. If you say that ....
So for the distance it's the same .... Logical ... for you.

Translation problem?
Yes the joystick speed is simply joystick position inc your programming settings. Its the guestimated speed the chair would be traveling if no slope, ramp, turning, or against a wall. It measures your joystick position, and displays that incorrectly as speed. And it has no other way to measure distance either. So it calculation is the (nessasarily inaccurate speed reading) over time. That gives the distance measurement.

Thats how it ACTUALLY works. You might not like it but I can demonsrate to anyone that this is true. And so can you do the same thing yourself. But you refuse because you will now look stupid. And this was the correct answer to the question: Re: How do chairs know their speed? And "specifically R-Net".

The rest of your comments are just insults because you cannot understand logic or reasoning. Its just matts. The chair HAS NO WAY TO MEASURE SPEED OR DISTANCE properly. It doesent matter if it matched the map on that occasion. Its still never going to be a reliable speed indicator. It cannot be so. And so likewise cannot be a reliable distance indicator. Even if all the errors cancel each other out on this ocasion. Thats YOUR failure of logic and reasoning.

[shirley_hkg]

It drives by pulse modulation . Does it know the rpm of the motors ?

[LROBBINS]

Another chair that might actually be able to measure speed is the invacare gearless-brushless (or any other brushless such as Will's homemade chairs, for that matter). Their controllers actually do get RPM information.

[Burgerman]

Yes the invacare brushless dynamic could know that. I dont think its used that way though. I cant say for sure. No longer have one.

It drives by pulse modulation . Does it know the rpm of the motors ?

No the pulse is just a constant frequency wave, and is simply the pulsewidth in place of actual voltage. At a frequency above where your hearing extends to prevent you hearing it "sing". Often 18 to 25K. A 50% pulsewidth = half battery voltage for e.g. Its just a way to regulate the motor power by chopping on/off power at high frequency.

It has no idea if its connected to a light bulb, or a motor. Or a resistor. It has no way to know real speed at all*. It simply presumes that voltage = rpm, and that half the voltage equals half the speed etc. With when unloaded is approximately correct. Povided you enter the right chair design speed into the programmer.

*UNLESS the system has an encoder fitted. I never saw one, but theres a special encoder module that allows real wheel speed or motor speed to be used. https://www.cw-industrialgroup.com/Prod ... der-Module IF theres a sensor installed on each motor.
This is for accurate tracking with alternative controls. And I dont think its implemented to measure speed on R-Net display either. Certainly theres no programming settings that relate or would allow it to be configured for this use. And the gyro module basically is a modern way of doing this. Myra may use it to connect their own electronics. To aid with servo steering angles.

[slomobile]

So the question was about control systems that ACTUALLY know wheel speed. My purpose for asking was to figure out how to implement a control system for an articulating 4 wheel drive chair. It would require knowing position of each wheel down to about 1.5deg and writing a control law to be programmed into the controller. I do not have an Rnet dongle yet, but hope to afford it soon. I also do not know how to program anything besides parameters into Rnet joysticks and power modules. Does anyone here have a resource for that?

It seems that most chairs just show the commanded speed, so they are OUT of the discussion.
Any brushless system has to know their speed in order to do electronic commutation. So they are IN. Are there any Rnet brushless systems?

Systems implementing https://www.cw-industrialgroup.com/Prod ... der-Module are also IN.
Any info on how to use it?

Fred005, your X8 is IN. It makes sense that it would have encoders given its off road purpose, monitoring slip and grip could improve traction. If you find out any more about how it operates, I'm all ears. The odometry compared to GPS track was convincing data. If you have your chair's programming file available, perhaps we could learn something from it. I'm trying to understand what principles make for a good off road chair, and the X8 is one of the best examples I am aware of.

[Burgerman]

So the question was about control systems that ACTUALLY know wheel speed. My purpose for asking was to figure out how to implement a control system for an articulating 4 wheel drive chair. It would require knowing position of each wheel down to about 1.5deg and writing a control law to be programmed into the controller. I do not have an Rnet dongle yet, but hope to afford it soon. I also do not know how to program anything besides parameters into Rnet joysticks and power modules. Does anyone here have a resource for that?

There isnt one. You can only use the programmer to choose some parameters.

It seems that most chairs just show the commanded speed, so they are OUT of the discussion.
Any brushless system has to know their speed in order to do electronic commutation. So they are IN. Are there any Rnet brushless systems?

Only on some pug and play tiny folding chairs. And invacares now junked brushless system. It was useless. I sent mine back for a refund. Gearless. So needed 190battery amps to turn a chair in place EMPTY. Nedless to say it didnt have adequate torque and kept shutting down and murdering its batteries in my room just trying to move around. Worked great at very low current on the flat. Thats invacares brushless system.

Systems implementing https://www.cw-industrialgroup.com/Prod ... der-Module are also IN.
Any info on how to use it?

They decide what it does and how. Its really only used to help track straight with those that cant steer proportionally.

[slomobile]

They decide what it does and how. Its really only used to help track straight with those that cant steer proportionally.

https://www.cw-industrialgroup.com/geta ... t_brochure
contains a list of EXPANSION MODULES
Omni Color LCD Universal Specialty Input Device
IOM Input and Output Module
Bluetooth Modules Bluetooth Mouse Control for Windows, Android or iDevices
Encoder Module Motor Encoder Interface
ASM Accelerometer Stability Module
Gyro Module Gyro Stability Module
Development Module Development platform for R-net compatible third-party modules

That last one looks interesting. Trying to order it now.