Which Controller (edit: hub motor) for my daughter's wheelchair?

Buk___ said:
He also said that he would either cancel those motors or sell them on having found out that direct drive motors were not right for his application.
Too late to cancel, not too late to sell them, but I think I'm still going to use them for testing purposes. I'll explain in a separate thread.

Buk___ said:
And whilst he does talk about 20" wheels; he does not say that he has already brought those, or is committed to them.
Sorry Buk. I have purchased 20" wheels. But hey, this is a prototype, and wheel diameter might change in the future.

Buk___ said:
If you assume 200kg for the chair, but spread the load across 4 motors; recognise that you aren't looking at 6 hours of constant WOT, but rather short bursts of (say) 20% power with cool down time between; then the demand per motor is so low that even when running them at speeds where they are only 50% efficient, heat would not be a limiting factor.
I haven't done the math yet as I'm still working out some of the intricacies in the design but my gut instinct tells me that 200kg is probably about right.

Buk___ said:
As irrelevant, out-of-date arguments between 'helpers' does nothing to help this worthy venture, I attempted to bow out. I do so again.
Please don't bow out Buk, I've really valued your input thus far.
 
Your direct drive motors will be fine. The trick here is to run a low enough voltage that will shift the efficiency curve so that the motors operate with an acceptable efficiency at low speed. What is the lowest top speed that you are willing to accept? This will dictate the voltage required for a given motor winding.
 
dogman dan said:
If you are going to use a hub motor, I think 4 of the smaller, slow wind, geared motors would do anything sane you'd like to try. ( at 700w each) Driving up the rock staircase though, I'd have to understand better why it has to be driven there.
Because it's the easiest way up the highest mountain in the country. It's iconic. And it will be a huge achievement for my daughter.
Some more background info here http://mountainwheelchair.com/the-team/ada/

dogman dan said:
Sure, wanting to get there I understand. But maybe it does not have to be in a completely independent wheeled vehicle the last mile?
I understand what you're saying here, and I agree, but I want to leave that as a last resort, rather than something I factor in during the design stage.

dogman dan said:
I also question the decent.
I'm questioning it too. I've rode down that path on a bike myself. Both on a downhill bike and a normal MTB. It isn't easy. I've also seen my smaller prototypes roll over many times when they come down off too large an obstacle.
This is partly why it needs to be radio controlled; so that I can test it driverless first.

dogman dan said:
You think I'm not serious but I am
I think you're absolutely serious.

dogman dan said:
A decent chair that can do almost anything less than that rock staircase will do. Up and down 20% grades will be no problem for it. That means typical logging roads in the mountains, and many quad motorcycle routes.
...but not the intended path, hence I'm having to custom build one.

dogman dan said:
If you end up with the dd motors, you will end up with a thing able to go about 60 kph, so make sure your controllers can be limited in speed, but not amps. Controllers that can direct connect with a cycleanalyst would be good. Not sure if that means 4 CA's, but it might.
I didn't know that was possible. Thanks for pointing it out.

dogman dan said:
I see no need for 6 drive wheels, but idling wheels in the middle might be ok.
Please see my previous post.
 
danielrlee said:
Your direct drive motors will be fine. The trick here is to run a low enough voltage that will shift the efficiency curve so that the motors operate with an acceptable efficiency at low speed. What is the lowest top speed that you are willing to accept? This will dictate the voltage required for a given motor winding.
Lowest top speed. 3kph I guess.
...on a 15 degree (26%) gradient
 
So, my feeling for the moment is this...

6 x 1500w DD motors have already been purchased and are on their way.
I appreciate they're not the best solution for the task, but seeing as they will soon be sat in my house, I think I'm going to give them a go anyway as if nothing else it will allow me to conduct some tests.
To begin with, I'll only build the drive system and a platform for the batteries, controller and RC receiver.

Once that's done, I'll drive the thing around (radio controlled) without the added weight of the passenger or a large part of the frame which makes up the seat.

If possible, it sounds like it would be good to get the windings changed on the hubs I've purchased and also run them at a lower voltage than they're intended for.

If the tests go well and it seems likely that it might actually get up the intended path then if the motors are getting hot I'll look into wheelchair/wheelbarrow/geared motors.

Sound sensible?
 
Glyn said:
danielrlee said:
Your direct drive motors will be fine. The trick here is to run a low enough voltage that will shift the efficiency curve so that the motors operate with an acceptable efficiency at low speed. What is the lowest top speed that you are willing to accept? This will dictate the voltage required for a given motor winding.
Lowest top speed. 3kph I guess.
...on a 15 degree (26%) gradient
By my estimates, 12V would give you a top speed of around 8mph and would put 3mph at around 40% of your 'no load' speed. This would be an improvement, but not an efficient enough zone IMO.

6V would give you a top speed of around 4mph and would be perfect for your usage with 3mph at around 75% of the 'no load' speed. This would keep the motors running at their most efficient.

Running such low voltages would mean that you would need to run higher currents to make up the power.

The biggest problem with the plan is sourcing controllers that would accept such a low voltage. I suspect RC controllers would be perfect, but perhaps someone else will be able to advise further.

EDIT: I've just realised that I've based the above calculations on using 26" wheels. Running 16" wheels would mean that 12v will give you around 4-5mph top speed. I'm sure you get the idea now anyway.
 
Glyn said:
Presumably then less heat is produced. [At lower voltage] However, I also presume it would produce less torque or have I not understood how torque is achieved?

Both torque and heat are caused by the flow of current. Torque is linear with current, but resistive heating is proportional to the square of current.

As you lower voltage (or as back EMF rises with speed), you can get into a regime where the motor is unable to pass all the current your controller can provide. The motor won't make any more torque at that point, unless voltage rises or motor speed falls.
 
This is really interesting and likely to make quite a difference so if you could keep humouring me please guys then that would be awesome.

I think an important part of understanding that I'm lacking has something to do with volts and current, so let me see if I can cobble my thoughts...

Power (watts) is a combination of volts and amps.
To produce 1500w at 48v would need a current of (1500 / 48) = 31.25a (I appreciate there are many variables to consider and this a simplistic view).
It seems to me like what you're saying is that the same motor could produce 1500w at 12v if it was supplied a current of (1500 / 12) = 125a.
Is that right?
One of my earlier prototypes uses a 24v controller with 150a per channel which suggests a 12v version is not too far fetched.
 
The question to answer is whether the motor you're using will accept enough amps at 12V to produce the power you require. If not, you'll have to use higher voltage.

All your cables, connectors, power traces etc. will need to be 16 times bigger to withstand 125A than they would to take 31.25A.
 
Glyn said:
Reading between the lines in your above statement, are you saying that the motors windings dictate it's optimal rpm? This would suggest that the windings can be modified for higher torque/lower rpm?
Yes, but as Chalo pointed out it's easier to lower the voltage and make the wheel smaller, which effectively does the same job regarding speed.

It doesn't change the torque, though; you'd still need either a wider stator (and magnets), or larger diameter rotor/stator, or both, to increase the torque.

(the more wire that runs parallel to the magnets, vs the end-turns, the more work the current thru the wire gets to do, vs how much work is wasted at the ends.

the larger the diameter of the stator/rotor, the more leverage it has so the more force is multiplied for higher torque, all other things the same.)



Having already purchased some motors (too late to cancel now - already been shipped) how easy would it be for a novice to modify the windings?
It's not techically difficult to wind a motor--but figuring out what windings you would need, including what thickness of wire (and/or how many strands of it) will fit in teh stator with the winding pattern you need, and then actually doing the winding, is not necessarily that easy.

As a novice, I've tried to rewind a couple of motors, and I gave up the first time because it was actually really hard to pack the wire in there correctly (all I was doing was a repair on a burned winding on a brushed motor, so no math was involved :) ). The second time I eventually gave up because I kept nicking the wire and shorting it to the stator (and itself thru the stator). It also hurt my hands to keep pushing and pulling the wire tight enough. :/ I keep envisioning building/rewinding some motors to do exactly what I want on my trike, but it's a lot of work that I don't know for sure would work when I was done (if I screw up or short windings, etc).


But it's definitely possible; you can look up threads where people have done this. "rewind" would be the search term I'd use first.


It's just not the first thing I'd do to fix the problem--I'd use the brushed motors with gearboxes, mostly because I like to overbuild things, not necessarily because it is the best option. Most likely, custom-wound motors would be the "best" solution, but not necessarily the most practical.


FWIW, I'd probably try the stuff you already have first, just to see how it performs. You never know--it might be "good enough" to work, even if it gets hot and wastes power (requiring more battery than you'd like to get the range you need). I tend to try what I've got at hand, learn from what it doesn't do right, and what it does do right, and go from there--but I'm more of an iterative-artistic type than an engineer type. ;)


Before you act on any of my information, you should look thru my build threads and see if you're willing to go thru that kind of process. :lol:


I've seen posts on this forum where people have requested custom windings from the manufacturers. Does anybody know how willing they are to do this?
Depends on what minimum order they have; almost certainly you'd have to buy some dozens (or hundreds) of motors to get a custom winding.

If they happen to already have motors wound for the speed you want, you could just buy those.

Or, as Chalo pointed out, get motors (that already have the torque you need) that if put in a small enough wheel at a low enough voltage would spin at the correct speed. So...if you need them to spin at 5MPH in a 20" wheel at 24v, for instance, then get motors wound for xxMPH in a xx" wheel at xxV. The three "xx" parts you can work out proportionally from what's available from a vendor vs what you need it to do. (sorry I'm terrible at math so I cant' give you a formula).


Would changing the windings be able to produce a hub motor which is suitable for my application?
Yes...but probably not *just* changing the windings, unless the motor were already capable of producing the torque you want. The goal of changing the windings is just to reduce the full-loaded-speed of the motor to what you want out if it. It still has to be able to create the torque to start with.

(one way of creating more torque is run a motor at a *higher* voltage than it was listed for, and apply higher throttles to get the higher torque--but this also means it's wasting more power as heat than a motor designed to create that torque with less current, *and* it's going to spin faster at that higher voltage, if left at full throttle).

Motor relationships are not quite as complicated as they seem, but it's easier to see how they work when playing with simulators like Justin's at http://ebikes.ca/simulator and comparing different ones or the same one with different parameters, using the Compare System A vs System B options.

However, wheelbarrow motors(!!! :mrgreen: ) might work. Something like this 18" 48/60v 1200w wheelbarrow motor might work.
I see nothing that says what speed it is intended to go at what voltage. You need to know both of those to know if it will work for you, because you'd want it to have a very slow max speed at the voltage you will run it at.

It also says nothing about how much torque it has (foot-lbs or newton-meters) under what conditions. You need to know that because you need higher torque to climb the hill.


are there companies willing to do it?
You can probably find motor-rewinding companies even in your local area (wherever that is), but it will be expensive.



I guess the primary reasons for using a hub motors are space saving (fits within a conventional fork) and simplicity. Buying a hub motor means I don't have to fabricate complicated mechanisms for transferring power from the motors to the wheels.
Its definitely simple, and is the primary reason I'm using hubmotors now (I had various problems using motors to drive the chainline, mostly because I didn't know how to build frames to prevent twisting; I do now, though I still haven't gone back and done the chaindrive thing yet. Is in the plans for the future, eventually).


My thoughts about fabrication apply here. For motors with an output shaft to drive the wheel directly, these seem to be intended for much smaller caster-type wheels. If I attached the 20" wheels I'd lose a lot of torque.
Unfortunately, yes, to both torque and complication of fabrication.


How is this achieved? fans? liquid cooling?
With brushed motors, the easiest way is to put a fan on one end of the motor that sucks air thru it, and put slots on the opposite end for intakes. You can look at various DIY car /etc conversions using forklift motors to see the types of fan rigs used on those, and emulate something similar in smaller format.


So, let's say I was using 15ah 3.2v Headway LiFePo4 cells. I'd need to wire 15 cells in series to achieve 15ah at 48v.
You'd actually want 16s (not 15s), for a typical LiFePO4 "48v" pack.

You'll also probably have to parallel a number of these packs to support the current draw you are likely to need. (or make parallel groups of multiple cells, then series those).


Each cell has a Max Discharge Current (Continuous) of 5C. Wired in series, am I correct in thinking that they have a combined discharge rate of 5C, not 5C per cell?
The C-rate applies to cell or a set of paralleled cells exactly the same, but not seriesed cells. If one cell has 15Ah, then 5C is 75A. If you parallel two, then there's 30Ah, which at 5C would be 150A.

My personal take on Headway stuff in general based on memory of posts around the forum is that Headway overrates them, and the voltage sag you get at the "rated" C-rate is high enough that I wouldn't use them at that rate. But I have no direct experience with them. There are members that have builds using them, though, so you can look up their build threads and see what they have to say. (lots of posts about Headway)
 
Glyn said:
I'll explain in a separate thread.
I recommend keeping the whole project in the same thread, so people can keep following it. If you make multiple threads for the project, those already following won't necessarily know to look for others, and you might lose some poeple that were helping in previous threads.

Plus, new followers that don't know about the previous threads are going to repeat stuff said in previous threads, potentially wasting their time and yours.

But it's up to you how to post your stuff. ;)
 
Glyn said:

I don't know those specific controllers (and have no personal RC controller experience, but have read around a bit on them here on ES and places like RCGroups), but keep in mind that many RC controllers are meant to run motors that spin up to full speed rapidly, like propellers on helis and planes, so their high loads are very short term, and they don't have a lot of heat dissipation ability.

They also don't all have current-monitoring and limiting (at least not sufficient limiting to save them from overcurrent draw when used on things like ebikes/etc).

If you run wheel motors on them, especially on something like this project, I'd be afraid of blowing them up.

So before you go with any RC controller, you'll want to be sure it is capable of self-limiting current quickly enough and sufficiently enough that it will remain within it's safe heat limits.

You'll also want to be sure it's easy to remove heat from the controller's FETs, because a lot of those controllers ahve FETs built onto the PCB, so the PCB itself has to do much of the heat dissipation, which it cant' do as quickly as a heatsink might be able to do once the PCB reaches a heat-saturation point. If the heat is trapped, the FETs overheat and fail, probably resuilting in the destruction of the PCB as well (meaning no repair possible).
 
danielrlee said:
Running 16" wheels would mean that 12v will give you around 4-5mph top speed.
...
Running such low voltages would mean that you would need to run higher currents to make up the power.
How does this work? If the motors are rated at 20 amps / 45a max (48v), how is the current affected when running them at a lower voltage. Will passing a higher current through them kill them? Will they still draw their max current even at this low voltage? I've been looking at at this Trackstar 200a (1000a max) ESC which can operate at 12v.

https://hobbyking.com/en_us/turnigy-trackstar-1-5th-scale-sensorless-200amp-8s-opto-car-esc.html
1a.jpg


This ESC has been put to good use in this racing go-cart:
cmvslide.gif


The guy who built the go-cart has a comparison of 200a RC ESC's here: http://www.etotheipiplusone.net/?p=3096. I didn't find anything useful to help move this project forward but an interesting read nonetheless.

There are however lots of bad reviews for this ESC. I then arrived at this similar, but more reliable with better quality Castle Creations Mamba XL ESC. The description on the manfucaturers page does specifically say "for RC vehicles weighing up to 30 lbs" (13.6kg) and also "not intended for human or animal propulsion".

So depending on the answer to my above question, I'm thinking a skateboard ESC might do the job. Looking at the pics on AliExpress, it doesn't look like the motors I've purchased have hall sensors, and this Skateboard ESC (50a continuos/200a peak) is programmable, has regenerative breaking, is intended for human propulsion and according to the blurb, has great start-up torque with sensorless motors.
 
amberwolf said:
Glyn said:
I'll explain in a separate thread.
I recommend keeping the whole project in the same thread, so people can keep following it.
Sorry, I meant separate post: https://endless-sphere.com/forums/viewtopic.php?f=1&t=91933&p=1341788#p1341157
 
amberwolf said:
I don't know those specific controllers (and have no personal RC controller experience, but have read around a bit on them here on ES and places like RCGroups), but keep in mind that many RC controllers are meant to run motors that spin up to full speed rapidly, like propellers on helis and planes, so their high loads are very short term, and they don't have a lot of heat dissipation ability.
I think this might be the reason that Mamba XL says not for human prepulsion.
"[with a specific motor it becomes] the perfect combination for large-scale, ultra-high-performance RC vehicles. Have you ever seen a 1/5th scale Baja doing 70 mph? If you’ve got the traction, this Castle combination can get you there".
 
danielrlee said:
Running such low voltages would mean that you would need to run higher currents to make up the power.
Glyn said:
How does this work? If the motors are rated at 20 amps / 45a max (48v), how is the current affected when running them at a lower voltage. Will passing a higher current through them kill them? Will they still draw their max current even at this low voltage?
I've been searching for an answer to this but struggling as I don't know the correct terminology to use to get a suitable response from Google.
Taking a hypothetical motor with the following ratings as an example:
1000w
50v
20a

If I were to half the supply voltage to 25v, which of the following statements would be true?

A. The motor would still produce 1000w by doubling the current it draws to 40a (25v x 40a = 1,000w).
Or
B. The current would remain at 20a and therefore the power would be reduced to 500w (25v x 20a = 500v).
 
I'm not sure I know how to interpret the outcomes correctly but if you're suggesting that I can answer my question by looking a the graphs then...

Take a motor doing 25kph.
at 48v, the load is 180w and the motor power is 1561w
at 24v, the load is still 180w but the motor power is reduced to 308w.

I can interpret that in two ways but I don't know which is correct without answering my previous question.

Interpretation A
If I assume that the load (180w) is the current drawn from the controller, then:
Changing the voltage makes no difference to the current drawn but the motor power is significantly reduced.

Interpretation B.
If I assume the motor power divided by the supply voltage tells me the current drawn then:
At 48v, the current is (1561w / 48v ) = 32.5a
At 24v, the current is (308w / 24v ) = 12.8a
In this instance, changing the voltage reduces both current and power.

So in conclusion, after looking at the simulator, I'm still none the wiser.

*edit* Just noticed the table titled Electrical. Looking at that, as voltage is reduced, so is everything else. Current draw, torque , and importantly temperature are all reduced. In which case, I likely don't need a controller rated at anything over 50a. Is this correct?
 
amberwolf said:
Use the http://ebikes.ca/simulator to test different configurations with the same motor and you can see how it works. :)

The best (possibly only suitable) motor for the OPs application is a geared motor -- whether geared hub or worm geared -- and the simulator does not handle those correctly.
 
Glyn said:
I'm not sure I know how to interpret the outcomes correctly but if you're suggesting that I can answer my question by looking a the graphs then...

Take a motor doing 25kph.
at 48v, the load is 180w and the motor power is 1561w
at 24v, the load is still 180w but the motor power is reduced to 308w.

I can interpret that in two ways but I don't know which is correct without answering my previous question.

Interpretation A
If I assume that the load (180w) is the current drawn from the controller, then:
Changing the voltage makes no difference to the current drawn but the motor power is significantly reduced.

Interpretation B.
If I assume the motor power divided by the supply voltage tells me the current drawn then:
At 48v, the current is (1561w / 48v ) = 32.5a
At 24v, the current is (308w / 24v ) = 12.8a
In this instance, changing the voltage reduces both current and power.

So in conclusion, after looking at the simulator, I'm still none the wiser.

If you apply the full voltage to any motor, it will draw as much amperage as the controller can supply. The power consumed will be V * I drawn.

But how much of that power drawn will be translated into useful work depends upon the load, the motor speed, any gearing involved, and the motor's ability to handle the drawn current without melting.
 
Buk___ said:
If you apply the full voltage to any motor, it will draw as much amperage as the controller can supply. The power consumed will be V * I drawn.
I don't think I understand what you're saying here sorry Buk.
Using the motors I've purchased as an example, they're rated at 48v, 20a continuous, 45a peak.
Are you saying that if I connected them to a 48v controller capable of supplying 100,000a, then they would indeed draw 100,000a?
 
What Buk says is true only if the motor is capable of drawing more current than the controller can provide. But especially as you lower voltage to slow the motor, that may not be true at all. For any given voltage, there is a maximum current that a motor is able to transmit. As voltage rises, this value rises with it.
 
Chalo said:
For any given voltage, there is a maximum current that a motor is able to transmit. As voltage rises, this value rises with it.

So therefore the opposite must also be true; as voltage drops, the maximum current that a motor is able to transmit drops with it. And if said motor has a peak rating of 45a at 48v, if I lower the voltage then this peak rating will also drop and therefore I don't need to find a controller capable of producing more than say 50a. Is this correct?
 
Glyn said:
Chalo said:
For any given voltage, there is a maximum current that a motor is able to transmit. As voltage rises, this value rises with it.

So therefore the opposite must also be true; as voltage drops, the maximum current that a motor is able to transmit drops with it. And if said motor has a peak rating of 45a at 48v, if I lower the voltage then this peak rating will also drop and therefore I don't need to find a controller capable of producing more than say 50a. Is this correct?

The motor's rating is different from how much current you can stuff through it. The rating is how much current it can take continuously in normal operating conditions without overheating. It's a lot less than the maximum possible. But cut the voltage by a factor of 4, and the motor might not let through 45A even if the controller can provide it.

There's another concept that might be useful to understand in this case. Battery amps versus phase amps. The controller synthesizes an AC wave form for the motor's three phases. The current on these phases can be more than or less than the current from the battery. That's a function of both the physical limitations of the controller and its programming.
 
Back
Top