Brushless E-bike hub motor questions. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Brushless E-bike hub motor questions.

previous 1 2 3 4 next

Move Thread

LAN_403

Ash Small

Sun Apr 26 2015, 12:31AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Thanks for all the above replies. Legally, unless I register it, The 'legal' speed limit for an unregistered wheelchair in this country is 4MPH, which is 6.4KPH (maths seems easier in SI units).

I work out that 26" wheels have a circumferance of ~80", which is ~ 2 metres.

6.4 KPH = 6,400metres per hour, which is slightly more than one hundred metres per minute. If wheels are two metres circumferance, that's just over 50 RPM.

(That's the 'legal' limit

)

I want the ability to climb inclines, and I mentioned it will be fitted with mountain bike tires

Thanks BB for reminding us that a 'big' motor will have more torque at any RPM. I think some of the smaller ones that drive the existing bike chain with an ~11 tooth sprocket, and use the bike's original gears need to be spinning quite fast to generate any significant power (power equals torque times RPM).

The distributors just seem to specify watts and don't mention torque or RPM.

I don't mind 'forking out' for big motors if they are going to have sufficient torque to get a wheelchair up to 4MPH from a standing start on a hill. They will be far more reliable in the long term (at least, that's the theory) and, at 50 RPM, won't consume much power. Generally, the faster you go, the less efficient it is. Is that the case with these motors, running from lithium ion batteries???

Also the gears and chain drive on smaller motors increase weight and decrease reliability.

As Bjorn pointed out, I need figures for torque, and then I need to do some more maths regarding inclines, acceleration, etc.

I'll have another look at the figures posted by ... later. How reliable are these figures? Can I reliably use them?

I'd assume two of the 250 Watt hub motors would do, but I should really do some maths with reliable figures

Thanks again for the assistance.

.

...

Sun Apr 26 2015, 12:52AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

As to the reliability of my numbers, whenever you are dealing with import ebike parts it will be hit or miss since they change up the designs fairly regularly to save costs. But those numbers should in general be fairly standard for ebike parts, the particular tire I have is

Ash Small

Sun Apr 26 2015, 01:38AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

... wrote ...

.Also, I personally strongly suggest that you buy controllers for them off ebay as well (they are incredibly cheap, something like $30 for a 50a/50v controller). They are quite easy to interface to, you plug the motor in, plug the battery in, and put 0-5v into the throttle to get 0-50amps out (yes they are current mode controllers, not voltage mode like a RC style controller). Almost like they were designed for this application

I did mention above that I intend to incorporate 'pulse skipping' for steering, as I can't think of any other way to achieve steering. I'd have no idea how to incorporate pulse skipping into these controllers, so I don't see any alternative but to build my own.

Also, pulse skipping is the easiest way I can see to limit max speed to ~4MPH.

Also, if the 36V, 250W motor in my original post can get a bycicle moving from a standing start on a hill, then I'd assume that two would get a wheelchair moving.

.Post edited, see below

hen918

Sun Apr 26 2015, 09:44AM

Registered Member #11591 Joined: Wed Mar 20 2013, 08:20PM
Location: UK
Posts: 556

BigBad wrote ...

hen918 wrote ...

Sorry Ash, these hubs are designed to be peddled up to ~10MPH before power is applied, they have next to no torque at speeds of a few MPH, and are essentially very high geared, for efficiency at ~15MPH.

Don't understand.

How do you design an electric motor with low torque at zero speed and high torque at high speed???

The only way I could see that is if the drive circuitry throttles it.

I meant it had low torque across the board (like you would expect), and is not designed, and probably not capable of, accelerating from a standing start.

Ash Small

Sun Apr 26 2015, 09:56AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

hen918 wrote ...

.I meant it had low torque across the board (like you would expect), and is not designed, and probably not capable of, accelerating from a standing start.

I can see that if it was struggling to get going, and power was applied to one set of coils for a prolonged period, and that if they rely on spinning, or just airflow, for cooling, then you could burn coils out subjecting them to heavy low speed, or stationary, loads,

I'm hoping that ... can give me more information regarding how his 350W motor copes with standing starts on steep gradients without any 'pedal assist'.

If his copes fine, then I can't see any real problems using two 250W motors.

If his struggles, then maybe I should consider either using more powerful hub motors, or look for another solution.

EDIT: Looks like I mis-read ...'s post, He has the 1000W motor, but in the link he posted, it appears to be a 350W motor. Sorry about the confusion.

If his 1000W motor is ~50Nm then two 250W motors should be good for ~ 25Nm.

BigBad

Sun Apr 26 2015, 02:57PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

hen918 wrote ...

BigBad wrote ...

Don't understand.

How do you design an electric motor with low torque at zero speed and high torque at high speed???

The only way I could see that is if the drive circuitry throttles it.

I meant it had low torque across the board (like you would expect), and is not designed, and probably not capable of, accelerating from a standing start.

That doesn't make any sense. Apart from a bit of stiction you need virtually no torque at low speed to move, and torque requirements increase with speed, not decrease.

...

Sun Apr 26 2015, 03:20PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Not sure where you got 350w from, the motor I have is a '1000w' (looking at the ebay link there is 350w in the auction title, but that is just there to help people find the motor, the auction only has '1000w' motors)

One thing you might want to double check before buying a motor is to confirm that it does not have an integrated freewheel, some hub motors have an integrated 1-way hub (so that you do not have to pedal against the motor while not using it), which may or may not be what you want for your application.

As to 'pulse skipping' I am not entirely sure what you mean, but if you mean that you are planning to not use PWM then your are going to blow up the motor. The winding resistance of these motors is on the order of 0.1-1 ohms so you are going to be running hundreds or thousands of amps stall when the mosfets are on. The standard way to run these motors is to PWM the high-side devices in the h-bridge at a few kHz (these motors have quite a lot of inductance, so even a 1khz is fast enough for good control), and use a feedback loop to keep the motor current (torque) below a setpoint, then if speed control is needed add another loop around the current loop for speed control.

As to motor startup, some motor controllers (in particular the ones without sensors) really do not like starting from a standstill, and are truly designed for pedal assist. Literally they have a sensor which are connected to the pedals so that the motor is controlled by peddling. That said, the controllers which do have sensors seem to be able to startup from a stop just fine. The motors themselves will work fine at low RPM, they are sealed motors so they are immune to the issue that fan cooled motors have where the fan does not run at low rpm, thus causing them to overheat at low RPM. You just will not be able to get very much mechanical power out of them, since you are limited to whatever current the motor is rated for (probably 20a for your 250w motor)

Ash Small

Mon Apr 27 2015, 06:20PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Yes, I've given this a bit more thought. Sorry about the '350W' mistake, I was trying to take all the 'mportat' bits,

The more I think about this the more sense it makes to get two of the 1000W motors and two of the standard controllers.

I think that with a suitable resistor bridge, or something I can probably manage to control the steering.

I'll see if the supplier I was thinking of using stocks the 1000W ones, and I'll have a look on Ebay.

Only other question is what barreries to get.?...Presumably Lithium ion, but which ones...?

EDIT: I didn't mention my reasoning for getting the biggest ones I can, it's so It will pull away on steeper inclines, and will be able to cope with rougher terrain

EDIT: This looks the part:

Shame it's 'out of stock'.

EDIT: I've ordered two bare hubs. They should be here on thursday. I've not ordered controllers yet.

EDIT: Found this on the Wikipedia page on BLDC motors:

"A motor with windings in delta configuration gives low torque at low speed, but can give higher top speed. Wye configuration gives high torque at low speed, but not as high top speed"

Ash Small

Thu Apr 30 2015, 06:18PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

These arrived today:

The only place I seem to be able to order 48V 1000W controllers on Ebay UK is direct from China (three to four weeks delivery). I had to get rear hubs in the end, as front ones were sold out, but I don't think it's a problem.

As I see it, for each direction of rotation, you have six possible states from the sensors and six possible states for the stator. Each possible combination from the sensors corresponds to a three digit binary number between 1 and 6.

A logic array 3x6 would output the correct state for the stator for any sensor state. for a given direction of rotation.

Once you have the correct stator state, you apply your chopped PWM signal at ~1kHz, or whatever the required frequency for these motors is.

EDIT: Are these like stepper motors, where the rotor is magnetised in place, by sending pulses through the stator windings, and will I de-energise them if I strip them?

(Work in Progress)

...

Fri May 01 2015, 05:11AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

You might want to try looking around on various e-retailers for stock controllers, I know in the US you can get them overnight through amazon for about double the ebay rates

Do not get too excited though, the firmware on some of those units are pretty bad and they may not behave well at low RPM. I have a program going at MIT where we reprogram the microprocessor inside commercial controllers to get around this, but there is so much variation in the controller designs that we have given up on the project for the most part (identical controllers bought from the same vendor have had completely different microprocessors in them!)

You are correct, driving a sensored BLDC motor is actually pretty easy, you can do it using a some simple logic. You can also get controller chips which do it all together, usually designed for running CD-rom drives. I had a commercial (goped brand) BLDC scooter which literally had one of these guys inside

Sadly those controllers do not have integrated high side gate drive, so you would need to add your own gatedrive, either using a prepackaged device (something like a fan7888) or go discrete (like

circuit 3 was roughly what the goped used). Gives you nice things like current control, braking. In any case, the datasheet is worth a reed as it goes into the details of how to avoid common pitfalls with BLDC motors. Of course you can also replace that chip with a microprocessor, with careful use of timers and PWM an arduino can be made to work quite nicely for bldc commutation.

PS - I would recommend a PWM frequency of ~5khz for these motors, 1khz is a little low. Ideally you would go to 25khz to minimize audible noise, although the gate drive requirements get a lot worse at those frequencies. May as well implement sinusoidal drive while you are at it

As to the motors, BLDC have permanent magnets and it is OK to disassemble them.

Unrelatedly, you should go and spend a day reading Charles blog. He basically builds vehicles of similar class to the one you want to build professionally.

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob