PDA

View Full Version : Wheelchair mobility bot help



Mmnimer
03-26-2018, 06:57 PM
http://photos.app.goo.gl/eJAniapz3rQJOX0n2Dear all. I'm working on a project for my niece who had CP for a long time we have had very difficult times going to the beach and making her enjoy her day out. Looking around for solutions I came up with a company called freedom traxx who sell an all terrain mobility device.

Being priced at over 5,000$ which I definitely can afford I went around making her my own version of a sand Stormer as she calls it.

I've purchased all parts made all necessary fabrications and have made huge progress. From this point onwards, this is where I really need all the help I can get from you. I'm not an engineer nor have any technical background but like to use general sense.

I've purchased a wheel chair (made in China one) with a 200W motor. After dismantling this donor chair, I made all the modifications and got a working prototype by using a snow blower tracks system. The problem is that once I added any weight the sand Stormer will not turn. I guess it must be either a motor problem where it is not powerful enough. Another reason can be all the gears and sprocket I've used. Not sure. Might be a battery.

Having a look around I noticed some heavy duty motors like the invacarevTDX storm wheelchair motor and the quantum q6 edge 350w or 1000w motor. There is also another 4500rpm invacarevTDX motor. Could one of those be my answer.

I've attached photos to give you an idea of where I am so far.

jwatte
04-02-2018, 11:12 PM
It has been my experience that home-built options always end up being more expensive than buying the ready-made solution.
The reason for this is that there really is an advantage of mass production; there's also an advantage of having a good design cycle with appropriate tools and testing environments.

Is it your battery, your motor, your control electronics, or your gearbox? Yes, most likely!

Treads are much harder to turn than wheels. This means that you need more torque. To get more torque, you either get a motor with more torque, or you get a gearbox with higher gearing. The gearbox of course needs to also be capable of sustaining the torque you will run through it, continuously. If the chair ends up running too slowly with the given gearing and motor, then you need a more powerful motor that can run faster at a given torque limit.

Now, to deliver enough electrical power to the motor, you need a battery that has the appropriate voltage for the motor, and the appropriate amperage capacity (which is different from the appropriate amp-hours energy density.) Typically, for big things like this, you'll want several big lead-acid batteries in parallel, or a beefy lithium-ion battery. Lithium typically lasts longer, has higher ampacity, and weighs less, but it does require some care in charging (and preventing over-discharge) to not be ruined or catch on fire.
So, you need battery charging and management systems. Perhaps the Chinese wheelchair came with something like this, but my guess is it's both slow, and barely up to the task for the battery for the 200W motor; and 200W isn't very much. (There's probably more in your garage door opener, or garbage disposal unit.)
There's also the question of wiring; as you increase currents, you will need thicker wiring, to avoid heating up the wire too much (maybe burning/causing a fire!) or at least avoid losing power to wire resistance.

Is that enough? No! You also need control electronics that can deal with the appropriate voltage and current for your motor, as well as any additional overvoltage that happens when the motor stops (or the chair is pushed manually.) The control electronics that came with the original chair is unlikely to be suitable for anything more powerful than what came with the chair.
Additionally, you need the right kind of control unit for your motor/motors -- brushed motors versus brushless motors are very different.

There's math you can run to calculate out all of this. The necessary information is typically available on data sheets from the manufacturers of the components. (It's less likely that end-user sold goods will have this information.)
Motors come with power and torque curves, as well as voltage and current ratings, in the data sheet for the motor. Control electronics come with maximum ratings. Batteries come with capacity and ampacity ratings. (Note: Trying to draw too many amps from a battery will overheat it, and may set it on fire, no matter what the chemistry.)

Run the numbers -- load times lever size plus margin plus losses, equals necessary torque out to the treads.
Make sure your gearbox is rated for this torque. Then divide by gearbox ratio; this is how much torque you need at the motor. Look at the motor ratings to make sure it's sufficient. If not, increase gearbox ratio.
Also remember that turning in place for skid steer easily uses 2x the torque of just going forward; in harsh cases you may need as much as 10x as much torque! (grass, etc.)
Make sure to select torque for continuous duty, not stall torque, for the motor. If only stall torque is listed, continuous duty torque is between 1/5 and 1/4 of stall torque.
Once you have a motor and gearbox that work, look at maximum voltage and current. Select control electronics that EXCEED those ratings, because you don't want it to blow up while trying to climb up a steep hill.
If you have multiple motors, of course compensate for this!
Now, select batteries that can deliver the necessary voltage and amperage, and that have enough of a runtime with the given load.
For example, if you need 48V and 30A, then you need at least 12S LiPo batteries (44.4 V,) more likely 14S (51.8 V,) and to run them for 1 hour, you need at least 30 Ah rating (which means about 120 Wh rating per cell, times the number of cells in the battery.)
You need a balancing charger / battery management system that is suitable as well -- enough cell count, and enough amperage capacity.

tician
04-08-2018, 10:05 PM
Very cool project, but if you spent more than ~$600 on the electric wheelchair that you scrapped for the motors, motor controller, and batteries, then you wasted quite a bit of money. From what I can see in the images, the motors are 24V and I am assuming they are powered from a pair of ~7Ah 12V sealed lead-acid batteries. Using small lead-acid batteries like that means the sand stormer is not going to last long or run well, as those batteries simply do not have the energy capacity, power limit, or life expectancy needed for any system other than low power emergency lighting (even then, they still have to be replaced every few years even without any use). Small, low-capacity lead-acid batteries like that are lucky to last 200 charge-discharge cycles or 3 years (whichever comes first), and the amount of current they can actually supply drops very rapidly during the discharge cycle.

Also, if the sprocket attached to the motor output is larger in diameter than the sprocket on the input of the tracks, then it is gearing for increased speed and decreased torque which is not what you want in this situation. I am assuming the motors taken from the electric wheelchair were directly driving wheels 10~12 [inch] in diameter, so they should be able to drive the tracks pretty well when geared down (small sprocket driving larger sprocket) to increase torque to compensate for the increased friction/torque losses caused by the solid rubber belt tracks.


If I were starting from scratch using only 'off the shelf' components, then I would likely grab a pair of BaneBots BB-220 gearboxes in 64:1 ratio (http://www.banebots.com/product/BB220S-444-4.html) mated to a pair of low-Kv brushless motors like the KEDA 53-20 195Kv (https://hobbyking.com/en_us/kd-53-20-brushless-outrunner-195kv.html) and controlled by a pair of EqualsZero Brushless Rage (http://e0designs.com/products/brushless-rage/). That would give a gearbox output speed of ~70 [RPM] at 24 [V] nominal system voltage and a torque of ~46 [lbf-ft] or ~62 [N-m] when limiting current to 20 [A] (current limiting would require an external current sensor on the Brushless Rage power leads sending a signal to the central controller that will limit the PWM range of the control signal sent to the Brushless Rage). Using only a shaft coupler between the gearbox output shaft and the track input shaft (no additional gearing) and assuming a track outer diameter of ~10 [inch], then the nominal top speed would be ~2 [mph] or ~0.9 [m/s] and the driving force per track would be ~55 [lbs] or ~25 [kg] which will be able to easily push the sand stormer along at a decent clip. The Brushless Rage can handle system voltages up to ~60 [V], so speed could be easily doubled by using a 48 [V] battery pack.


As for batteries, I would probably go with a LiFePO4 system simply because it is almost entirely impossible for them to catch fire or explode even if you drive a metal spike completely through the cell. This makes them much safer to charge unattended than most other Lithium-Ion or Lithium-Polymer chemistries. The downside is that LiFePO4 can be a bit more expensive than other Li-Ion and LiPo battery chemistries because there are fewer factories producing them, and LiFePO4 also has a lower energy density but that does not matter in most ground applications where a little more weight can be a good thing (can be used as ballast to keep the vehicle from rolling over). Batteryspace tends to be my preferred source for high-capacity LiFePO4 packs, but Headway also makes 10~15Ah cylindrical cells that are quite common in electric bicycle conversion kits. For the sand stormer, I would probably go with two 4S 40Ah LiFePO4 packs connected in series for a system voltage of ~24 [V] and capacity of ~40 [Ah] which would be enough to power both motors at 20 [A] (24*20=480 [W] each) for up to one hour and the battery pack will last ~2000 cycles and/or several years of use. Batteryspace also sells BMS and chargers for their bare packs, and some versions of the battery packs come complete with BMS and charger leads already installed.

If you need to keep the initial cost down, Hobbyking sells several large capacity LiPo battery packs (https://hobbyking.com/en_us/multistar-high-capacity-6s-12000mah-multi-rotor-lipo-pack-1.html) (https://hobbyking.com/en_us/turnigy-graphene-professional-12000mah-6s-15c-lipo-pack-w-xt90.html)
that could be locked away in a vented, heavy gauge metal box to protect against the possibility of fire and direct any possible flames away from the chair. The battery packs would need to be connected in parallel (and/or split between the two motors) to increase the overall capacity, and would absolutely have to be charged individually using Li-Ion/Li-Po specific chargers with balance charging capability to ensure none of the cells exceed any of their voltage or current limits during charging and that they end up at the same pack voltage prior to connecting in parallel to prevent sparking. Current limiting functionality would be absolutely necessary if using the Multistar LiPo packs since they are made using 'energy' cells (high energy capacity, low output power) instead of 'power' cells (low energy capacity, high output power). Tracking cell voltages during use would also be essential for the LiPo packs, but can be accomplished with simple battery monitors / 'battery savers' that plug into the balance-charge connector and beep loudly if any cell voltage starts dropping too low to alert you to the issue so you can stop driving too hard and/or recharge the batteries.