Wheelchair inefficiency

[Jeremy Harris]

Interesting project. I did a lot of work on drive train losses a couple of years ago. Gear reductions can really eat up power in a big way if the reduction ratio is high, with the losses being highly non-linear with respect to output power. I found that there was a fairly big fixed power loss component, plus a slightly variable rpm-related loss component, but that it was the fixed loss that was the killer for efficiency. It's easy to end up with a sub-10% efficiency system at low speeds, but this may well then get up to maybe 20 to 30% efficiency at higher speeds.

I helped out with a few mobility related projects a few years ago and found that generally things like mobility scooters etc aren't engineered for efficiency, they seem to have been designed primarily for robustness and reliability as the key design drivers. It's also tough to get good efficiency when one of the key safety requirements is that the drive train has enough friction/stiction to limit the rate of acceleration/deceleration when power is applied and to damp out the action of the safety brake coming on and off as power is applied/removed. One big downside of putting a low loss reduction drive in place of the gearbox is getting the wheel braking to feel OK for someone with limited motor skills - the thing actually feels safer to some users if it has a sluggish response.

One way to get a "good" response (probably the reverse of what we'd call good from an ebike perspective) would be to use a controller with very good and responsive regen coupled with a fairly high torque motor and a reduction drive with no backlash. The controller could then provide the drive train stiffness that's needed to instil confidence in many users. The reduction drive would perhaps best be made using toothed belts, as chains would be too messy for a lot of users. Finding a good controller with regen that is tunable, and with an electronic braking facility (where some power is applied when the motor is near stationary, to keep it feeling stiff as regen stops being effective) would be the challenge, but not insurmountable. I believe that some of the newer electric wheelchairs are using clever brushless controllers like this, presumably for the same reasons, to maintain the needed stiff drive train feel and yet improve overall efficiency.

 

[woodygb]

The Invacare gearless brushless ..refered to earlier in this thread ...is ..as suggested in it's title blurb... direct drive.

Invacare’s Gearless-Brushless
GB™ motors are not only a key
component of the TT® TrueTrack
technology, the design is 75%
efficient resulting in 17% more
battery range than Invacare’s
4-pole motors

Not really a quantum leap in efficiency as reported by Sam.

 

[Jeremy Harris]

IMHO there can't really be a "quantum leap in efficiency" when the core requirement needs power to provide the desired response and feel. Using electronic braking (or more accurately, electronic rotor position feedback) to give the required stiffness in the drive train is going to absorb a fair bit of power, maybe a bit less than using a relatively high frictional loss gearbox to gain the same effect, but still a fair bit of power, all the same.

I think that one thing some able-bodied people under-estimate is that it takes a fair bit of power just to provide the sort of stability and steering response required

 

[SamTexas]

It's good to have you back Jeremy.

IMHO there can't really be a "quantum leap in efficiency"...

Quantum leap efficiency: Of course there is not such thing. Woody used that term mainly because he was comparing it against the efficiency of existing wheelchairs (14.5% by my calculation). The efficiency of my "proof of concept" is exactly what I expected before I even built it. It's just basic expectation based on previous ebikes experience.

... when the core requirement needs power to provide the desired response and feel. Using electronic braking (or more accurately, electronic rotor position feedback) to give the required stiffness in the drive train is going to absorb a fair bit of power, maybe a bit less than using a relatively high frictional loss gearbox to gain the same effect, but still a fair bit of power, all the same.

I think that one thing some able-bodied people under-estimate is that it takes a fair bit of power just to provide the sort of stability and steering response required

You might be on to something here. Something that I have miserably failed to understand. I would appreciate your inputs as to where I went wrong in my thinking:

Response and feel: The 280 lbs 3G Storm Torque SP RWD wheelchair (wheelchair from this point on) I have been playing with has variable speed control (analogic to the 3-speed ebike switch). Regardless of which speed I choose, the energy consumption is the same at the same speed. It just takes longer to reach that speed at the low setting. While the lowest setting gives a very slow and thus reassuring feel, the highest setting is extremely jerky and startles even an able-bodied as myself. So between the smoothness of the low setting and the jerkiness of the high setting, I don't see why it takes more power/energy to do it either way.

Stability: I agree that the wheelchair is stable. But that stability is achieved using the most rudimentary method: Ballast. As mentioned above, the wheel chair weighs 280lbs. My trike will have a low center of gravity and hopefully will not weight more than 120lbs complete.

Steering response: I understand why wheelchairs use the "tank steering" method (no steering wheel and small turnaround radius). But as far as response is concerned, I could not think of a worse way to steer. But that's beside the point. What I don't understand is why this steering method is energy hungry? It's just a matter of NOT running one of the motor (therefore braking it), or running both motors in opposite direction.

 

[woodygb]

The "terrible response" that you refer to is a consequence of the wheelchair manufacturers programming of the controller and once programmed "correctly" it becomes very responsive.... some disabilities will of course require different settings.

http://www.wheelchairdriver.com/powerchair-programming.htm

 

[SamTexas]

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

The "terrible response" that you refer to is a consequence of the wheelchair manufacturers programming of the controller and once programmed "correctly" it becomes very responsive.... some disabilities will of course require different settings.

http://www.wheelchairdriver.com/powerchair-programming.htm

Thanks Woody. I'm not trying to build a mobility scooter for the mass. I'm building it specifically for my BIL. He's handicapped but not totally. His hands and arms are fully functional. His legs are extremely short and extremely weak. He can still walk at 1mph with the help of a walker. He has 2 wheelchairs that he can use inside the house.

What I want to give him is a means of transportation outside of the home. To the grocery store, the coffee shop, friends' houses, etc... within a 10 miles radius. He's currently doing all the above with the wheelchairs with great difficulties and danger, mainly because the sidewalks and roads in his neighborhood are horrible.

 

[JennyB]

Crazy idea, but... Something like a minibike that fits under the chair. Chair wheels lift when under way, bike wheels lift for slow, tight corners?

 

[Jeremy Harris]

During the time I was playing around with stuff a bit like this, one thing I quickly learned was that every users needs were dramatically different, yet the manufacturers had to produce a more or less standard bit of kit, albeit one that could be tweaked and adjusted to some extent to suit individual users needs. Efficiency never seemed to be very high up the priority list, probably because the majority of users never come close to using all the capacity available. Some do, for sure (I have a paraplegic friend who could easily use any amount of battery capacity you could fit in a chair in less than a day!), but again the manufacturers design for the majority.

I believe they deliberately used high ratio, high power loss, gear boxes because they were a simple solution that massively eased the control problem. It's hard to get something that's overly responsive when there is such a large gear train involved. Once they'd done this originally it became "custom and practice" to use heavily geared motors, as it makes the electronics relatively simple.

I understand that the newer machines are using direct drive with better electronic control, but there is a fair bit involved in getting the control system to have the right sort of response and feel, I believe. The advantages are better efficiency and (hopefully) a way to programme the controller to suit the individual user, which should make a big difference. I worked with one chap who had such limited delicate hand movement and strength that he needed a really slow control response. He hadn't got the delicacy of touch or hand speed of movement needed to control a machine with a fast response. On the other hand, my paraplegic friend loves speed and wants the fastest chair possible!

I tried one neat idea that another disabled friend had that showed promise, a lightweight Quikie chair with a clip on front hub motor (an old Crystalyte 300 series hub motor, I believe) that turned it into a front wheel drive trike. Such a system might suit your BiL, Sam, as it would be fairly easy to DIY. It was great, as it was fast and would go over fairly rough ground pretty well, as the front wheel had a fat moped size tyre. He could also clip and unclip the front fork, wheel and front fork mounted battery pack himself in a minute or so, and convert the chair back to manual. The only real problem with it was that the controller wasn't well-matched to the requirement and still seemed to use speed, rather than torque, control. This made going over kerbs etc a bit challenging, as it needed a big squirt of throttle to get up the kerb, then the throttle had to be closed quickly, or else the chair would shoot off at high speed. Had I had the chance I'd have quite liked to engineer a proper controller for it, so that the throttle could set the desired torque, or maybe power, rather than speed.