Bafang G3xx series high reduction ratio geared hub motors

The Bafang G3xx series seem to offer some interesting motors that fit standard dropout spacing with disk brake compatibility - There is already a thread for the G310/G311 https://endless-sphere.com/forums/viewtopic.php?f=2&t=92124 so I won't mention those here aside from listing it's specs for comparison sake: 11:1 reduction ratio, 135mm diameter, 138mm width, Inrunner with 16 poles, 30nm max torque, 2.7kg. https://www.bafang-e.com/cn/components/component/motor/rm-g310250dc.html. (I'll call this Mini size). Read other thread for details but it looks like it has ridiculous power density, especially with oil cooling.

One of the other high reduction ratio motors in this series that seems interesting is the
FM G370.250.D/V https://www.bafang-e.com/en/components/component/motor/fm-g370250dv.html.
Similar to the G310, 16 pole count inrunner, in a smaller 100mm diameter motor (I'll call Micro size), the front motor version is standard 100mm width. Why is this motor interesting? It has a 12.5:1 reduction ratio and despite weighing just 1.7kg can output 42nm of torque! (Supposedly...). Like the G3xx it's rated for 250watt
Unfortunately the rear version RM G370.250.DC https://www.bafang-e.com/en/components/component/motor/rm-g370250dc.htmlis a wider 145mm, more than I would comfortable stretch a standard rear dropout.

The other motor of interest is the RM G360.500.DC https://www.bafang-e.com/en/components/component/motor/rm-g360500dc.html.
This is a larger 170mm diameter (I'll call standard) motor, ie similar in size to a mac. This one has a 12.8:1 reduction ratio, weight of 4.1kg. Standard (with stretch) dropout width of 138mm. Rated for 500 watts and 50nm of torque this seems to be the bigger brother of the G310. The pole count for this one is very low, just 8 poles. I THINK this one is an inrunner too (it's described as "having an Inner rotor" on the Bafang website). If the magnets don't get thrown off at high RPM like the G310 this one may be able to put out some decent power/speeds for it's size without hitting super high eRPM. I think this may be an issue with this one though due to the increased diameter causing more force + a higher reduction ratio meaning higher RPM.

These motors seem to be quite new (2018?) so I can't find much information about them apart from on the official bafang website and 1 short thread on 1 russian forum https://electrotransport.ru/ussr/index.php?topic=56542.0. (Interestingly after a google translate they mention the G370 as possibly running hot, although this could have been an early design)
I also cannot find any suppliers aside from 1 Alibaba supplier, Yongkang Gomax who lists all their motors as "2018 Bafang" (why bother actually putting the model name in the title where it is searchable? :x ).

Anyway to save others going through every single one of their product pages working out which is which I think its as follows
Front G370 (micro) https://cngomax.en.alibaba.com/prod...08/2018_Gomax_Bafang_Brushless_Hub_Motor.htmlNote, there is no model number, this is purely based on the torque figure and weight or https://cngomax.en.alibaba.com/prod...ang_Mid_Drive_Brushless_Geared_Hub_Motor.html
Rear G370 (micro) https://www.alibaba.com/product-detail/2018-Bafang-Hub-Motor-48V-500W_60777932423.html
Rear G360 (standard) https://www.alibaba.com/product-detail/2018-New-Brushless-Geared-Bafang-8Fun_60775162265.html

They also seem to supply the G310 (the only supplier I've found other than Grin).
Front G310/G311 (mini) https://cngomax.en.alibaba.com/prod...Use_For_Electric_Bicycle_Motor_Brushless.html
Rear G310 (mini) I don't think they supply, at least I can't find it on any of their listings.

If anyone finds any more information on these motors feel free to share, they look very interesting with their high reduction ratios. If they are anything like the G310 they should also be able to put out some decent power, especially with oil.

Nice find!

I believe these ratings are extremely understated. I have no idea how you could manage to only produce like 500w with a 4.1kg dual reduction motor like the 360. ebikes.ca's dyno data seems to show that the G310 can sustain 700w for a long period of time, and that's about half the motor.

This could be more powerful than the MAC

The magnet flingoff issue could be just as bad with the big boy, if not worse due to the higher ratio.
If it has the power density i think it has though, then the critical RPM could match it's power capabilities in a 26" or 700c wheel.
if you want high power density in a smaller wheel, you have to accept a lower top speed.... or just buy a MAC or DD or mid drive.

Agreed on the G360 being the one that really excites me.

I've found it's actually in a few production bikes already, specifically the Surface 604 Shred https://surface604bikes.com/products/shred-electric-bike. (Which if we weren't talking about motors actually seems like a decent ready to ride)
They are running it in a 27.5" wheel + Fat tyres @ 48 volt. They also have the torque rated higher than Bafangs 50nm at 65nm, although I would take this with a grain of salt as they specify using a 12fet 500w 25amp controller and I don't really think the maths adds up there.
They are also getting 28mph/45kph (speed limited, supposedly can set up to 32mph from videos) out of it which isn't too bad at all, probably pretty much voltage limited at this point anyway. (Your also probably not doing 28mph with just 500 watt though, unless your in full tuck). Anyone in the US want to go take a test ride?

Going off the specs + bike above, I think the motor itself would be happiest "offroad only" at 60V in a 29" wheel, 72V for a smaller 26"er. Although my solution pretty much always is just throw volts at it
I'm curious where eddie losses and iron saturation hit on this one, or if it does just hit magnetic explosion first, the 26" would probably be throwing magnets at 72v top speed.

Anyway I've emailed everyone's favourite BMSBattery to see if they can source one.

https://www.alibaba.com/product-det...ml?spm=a2700.7724838.2017115.1.2d41c74cLJjYd7

Would love to have that motor in a 24" rim with a PhaseRunner on 60V 8)

neptronix said:

The magnet flingoff issue could be just as bad with the big boy, if not worse due to the higher ratio.
If it has the power density i think it has though, then the critical RPM could match it's power capabilities in a 26" or 700c wheel.

Click to expand...

Excuse the newb blundering in, but would kevlar wrap be a solution to the fling-off issue?

High resistance to heat and most chemicals, very low stretch, electrically / magnetically inert.
I dont know what the clearance is, I'm assuming the allowance for thermal expansion & eccentricity / runout is way more than a monolayer of fine thread needed to hold the mags at a zillion rpm? Material cost near zero.

If rpm were unconstrained by fling-off what is the next bottleneck and how far away is it - controller erpm?

Welcome to the forum. That's a way less bumbly question than the first ones i made here :lol:

I'd imagine that a fine sheet of kevlar inside a motor spinning at thousands of rpm would be quite a feat to figure out how to make reliable.
If the magnet glue gets loose enough that they're ready to fly off their ring, then the kevlar would have to support them afterwards..

You might as well do it right, and forget using glue.. hold the magnets in another way.. in some kind of slot.
This would require re-engineering the motor entirely.
The added power also means you'd have to make the clutch and gears thicker..
Now you're in a real mess of re-engineering.

But let's say you did that.
Your limit is now the controller's eRPM... and also the thin-ness of the laminations.

My guess is that these motors have 0.35mm lams. Spinning faster would mean you'd want 0.27 or 0.2 so that the eddy losses in the stator don't get atrociously bad.

Another problem is that they shed heat poorly. If you wanted to design a 10lb geared hub motor that did 2-3kW all day, you'd have to design it from scratch so that it had oil cooling of some sort.


These dual reduction motors are fantastically power dense, but you'll never hit the full potential of the copper, iron, and neodymium in them.

Direct drive motor in a 20" or smaller.. or chain driven motor hooked to a sprocket on the rear wheel is still the best way to get huge power per lb.

But for a rider that doesn't demand high power or state of the art motors, these are still damn good motors. If i ran a company that built ebikes, i'd install nothing but these on the bikes. Especially considering how much power per lb you get VS a mid drive.

Yep certainly agree Proper Engineering (tm) is always preferred. Modifications get especially interesting when something cheap or easy brings a significant improvement, eg ferrofluid. The Kevlar thing I was envisioning just spinning on a layer of thread & dosing with appropriate adhesive as a way to make an extremely strong wrap (well, materials wise at least).

0.045" thread has a breaking strain of 225lbs (1.14mm/102kg) so by the time you have a few dozen from one side to the other plus bonding you're holding some pretty incredible forces. Negligible stretch is also key.

I should stress, I have no idea what I'm talking about :lol: , I just had occasion to use some kevlar thread recently & even though we all know its a remarkable material its still amazing to hang my fat arse off the ground on a 1mm thread.

Of course even if this supported infinite rpm its all pointless if it immediately hits the next limit such as strength eleswhere, heat etc, I wonder what this motor would do sans the mag fling off issue.

I'm hassling suppliers at the moment for the RM.G360, if anybody has specific questions about them I can ask while I'm doing so.

wil said:

The pole count for this one is very low, just 8 poles.

Click to expand...

I'm hassling suppliers at the moment for the RM.G360, if anybody has specific questions about them I can ask while I'm doing so.

Click to expand...

Any information about the rotor type would be interesting, lam thickness, IPM vs SPM etc.
The low pole count might be an indication of IPM rotor, like BBSHD has.
This thread shows the differences between SPM vs IPM rotors of Bafang mid-drives - 16 vs 8 poles.

Now about hub-motors - as the smaller G31x has 16 poles,
how does the larger diameter G360 rotor with it's 8 poles look like?

Has anyone managed to buy the FM G370.250.D/V motor?
If so, does it live up to its datasheet?
Where did you buy it and at what cost?

I have good reason to believe that Justin over at ebikes.ca will be either experimenting with, or offering the large dual reduction hub for sale in the near future.
We may even be lucky enough to get data in the ebikes.ca motor simulator.

neptronix said:

I have good reason to believe that Justin over at ebikes.ca will be either experimenting with, or offering the large dual reduction hub for sale in the near future.
We may even be lucky enough to get data in the ebikes.ca motor simulator.

Click to expand...

In that case hopefully we get a full characterisation of the G360 then! Much better then anything anecdotal I'll get.

Unfortunately I haven't been able to get any more detailed information from suppliers over that which was already in the Bafang datasheet.

I have a motor and spare gears on it's way from GoMax (the one I linked to above). I ordered a 260 RPM (slowest available) 500watt version, going in a 26" rim. This will take a month+ to arrive though due to lunar new year.

The only other supplier that would sell me one was cnebikes, but as they did not actually stock them it would be a special order with a 40 day lead time.

Wow - and some still think hub motors are dead... Was going for a heavy mac mod to add a 2nd reduction, but this could mean a lot less troubles. Please keep us posted about the new G360. Definitely following this.

Hi,

I'm interested in the G370 motor from Bafang, do you know where I can order it ? I haven't found it on Aliexpress.
Thanks.

Jil

So my G360 has finally arrived! Chinese new year, unresponsive sellers and unknown shipping statuses aside...

The speckles are bits of styrofoam, not chips in the paint. It has a Higo plug, I would prefer plain bullets as I can't imagine you can get many amps through one of these. (Anyone know the limit?)

The motor appears very similar in construction to the G310, just bigger (thread over at https://endless-sphere.com/forums/viewtopic.php?t=92124)

- Markings, 15 speed code which in BPM and CST is 205 RPM The QR code translates to: RM G360.500.DC 14 1 008F5T2280001 (Different speed codes, 215RPM?)
I ordered the slowest wind 500 watt version because I will be running at 20s->24s. Should be good for around 65km/h on 20s (My slightly slower wind SWXH 201rpm gets around 55-60km/h at 1800watt)

Disassembly:
Undo 6 hex bolts on the side without the cassette, then pry it off. It is a slight interference fit so I pried apart with screwdrivers.

O-ring around perimeter of gear side of motor, nice! Definitely an improvement over previous Bafangs.


Inner of gear side, it is easy to get to too! Just 6 bolts and it pulls off. Sealed bearings too (well at least shielded). Should be easy to replace the gears should they melt, although getting them might be difficult, the vendor I bought from was very reluctant to send gears, supposedly Bafang is cracking down on non authorised dealers repairing.


Pedal side cover off. To remove you need to undo 6 philips screws holding the cassette on and use a gear puller. (Or, don't use a gear puller, smack shaft first onto the ground a few times. I had to 3D print a piece to protect the wiring to do this)It is a tight interference fit between the case and the cassette base but once separated it is easy to get the rest off.

The cassette itself absolutely SUCKS to get off, 2 17mm nuts hold it on, but they aren't actually nuts, they are washers with 2 flat faces so sockets don't fit. Plus due to them being set into the the casette itself it is near impossible to get any contact. Eventually I ended up using a set of plyers at a 70 degree angle and mangled the faces of the nuts getting them off... Thankfully you don't actually need to remove the cassette to get to the motor internals (Wish I knew this before struggling with it so much...)


Speed sensor, there is a matching magnet glued into the outer case


Matching magnet in outer case.


Wiring is potted on


Finally at the internals! Like it's smaller cousin G310 it's definitely an Inrunner! Due to the wiring being potted on the case above you can't really separate further than this. Also of note, the magnets aren't just glued on! They are set into the rotor, so shouldn't fling off at high RPM. Also the motor may be a bit "Reluctancy" due to them being within the rotor, rather than on the surface, although from what I've read the magnets don't look deep set enough for it to really take affect.


It's like Bafang designed this with either a speed or temperature sensor in mind, those 2 through holes in the middle of the image are ground and direct to the S output. I'll just slice the S trace and wire a thermistor straight there, easy!


Copper fill factor could probably be improved a touch, but for a commercially produced motor it is fine. The wire is around 0.4mm thick


The stator laminations themselves. Visible is 13.7mm, overall thickness is 17mm


36 lams in 13.7mm = ~0.38mm lams


integrated thrust bearing & helical gears

Will be a bit before I get in on the bike, I THOUGHT I was buying a 26" laced hub, but evidently not. I currently only have a em3ev 6FET too, so limited to 25amp, I need a decent FOC controller!

I want to oil fill this (after great success running an oil filled SWXH), but I'm concerned that the top side of the windings won't be covered with oil and cooled as much due to them being sealed inside an extra case vs the SWXH, there will be less splashing around.

If anyone wants any other specific measurement etc let me know

wil said:

So my G360 has finally arrived!

I currently only have a em3ev 6FET too, so limited to 25amp, I need a decent FOC controller!

I want to oil fill this (after great success running an oil filled SWXH), but I'm concerned that the top side of the windings won't be covered with oil and cooled as much due to them being sealed inside an extra case vs the SWXH, there will be less splashing around.

Click to expand...

Looks like it is a very well designed motor, thanks for opening it and showing us!
I think a 6FET Nucular controller would be great with this motor: https://endless-sphere.com/forums/viewtopic.php?f=31&t=75494
And being an inrunner with very good heat shedding to the outside of the motorcase would filling with oil really be beneficial? Maybe some sort of hubsinks/aluminium fins (oil filter cooling fins?) mounted to the outside of the case would do the trick?

What I like most about ES. Member buys brand new motor and tears it down completely and documents it for all to see. Qudos!

But my biggest beef with Bafang, as well as other hub makers, is that they don't seem to have gotten the message that there are many more options for 32h drilling pattern rims available these days than 36h. At least in regards to bicycle type especially those designed for use as tubeless. I would really like to spec the smaller G motors for front hub conversions but without them being 32h it just isn't happening for me.

I use a Grin AnyAxle which is 32h and can use 9mm, 12mm, 15mm and 20mm axles types. I guess as long as I am wishing I would ask that Bafang would also go that route, oh and an integrated torque arm would also be nice.....

Nice teardown!

wil said:

The cassette itself absolutely SUCKS to get off, 2 17mm nuts hold it on, but they aren't actually nuts, they are washers with 2 flat faces so sockets don't fit. Plus due to them being set into the the casette itself it is near impossible to get any contact. Eventually I ended up using a set of plyers at a 70 degree angle and mangled the faces of the nuts getting them off...

Click to expand...

Same as with the G310 then. It depends on your luck. On one motor I managed to get these suckers off, on another they're stuck on too tightly. Justin mentioned his G310s always had them come off easily.

Thankfully you don't actually need to remove the cassette to get to the motor internals (Wish I knew this before struggling with it so much...)

Click to expand...

You do if you want to change the gears tho.

Also of note, the magnets aren't just glued on! They are set into the rotor, so shouldn't fling off at high RPM.

Click to expand...

Lucky you! In the G310 they're glued, see pic.

I want to oil fill this (after great success running an oil filled SWXH), but I'm concerned that the top side of the windings won't be covered with oil and cooled as much due to them being sealed inside an extra case vs the SWXH, there will be less splashing around.

Click to expand...

Go ahead! These inrunners benefit a lot from oil cooling. I'm planning to do the same on mine. What kind of oil are you using? It should be something thin that won't foul up or attack insulation/lacquer/seals/glue. I'm planning on using silicone oil.

How's the freewheel on your motor? The G310's has a quite a lot of drag, based on the two motors I had.

What would also be interesting is the noise level of your motor. It has helical gears on the first reduction only, right?

1N4001 said:

You do if you want to change the gears tho.

Click to expand...

Nope not on the G360! Looking at this vs the G310 teardown Justin did you can see that internally the motor is flipped, the inner gears are on the opposite side, to get to them you just undo the 6 bolts and take the one cover off, much more maintainable.

1N4001 said:

Lucky you! In the G310 they're glued, see pic.

Click to expand...

I was worried the G360 might be the same, but I had the suspicion that due to the lower pole count the G360 would have them set into the rotor (like the BBS02B rotor) and also the reason I thought it may be more of a reluctance motor from the outset.

1N4001 said:

Go ahead! These inrunners benefit a lot from oil cooling. I'm planning to do the same on mine. What kind of oil are you using? It should be something thin that won't foul up or attack insulation/lacquer/seals/glue. I'm planning on using silicone oil.

Click to expand...

On my SWXH I just used regular car ATF. I have heard some reports of it being bad for the motor internally (none of these reports existed when I did it 5 years ago though ) My motor still runs fine after 5 years at 1800 watts. Would silicone oil not chew through the wiring insulation though? I just wired my thermistor in with silicone insulated wiring, maybe I'll need to change that...

1N4001 said:

How's the freewheel on your motor?

Click to expand...

Not sure yet, I need to lace it into a rim which is something I haven't done before... + I figure with Australia it may take me a bit to get the parts.

1N4001 said:

It has helical gears on the first reduction only, right?

Click to expand...

Correct, first reduction only.

wil said:

Would silicone oil not chew through the wiring insulation though?

Click to expand...

Nope, that stuff is pretty much inert. That, along with it being very thin and having decent heat-carrying capacity make it ideal for oil cooling a hub in my opinion.

I've seen pictures of insulation being turned into a flaky mess by ATF. If you use silicone insulation everywhere, I think it should be fine tho.

wil, thanks for the motor porn! can't wait to see what the performance is like on that bad boy.

It looks like a finely designed motor. And it looks like it has a shit-ton of thermal mass to handle lots of abuse.
It's unfortunate that they didn't increase the diameter of the rotor a bit by moving the copper portion outwards towards the edges, hence making more potential power!

The G360 is laced up and installed!
Out with the old:



In with the new:



The new G360 sits quite well, although it is definitely more noticeable than the SWXH. Excuse the shoddyness of the setup, it is only temporary

Riding characteristics: I'm running the exact same controller settings to begin with as I was with the SWXH, a EM3EV 4110 6FET, 25amp, 30 phase amp, on a 20s RC LiPo pack (everyone's favourite 5Ah hard cases). This setup peaks at about 1800 watt when fully charged.

As you would imagine with the same controller setup, it feels much the same as the old motor.

It is CONSIDERABLY quieter than the SWXH though, up to about 600 watt it's almost silent, above that there is a slight whine, but much less than the old motor. Previously I would cut power whenever I passed someone so it wasn't obvious I was "Performance enhancing", no need for that anymore

The motor also feels like it is straining slightly less with hills, even with the highly restricted phase current limit. This may just be because it runs so much smoother than the previous motor.

One thing is, with this testing I am running through a Z910 Higo adapter, which terminates in those awful crimped "bullet" connectors.


I don't have any actual bullets to build a new setup with at the moment, so those are just jammed into the female bullets on the controller, definitely not the best or most efficient connection there.

No load at 75 volt with rim and tyre is ~500 rpm with 110% speed, (I forgot I was setup with overspeed enabled...). This uses about 150 watt. In real riding sitting on about 55km/h on 26" wheels seems to be the max on 20s, maybe it is finally time for me to jump up to 24s My rear brake is rubbing a bit, at no load it was increasing draw by 125 watts... I'll just wait until the pads wear a bit and rub less I guess.

I rode for about 10 minutes at 100% power and the core got up to 55 degrees C. This included multiple acceleration from standstill and large steep hill sections. Using the ebikes.com trip sim my old SWXH should have gotten to about 90 degrees, a G310 75 degrees.
I wish I had a MAC to compare to, as it seems to be the closest competitor.

I'm waiting on a new torque arm before I increase power from here, I'm just running on v1 clones at the moment... at least they are waterjet cut ones, not stamped and are reasonably thick.

I am quite surprised, I was expecting the controller to have issues at top speed due to the high reduction ratio, therefore eRPM, but I didn't notice any stuttering or other issues.


Edit: oh and the freewheeling feels identical to the oil filled SWXH, I had the rear brake off to test so there was no extra drag from that.

Great results, Wil.
Are you planning on using a different controller in the future? Like a sinewave or FOC type that would be quieter and produce even less heat?

About time they improved geared hub motors. Excuse my ignorance, but this is a lower gear inside right? So at lower wheel rpm the motor has a higher rpm than has been the case.

Does this affect top speed? or is it paired with faster spinning wind? I'm the kook who has always been interested in lower top speeds, for hauling trailers, and such high load applications.

As said above, a DD in 20" wheel works real good for that type thing, and can have all the power you want, and reasonable good cooling.

But a motor with similar ability to haul a load up a steep hill, in 26" wheel would be cool. Regular geared motors max out at about 300 pounds of load, the smaller dd's at 400. A geared that can handle 400 pounds up a long hill would be very cool, even if max speed was only 15 mph.

Yes I want to move to a VESC based controller. I'm hoping that with the inset magnets the motor will respond well to a proper FOC scheme, hopefully top speed/torque curve should be able to be extended with some field weakening.

dogman dan said:

About time they improved geared hub motors. Excuse my ignorance, but this is a lower gear inside right? So at lower wheel rpm the motor has a higher rpm than has been the case.

Does this affect top speed? or is it paired with faster spinning wind?

Click to expand...

Yes this has a higher reduction ratio than standard hubs, my old SWXH was 1:4.42, rated torque 32nm. This new G360 is 1:12.8, rated torque of 50nm.

Both motors run at similar speeds though, the SWXH was a 201 rpm@36volts, this G360 is 210rpm@36 volts. The actual rotor internally will be spinning much faster than previously though so no, top speed is unaffected as they run a faster wind).

I get the slowest wind motors I can and just throw voltage at them to get them into a higher speed range while maintaining the extra torque per amp. There are less conductor losses and you can run smaller controllers this way. Climbing hills would be the same case, get the torquey-est motor you can and increase voltage until you get to the speed you want.

The one issue will be the gears, the Bafang rating (on the SWXH at least) seems to be quite accurate, so you are limited in low down torque by them. If you don't mind having to run a larger and heavier DD then you will definitely be able to get more torque from one with ridiculous phase amps.