Wheelchair inefficiency

[JennyB]

Welcome to ES, Burgerman!

24v poweerchair. Takes about 20 battery amps to turn on the spot. Measured on a thin carpet. Because the casters are fat, have a fat bloke sat on top! And remember that this is happening all the time. You are using motor power to steer or even to correct your course all the time. In a car or bike the motor eitther pushes or pulls. It doesent spend its time doing both manoevering and going nowhere, wasting amps, but not asctually covering distance.

Try and turn a power wheelchair set in free wheel mode on the spot with an adult sat in it and you will get the idea. Its bloody hard work! Think amps! When you were going forwards at max speed in a straight line, and on a LEVEL surface it is using the least power. Now add steering corrections (eats chunks of power as iy turns your mass directly) and add a little left to right slope. This is the killer. And power consumption goes through the roof.

Driving along the edge of a road SLOWS the chair, and can take 60 to 100 amps. Remember that the casters have say 1/3rd the weight. And want to naturally head to the gutter. To stop this, while traveling in a straight line you push really hard with one wheel, and at times even pull with the other. To "twist" the chair against gravity to travel straight.

Thanks for explaining that. I wondered if a Segway would suffer in the same way, having the weight more centred over the wheels. It seems, though that the X2, which has about the same performance and use as your chair, has a very similar power consumption. I guess the active balancing makes up for the lack of castor wheels! :x

 

[amberwolf]

Thanks amberwolf. 29.1wh/mi is still a horrible number at 15mph, but leaps and bounds better than the 58wh/mi at 3mph I experienced on my BIL's wheelchair. I looked at your above mentioned blog trying in vain to find the modification you did to the wheelchair motor for adaptation to your bike. Can you point me to the specific blog entry for that?

There's a lot of them, becuase it was an ongoing experiment. But some of the versions of the drivetrain are in my early posts here on ES, in the CrazyBike2 thread:
http://www.endless-sphere.com/forums/viewtopic.php?p=188579#p188579
and some posts before that link to blog posts about it.


Here's some posts describing some problems I had with the system and some solutions
http://electricle.blogspot.com/2009/11/chain-reactions.html
http://electricle.blogspot.com/2009/11/chain-reactions-part-2.html

previuosly, when i changed to the 4pole motor for better torque
http://electricle.blogspot.com/2009/10/technical-difficulties-mark-ii-4-pole.html

An older change to the drivetrain
http://electricle.blogspot.com/2009/08/drivetrain-v31-with-bench-test-video.html

I assume you have (at the minimum) disabled or removed the internal brake, correct?

Yes, I removed the brake completely (partly to drive something else from that end instead, like a fan, but I never got around to that). I've also removed it from the brushless motor I have now, but that one has no solenoid, just a manual lever. In the brushless motor, it is actually why it was tossed out--the brake didn't hold on slopes correctly anymore. (seems dumb to throw away a whole very expensive motor because of a simple pad or disc wear issue, but apparently it's commonplace, because insurance and/or medicare pays for it--and you wonder why costs are so high?!).

 

[Burgerman]

Thanks for explaining that. I wondered if a Segway would suffer in the same way, having the weight more centred over the wheels. It seems, though that the X2, which has about the same performance and use as your chair, has a very similar power consumption. I guess the active balancing makes up for the lack of castor wheels!

By definition - because its balanced, its C of G and centre of pressure are in the same spot. Or it would fall over! So it cant suffer at all from this.

Whats an X2??

Added, high end powerchairs typically have 24v x 100 or 120 amp controllers, each motor. So on a ramp, curb etc they pull up to 240 amps. So if you are using a serious wheelchair motor and controller you bertter be using high C rate lithium batteries too. On my chairs normal AGMs or Gel batteries cannot cope. So its Odyssey batteries only (2.5mOhm) less Peukert, etc And they are still crap! Hence the 78x 12Ah Headway cells. These have way better IR compared to lead and other lithium phosphate cells. Thats a 13s x 72Ah pack. (6p). So 3000 watt hours, no voltage real sag even at the Roboteqs 2x 150 Amps and around 43 volts.

 

[JennyB]

By definition - because its balanced, its C of G and centre of pressure are in the same spot. Or it would fall over! So it cant suffer at all from this.

Whats an X2??

The "off-road" Segway, which claims a top speed of 12.5 mph and a 12 mile range (or 24 miles on-road) from a 1 Kwh pack. That's good, but not dramatically better than your chair - despite lacking the tendency to turn downhill. I've seen an attachment for an unpowered chair that looked like the front end of an ebike with a front hub motor, and wonder how practical that would be.

 

[amberwolf]

I would guess that it will take less power to do the same thing as typical powerchair, despite the extra weight. It would be essentially an electric trike type bicycle at that point, without pedalling.

 

[SamTexas]

The "off-road" Segway, which claims a top speed of 12.5 mph and a 12 mile range (or 24 miles on-road) from a 1 Kwh pack. That's good, but not dramatically better than your chair ...

1000/24 = 41.7wh/mi Segway
3000/45 = 66.7wh/mi Burgerman

If the numbers you provided above are correct for the Segway (same assumption with Burgerman's numbers) , I'd say that the Segway is dramatically better, assuming of course that the two perform similarly.

Obviously both numbers are horrible by a factor of many when compared to the energy consumption of an electric bicycle.

 

[Burgerman]

The "off-road" Segway, which claims a top speed of 12.5 mph and a 12 mile range (or 24 miles on-road) from a 1 Kwh pack. That's good, but not dramatically better than your chair - despite lacking the tendency to turn downhill. I've seen an attachment for an unpowered chair that looked like the front end of an ebike with a front hub motor, and wonder how practical that would be.

It wouldnt. I need to turn on the spot, manoever indoors, turn around in a van etc. It would drastically improve total w/h per mile but isnt usable.

 

[SamTexas]

Update: I finally found the major source of inefficiency on the "3G Storm Series - Torque SP RWD".

Unloaded power: Both driving wheels off the ground and moving straight forward at full speed (WOT).
225 watts at 7.0mph! Or 32.1W/mph. Simply unbelievable. In this unloaded test three of the potential ineffiencies are removed from the equation: the weight of the wheelchair, the caster wheels effect and the motor steering effect.

For comparison, same unloaded test: my 26" wheel with a 9C 2807 Direct Drive Brushless hub consumes 76 watts at 36mph or 2.1W/mph.

 

[amberwolf]

That sounds like there is either some serious drag in the gearboxes (but shouldn't be that much or they will get pretty frickin' hot really fast), or something else is causing friction or other mechanical loading, that shouldn't be.

That's over 100W of power dissipated in each wheel/motor/gearbox system!

None of the powerchair motors I've had any experience with, gearboxes or not, use anything like that much power with only an off-ground wheel as a load.

 

[SamTexas]

Believe it or not amberwolf, those numbers are common. There is nothing wrong with mine.
http://www.wheelchairdriver.com/board/viewtopic.php?f=6&t=1858

But it sure means that improving the efficiency for these wheelchair should be a very doable task.

None of the powerchair motors I've had any experience with, gearboxes or not, use anything like that much power with only an off-ground wheel as a load.

What were your numbers?

 

[amberwolf]

Around 1-2A at most, using two series SLA (around 26V). Call it 25-50W, accounting for likely voltage sag. That's with the batteries connected directly to the motor, no controller at all. I can go back and dig out the motors and wheels to get exact numbers, though it might take a while. I can also rig up a power meter to test on the whole Pride powerchair I still have, but I have to get to it first. It uses a transaxle and a single motor, AFAICR.

 

[SamTexas]

Thanks amberwolf. No rush at all, but I would appreciate those numbers, along with pictures of the tested motors.

 

[woodygb]

Sam,

Have a look at this typical wheelchair motor and gearbox combo.

http://www.asidrives.com/pdf/adl100.pdf

You'll see that the no load amps is 5.2 amps = 124.8 watts loss in the motor and gearbox.

I've calculated this motors no load amps to be around 1.6 ... hopefully the maths and formulas I used are correct.

( 127 * 32 ) = rpm @ peak efficiency

Peak efficiency rpm = no_load_speed/(1+SQRT(Motor_No_Load_Current/(motor_rated_voltage/RΩ)))

4064 = 4500 / (1+SQRT ( 1.578 /( 24/0.175 )))

That puts the peak motor efficiency @ 80%...and makes this gearbox to be about 76% efficient @ that rpm ...this is a fairly typical value.

Regards Woody

 

[SamTexas]

You'll see that the no load amps is 5.2 amps = 124.8 watts loss in the motor and gearbox.

Thanks Woody. That's 10.4 amps for two motors or 250 watts. Slightly worse than my number (225 watts). It's beyond my imagination how the American electric wheelchair companies have been able to get away with horrible products like these.

http://www.goldenmotor.com/wheelchair-PW-2008EF.htm
China Golden Motors wheelchairs run at less than half the power (2A per motor LOADED). That's 96 watts at 4mph or 24wh/mile. Compare that to the number I measure on mine: 70wh/mile. Or to Burgerman's number: 66.7wh/mile. I'm not feeling patriotic at this moment.

 

[woodygb]

The Golden motor has been discussed over on the wheelchairdriver forum ... the specs don't really seem cut it for a wheelchair @ 24v.

IF my spreadsheets are correct PMDC motor/gearbox combos in wheelchairs are horribly inefficient .... around 30% efficient on the flat when driven @ approaching peak speed...(just rolling resistance and mass EDIT ..246 kg's used )..so minimal load.

Something like 172 watts into motor ... 65 watts into gearbox ...51 watts out of gearbox to move the mass ... 121 watts loss.

Put a hill or rise into the equation ( maintain the speed ) and the efficiency can rise by a factor of 2+

565 watts into motor ... 470 watts into gearbox ... 374 watts out to move mass up slope ... 191 watts loss.

 

[SamTexas]

Thanks for the follow up Woody.

The Golden motor has been discussed over on the wheelchairdriver forum ... the specs don't really seem cut it for a wheelchair @ 24v.

I don't know what that mean. Has someone over there actually tested those motors and proven that the specs were too optimistic? If so, please point me to thread with the hard data.

Something like 172 watts into motor ... 65 watts into gearbox ...51 watts out of gearbox to move the mass ... 121 watts loss.

That's much better, but still horrible (as you have pointed out). But I have a problem with the first figure, 172 watts. As you have pointed out earlier, the no load power is 250 watts (5.2 amps per motor). So how is it possible to see 172 watts loaded? And my own verified figure is 225 watts unloaded.

 

[woodygb]

MOTORS = 249.6 watts

MOTOR ...singular.... 5.2 x 24v = 124.8 watts


172 watts into motor ... 65 watts into gearbox ...51 watts out of gearbox to move the mass ... 121 watts loss
344 watts into motorS ... 130 watts into gearbox'S ...102 watts out of gearbox'S to move the mass ... 242 watts loss

Just type in "Golden" into the forum search ..there's not a lot of HARD info just experienced wheelchair "experts" with gut feelings ..

E.G. ...Joe builds and maintains soccer chairs ...his wife just recently competed for the USA in the wheelchair soccer World Champs in Paris.... The US Team beat the UK in the final. :x ....

 

[woodygb]

For what it's worth ...which may be nothing as I keep finding spreadsheet errors...
It's murder checking your own work! :x

 

[SamTexas]

MOTORS = 249.6 watts

MOTOR ...singular.... 5.2 x 24v = 124.8 watts


172 watts into motor ... 65 watts into gearbox ...51 watts out of gearbox to move the mass ... 121 watts loss
344 watts into motorS ... 130 watts into gearbox'S ...102 watts out of gearbox'S to move the mass ... 242 watts loss

Ok, that makes more sense. But it brings up another problem: It does not take 102 watts to move a 500lb 4-wheel vehicle at 5mph on level, smooth ground. It only takes about 50 watts (one single average healtly male can push a 2,500lb Civic/Corolla.) So the 121/242 losses are too optimistic. The real losses should be around 145/290 watts. So, 344 watts in and about 50 watts out for an overall efficiency of 14.5%!!! Once again, how could they (the manufacturers) get away with such inferior products for such a long time. All those so called engineers should be demoted to janitors.

 

[woodygb]

I'll have to do a little more research on the watts required to move a mass....as I find your 50 watts per 225 kg's a little low... but I'm no expert. edit :- research ( Google ) done ...Rolling resistance equation modified to output approx 10 watts per 100lbs.

Revised as per your estimation....edit :-well more or less .. just halved the mass ... motor watts in 145.2 , 33 out of motor & into gearbox , 26.1 out of gearbox, 119.1 wasted , 18% drive efficiency

 

[SamTexas]

Revised as per your estimation....edit :-well more or less .. just halved the mass ... motor watts in 145.2 , 33 out of motor & into gearbox , 26.1 out of gearbox, 119.1 wasted , 18% drive efficiency

No Woody. 145 is the loss per motor. Watt in is still 172. So overall efficiency is still 14.5% as I stated earlier. It really does not matter much. I just want to keep the math correct to minimize confusion. Bottom line is that both 18% or 14.5% efficiency are totally unacceptable (to me at least).

 

[woodygb]

Hasn't the Amps and volts required to move the reduced mass @ the same speed changed?

watts = torque x (2 PI/60 ) x ( rpm/gear_ratio )

 

[woodygb]

I just want to keep the math correct to minimize confusion

Indeed.
Motor specs

Level ground

Climbing max gradient.

 

[SamTexas]

I decided that improving such an inferior product is a waste of time. So I start fresh. My goal is to turn a tricycle with 20" wheels into a supreme mobility scooter.

Today I finish my first "proof of concept", using what I have on hands:
- A Trailmate tricycle with 24" wheels.
- A front 9C 2807 laced in a 26" wheel.
- A 40A Ebike.ca controller with a 27V LVC and a reverse option.
- 29.6V (8s), 32Ah homemade battery using laptop 18650 cells.

I had to "borrow" a 26" fork from one of my mountain bike so I can use the already laced front hub motor. The test result is spectacular:
7.2wh/mi at 7.5mph average. No pedaling of course. Compare that to the 58wh/mi at 3mph I got with the wheelchair.

On flat terrain, top speed is 15.5mph, and the power is around 250W (again no pedaling). That's 16wh/mi. The complete e-tricycle weighs 100lbs and I weigh 140lbs.

Now to the real work. First task is to find a solidly made tricycle with excellent rear brakes and 20" wheels all around. Do you have any recommendation based on actual use or ownership?

 

[woodygb]

Do you have any recommendation based on actual use or ownership?

NO...As I don't use one or currently own a wheelchair.

Seems that you have decided to build a mobility scooter...Fine.

I might however suggest that you investigate what a disabled person might require/need in the way of a wheelchair .
I therefore suggest a read of these 2 threads.

The real world and useless powerchairs
http://www.wheelchairdriver.com/board/viewtopic.php?f=2&t=2096&start=0&st=0&sk=t&sd=a
All Terrain Wheel chair design idea's?
http://www.wheelchairdriver.com/board/viewtopic.php?f=2&t=2076&start=0&st=0&sk=t&sd=a