Converting a Bakfiets cargo bike

MalcolmG

1 mW
Joined
May 9, 2018
Messages
14
Location
Auckland, NZ
Hi all,

We've had a Bakfiets long cargo bike for a few years now, and use it for carting some kids around (1 and 3) and the odd bit of grocery shopping. My wife and I are fairly fit and manage well without any e-assistance, but I think we'd use the bike more if it were electric. In a 5km radius of our house everything is relatively flat, but if I go much further things get a bit steeper. I could envisage doing 20-30km trips with the kids if I had e-assistance. I don't like to go too fast on this bike with the kids in it, and as I'm mostly looking for assistance for getting up hills I think a max assisted speed of 20km/h would be ok (but up to 30km/h wouldn't hurt)

So anyway, I'm thinking of adding a front hub motor (rear wheel has a nice 8 speed Nexus hub which I'd like to keep); the front wheel is 20" and although it currently uses roller brakes, there are mounting tabs on the fork for a disk brake so I will be converting.

At present this is what I'm thinking - I'd be interested in any feedback and/or recommendations for alternatives
Motor - Bafang G311 from Grin Technologies - standard speed windings, which will give better torque at low speeds than the high speed wind Grin recommend for a 20" wheel (I'm happy to sacrifice top speed).
Controller - em3ev 25A (anyone used these and have any feedback?)
Battery - I plan to make my own pack - I have access to a decent number of A123 LiFePO4 26650 cells, probably enough to do a 12s4p configuration, so 39.6V nominal and around 396W/hr.
Thumb throttle, CA2.4.

I'd be interested to hear opinions on this - I have very little experience with e-bikes (my wife bought one a couple of weeks ago, but that was the first time I've ridden one), but I am fairly technically proficient (mech engineer that does electrical engineering for a job, and I have a bevy of bikes and have worked on them for years)
 

Attachments

  • 50610856_10161424972760523_7864483580271919104_o.jpg
    50610856_10161424972760523_7864483580271919104_o.jpg
    209.5 KB · Views: 4,342
I have those controllers and still using them, work great.

If you are buying the hub from Grin, why not get it all there as a package? In most cases when you buy motor here, controller there and throttle somewhere, in most cases the wiring doesn't match colors or plugs are wrong. It fine if you understand how to match wiring. The colors are not always the same color for the same item, like a throttle. you can have a white or a green or any color they have handy, just for the throttle sensor. In most cases the red is red and black is black but not always.

Buy a kit if you can, a good one in your case as your kids are involved. A geared hub as you said Bafang. If your getting a controller from EM3ev, why not the whole kit. A MAC is a fine geared hub motor. A 10 or 12 turn is a beast on hills. In a 20" wheel I think a 10t on 36v would be a great combo. If from GRIN but a whole package there, they are both two of the most respected here on ES.

Dan
 
Which winding you choose will not affect the motor's torque capacity. If you want more torque, get a bigger motor. How fast you go is determined by programming and your throttle, so you don't need to go fast. The motor your suggest looks really small and underpowered even for the application.

A larger direct drive hub could also work, depending on the exact nature of your hills, and would have the advantage of total silence with a sinewave controller. The geared hub will make a lot of noise in comparison. Total silence is really nice.

396Wh is not much, so you might think about a larger battery.
 
Check out this thread, just a few lines below yours... https://endless-sphere.com/forums/viewtopic.php?f=3&t=101783

G311 is too weak for serious cargo, and the model has issues with magnet loosening at high speed (eg. in a small wheel). Maybe if it's super flat where you live but even than, I'd get at least a bpm or a mac geared motor. Better go with a heavier DD hub, than you have Regen and an additional brake.
 
MalcolmG said:
but if I go much further things get a bit steeper.

You're thinking, how much motor do I need? "A bit" more! You can't really use information like that for much, but if you can make up an approximate grade value, you can feed that into the motor simulator at ebikes.ca. It seems to be in part about RPMs, so it's good that you're in a 20" wheel.
 
DAND214 said:
I have those controllers and still using them, work great.

If you are buying the hub from Grin, why not get it all there as a package? In most cases when you buy motor here, controller there and throttle somewhere, in most cases the wiring doesn't match colors or plugs are wrong. It fine if you understand how to match wiring. The colors are not always the same color for the same item, like a throttle. you can have a white or a green or any color they have handy, just for the throttle sensor. In most cases the red is red and black is black but not always.

Buy a kit if you can, a good one in your case as your kids are involved. A geared hub as you said Bafang. If your getting a controller from EM3ev, why not the whole kit. A MAC is a fine geared hub motor. A 10 or 12 turn is a beast on hills. In a 20" wheel I think a 10t on 36v would be a great combo. If from GRIN but a whole package there, they are both two of the most respected here on ES.

Dan
The basic reason for getting from different places was simply that I like the look of the G311 motor from Grin, but their international shipping rates are pretty steep (about $200US to get a kit from them to NZ), whereas the EM3ev rates are much better. I will look more closely at the MAC motors that EM3ev sell, as I'm sure it would work out cheaper to buy the motor from them too.

Thanks for the feedback
 
flat tire said:
Which winding you choose will not affect the motor's torque capacity. If you want more torque, get a bigger motor. How fast you go is determined by programming and your throttle, so you don't need to go fast. The motor your suggest looks really small and underpowered even for the application.

A larger direct drive hub could also work, depending on the exact nature of your hills, and would have the advantage of total silence with a sinewave controller. The geared hub will make a lot of noise in comparison. Total silence is really nice.

396Wh is not much, so you might think about a larger battery.

I'm not sure if what you're saying is true regarding the effect of motor winding type on torque and speed. A quick check through the Grin tech motor simulator shows that the higher speed windings have a lower torque at stall but continue to provide useful torque up to a higher speed.

My main reason for going geared hub rather than direct drive is the greater starting torque (good for getting 160kg+ of rider/bike/cargo moving), and the fact that when it's not powered it will provide virtually no drag. Having said that, the regen braking and silent operation of a direct drive motor would be nice too.
 
Fork / frame issues with front hub motors

https://endless-sphere.com/forums/viewtopic.php?t=101819
 
qwerkus said:
Check out this thread, just a few lines below yours... https://endless-sphere.com/forums/viewtopic.php?f=3&t=101783

G311 is too weak for serious cargo, and the model has issues with magnet loosening at high speed (eg. in a small wheel). Maybe if it's super flat where you live but even than, I'd get at least a bpm or a mac geared motor. Better go with a heavier DD hub, than you have Regen and an additional brake.
Thanks for the link, some useful info in there. I guess my expectation (which may be wrong), is that because I'm doing OK with pedal power, the amount of assistance I need for what I'm doing isn't that much - hence looking at smaller motors (for the reasons you outlined in your thread around weight/handling). Having never ridden an electric cargo bike, I don't really know if my expectations are reasonable or if a "little bit of assistance" will be worthless for the effort.
 
donn said:
MalcolmG said:
but if I go much further things get a bit steeper.

You're thinking, how much motor do I need? "A bit" more! You can't really use information like that for much, but if you can make up an approximate grade value, you can feed that into the motor simulator at ebikes.ca. It seems to be in part about RPMs, so it's good that you're in a 20" wheel.
Yeah you're probably right, I should have a look at some of the hills that I currently struggle with and try to get some approx gradient values for them and run them through the simulator to get an idea of how well the different motors will perform.
 
DAND214 said:
The one thing about a DD is the drag or cogging, if you don't give some throttle. Yes a geared hub does whine but they have better torque at low speed.

Yes, I also would consider a larger battery. A little more is better than not enough.

Dan
That's a bit part of why I was hoping to avoid a DD (plus the better torque vs weight from a geared motor). It would be nice to have less noise and regen braking though.

I figure that if I start with the battery size I can make from existing parts then I can always make it larger later (although redoing connections and packing of the cells would be annoying), but given that the battery on my wife's bike is 3x the size and her bike is smaller/lighter, I think you're probably right that it might be a bit small.
 
john61ct said:
Fork / frame issues with front hub motors

https://endless-sphere.com/forums/viewtopic.php?t=101819
My forks are steel and as they currently have roller brakes I think the torque loading is probably much greater than what the motor will generate (admitted in the opposite direction). There is already a mounting location for a torque arm for the brake so I suspect I can probably fabricate something adequate.
 
MalcolmG said:
I'm not sure if what you're saying is true regarding the effect of motor winding type on torque and speed.

No, it's totally true. Basic misunderstanding by beginners. Read and learn: for a given phase current more turns will give more torque. But the amount of current carrying capacity (torque generating capacity) you give up by winding less copper more turns is exactly equivalent to what you gain from the additional turns. So yes, the faster winding will need more PHASE current (not battery current) to achieve peak torque. The maximum torque available from either winding on the same motor will be exactly the the same.

After choosing the correct motor for your application (you haven't, it's too small) you choose your winding based on the voltage you want to run and the speed you want to go.
 
flat tire said:
for a given phase current more turns will give more torque. But the amount of current carrying capacity (torque generating capacity) you give up by winding less copper more turns is exactly equivalent to what you gain from the additional turns. So yes, the faster winding will need more PHASE current (not battery current) to achieve peak torque. The maximum torque available from either winding on the same motor will be exactly the the same.
Excellent summary, will keep that as a reference - so others please chime in with any additions, qualifications etc.

> After choosing the correct motor for your application (you haven't, it's too small) you choose your winding based on the voltage you want to run and the speed you want to go

Say torque is (much) more important than top speed

big heavy cargo / tandem, major hill country, including starting from stopped halfway up,

any voltage between 36 and 72 is do-able, batteries will be at least 3kWh and willing to put out high discharge rate when needed

willing to put in CAv3 and/or whatever expensive controller to optimize / balance the various factors

what type of winding / phase current combo is best?

Torque is not the only priority of course, range is also right up there, as well as bank longevity.

A "slower" winding means more turns, and require a less stressful C-rate, is that right? More copper is good?

I also don't yet understand the distinction between PHASE current and battery current, good links to help me get up to speed would be appreciated.


 
More copper = more torque, other things equal.

More turns is a slower winding. Fewer is faster.

Phase current is current in amps sent to the motor from the controller. At lower speeds the controller doesn't need much voltage to shove current into the motor and can trade voltage for additional amperage. So motor current can be much higher than battery current, and motor voltage can be lower than battery voltage.

Battery current is current, in amps, drawn from your battery by the controller for processing to get sent to the motor. This current will always be supplied at your battery pack's voltage.
 
flat tire said:
More copper = more torque, other things equal.

More turns is a slower winding. Fewer is faster.

Phase current is current in amps sent to the motor from the controller. At lower speeds the controller doesn't need much voltage to shove current into the motor and can trade voltage for additional amperage. So motor current can be much higher than battery current, and motor voltage can be lower than battery voltage.

Battery current is current, in amps, drawn from your battery by the controller for processing to get sent to the motor. This current will always be supplied at your battery pack's voltage.
Thanks. So, if I am purposefully sacrificing top cruising speed, there is no point in going up to high voltages in the battery?

Actually want to get higher amps more easily out of the pack for torque, easier at lower voltages?

More efficient conversion by the controller as a result of Vin and Vout being closer?

I don't think I want to go lower than 52V (56-58 charging), does that seem right for my use case?
 
I'd consider 36V - why not? When you have the input parameters, try that in the simulator.
 
Is there really a compelling advantage to going that low?

I have other banks and infrastructure already set up for that nominal 48V range, would be a major PITA to have oddball 36V packs.
 
Not really, I guess - it's the same, isn't it, if you run a higher battery voltage and back off on the throttle.

Don't ask me, I just play with the simulator once in a while. I'd encourage our Bakfiets project engineer to do that too - and when possible check the motor winding options, to see if what he's hearing here is borne out by the simulator. I believe I've seen it come out either way, so it may make sense to check with the specific configuration.
 
MalcolmG said:
That's a bit part of why I was hoping to avoid a DD (plus the better torque vs weight from a geared motor). It would be nice to have less noise and regen braking though.

This is a big misunderstanding. Justin showed years ago that even under basic ride conditions (short trips, lots of stops and gos) the the amount of regen power you get during brakes is greater than the power loss due to stator drag. You can calibrate the controller so that PAS level 1 is just strong enough to overcome drag (around 1-2% of motor power) and just leave it continuously on.

I drive around 2 kids too here and I can't but stress the importance of an additional brake. That's a super advantage of DD hubs. If the usual 6kg hubs are too heavy for you, I suggest the lighter shengyi dwg09c http://www.syimotor.com/productDe_5.html
Only 5kg and plenty of power.
Otherwise I suggest a mac (4.3Kg), Greenpedel M58 (https://www.greenpedel.com/Product/M58R.html#pro_detail) or a shengyi x2 (3.2Kg). All geared though. The mac is surely the best of the lot but also super expensive. And you can have regen: ask for a locked clutch when ordering the mac, and you have both a geared hub and regen.
I wouldn't go with 36V. Lower tension means bigger wires.
 
qwerkus said:
I wouldn't go with 36V. Lower tension means bigger wires.
Great, so 48Vnom (51-58V) is a good range for this max-torque, lower speeds use case?

Or is there some advantage to going any higher?

 
This is a low power setup. He doesn't need big wires on any part of it.

You could choose an oversized direct drive motor like a QS205 if you wanted to conquer hills and not overheat, and do it all in total silence. Big motor doesn't have to mean fast. The motor will only accelerate as fast as its programming allows.
 
john61ct said:
Great, so 48Vnom (51-58V) is a good range for this max-torque, lower speeds use case?

Or is there some advantage to going any higher?

There is no absolute rule here. Higher voltage has the advantage of lower power losses with cables of the same diameter, but those losses tend to be negligible if cables are kept short. Note that on a cargo, depending on where the battery is located, cables might run a lot longer, and would make a difference. Also higher tensions overcome skin resistance easier (ohms law) and are thus more dangerous.

In the end it's rather a matter of motor windings. Most chinese motor are wound with a goal of 250-300rpm for 26" or 200-250rpm for 28" wheels. If you take a slower wound motor (more copper turns per pole) and apply a higher voltage, you can reach the same speed than a faster wound motor (less copper turns per pole) with a lower voltage. Hence the tradition to describe motor windings with a kv number: kv * voltage = rpm. Sometimes it can be hard to find a fast winding for a giving motor, which means that if you want to lace it into a small wheel (like 20"), you have no choice but to take the standart winding, and apply high voltage (like 60-72V). The opposite applies if you want to slow down a motor optimized for 26" wheels in order to use it in a larger 29er wheel.

The important thing here is that whatever winding / voltage / current combination you chose, basic motor characteristics will never change. A 500W motor will always stay a 500W motor, since it's power limits are definite by its ability to shed heat, and the strength of its components (in geared hubs, the clutch and the nylon gears).

See: https://www.ebikes.ca/learn/power-ratings.html

Where voltage is more relevant is in battery design. Say you need a 500W power output at battery level. In a 52V setup, that would mean 500/52=9.6A per cell. In a 36V setup, it's 500/36=13.9A per cell. This number gives you the minimum of parallel rows you need: using 10A max cells, the 52V setup can work with only a single row of cells (14s1P) while the 36V setup would need at least two (10s2p). Using cheaper 5A max cells, you'd need at least 14s2p in the 52V setup, or 10s3p in the 36V version, and so on.
 
Back
Top