Mid drive ?

[tomjasz]

The only way I've been able to have real-time data on power use was by using a shunt and CA3 from Grin. Let me confirm that what is really happening was kinda shocking. For example, I saw that running a 52T Chainring on a BBS01 36V spiked wattage to a level that would seriously overheat the Bafang mid-drive. Sure enough, we've seen fried motors from BBS02 and BBS01 users and 52T rings. Once again that ugly screen offers more than any pretty color...

 

[tomjasz]

Unless you buy an expensive hub drive bike i doubt it will last long. I zip stores making the hub drive for my 2014 izip e3 dash, so I put a rear hub bfang 500watt motor kit on from Amazon for 500$ and it just blew on my after only 800 miles in like 4 months or riding to work and friends house. My dash lasted Bout 3500miles then I stopped riding for 3 years went back to ride it after 3 years and the motor is shot. PROBLY BECAUSE I LEFT IT IN MY SHED IN NJ WHERE IT GETS REALLY HOT AND COLD OUT SIDE THROUGH THE MONTHS.

Motors don't die from being stored. I have 5 bikes that are in a garage ranging from -10F to 105F. There's more to your failure story...

 

[Johnny]

I apologize in advance for the lengthy post...

I do agree that hub motors are better for some uses, mid-drives for some others but I think that motor efficiency vs speed needs a bit more clarification. I ran the Grin Motor Simulator for their GMAC hub motor. It's efficiency peaks at 84.5% near 26.7mph with a power draw of 652W. BTW - You'll need to tap on the charts to get the speeds referenced in this writeup.

Next I took a look at the mid-drive Bafang BBS02 with same battery, controller, and wheel size. Being a MDM, gear ratio is now a consideration as well. I used my ebike's gearing as an example, 46t chain ring and a 11x42 cassette. To consider higher speed efficiency I went with the 46x11 gearing. This setup reaches 81.6% efficiency at ~39mph with a power draw of 449W. An efficiency edge to the hub motor of ~3 percentage points.

But what about absolute power draw? The hub motor is operating at 84.5% of its design efficiency, but needs 652W of power. Even though the MDM is operating at a lower 81.6% of its design efficiency it needs just 449W. There are lots of other motors that could be modeled but the point is that operating efficiency is not necessairly the same as minimum power usage. I personally have noticed this between my original DIY hub bike vs my replacement MDM. I use less battery power for the same rides on the MDM bike than I did on the hub drive bike.

Now, what about other speeds? I often ride in the 15mph range around town. Using the same Simulator setups but adjusting speed by tapping on the chart the hub motor still has an efficiency advantage but the power draw is now ~876W vs ~673W, a substantial 200W savings with the MDM.

YRMV but it is important to consider motor power draw. Just sayin'

I think you may be making a reading mistake in those curves.
First of all, at the steady state, what I mean is you are keeping your speed no acceleration or deceleration, the higher the efficiency the less power has to be used. In the simulator that is where the black load line and the red power line intersects. Anywhere else the bike is whether accelerating or decelerating(that is why you are mistaken about the consumed power).

You should play with gearing to pull the mid drive to the same speed as the hub motor at that intersection so that you see what the efficiencies at the best case will be. I changed the gearing to 46x15 , now mid drive and hub motors you have posted are steady around 26-26.2mph and both are consuming around 680W while mid drive is around %80 efficiency , hub is %84.4 hence mid drive in this case is consuming around 850 while hub is around 805-810, 40-45W difference.

 

[B_Money_]

Motors don't die from being stored. I have 5 bikes that are in a garage ranging from -10F to 105F. There's more to your failure story...

Sure of that?

 

[tomjasz]

Sure of that?

Yes...

 

[linklemming]

Hub-drive motor acts only on the rear wheel. Mid-drive motor acts on the chainring, and as I wrote before, the drive-train multiplies or divides combined torque of the rider/motor from the chainring to the rear wheel. It gives completely different riding experience. Once, I made a small calculation and it turned out the Yamaha PW-X2 motor working over the 36/51 gearing produces 120 Nm at the rear wheel alone. Add to it the rider's own torque, also amplified.

P.S. That's why a hub-drive motor e-bike requires far more power for the same climbing capability than the mid-drive one. More powerful motor is heavier and requires far larger (heavier) battery for the same range. That's why all e-MTB from big name brands (that may actually be small companies) are mid-driven.

For those who want to see the torque multiplication/division for the pw-x and 10-51t 12 speed shimano cassette, here goes

and to put these number into perspective at a cadence of 90rpm

Im curious how much you use that 51t gear? I use a 50t on my acoustic FS MTB (32t front) only on one or two hills and its so steep that once I stop, I have to walk up the rest of the way.

I have only encountered torque multiplication once on my brose eMTB in 2 years(it was so steep I couldnt continue either going up or down and my feet would come out of my stiff sole cycling shoes when I attempted to go up). I have ridden this bike over 6k miles and it easily sees inclines >30%

Please dont 'over generalize' hub motors. You are attributing traits to hub motors that really only apply to cadence based systems. Should I over generalize all mid drives based on bbs02 or bbshd? Of course not, that would be silly.

Your 'over generalization' of "P.S. That's why a hub-drive motor e-bike requires far more power for the same climbing capability than the mid-drive one. More powerful motor is heavier and requires far larger (heavier) battery for the same range. That's why all e-MTB from big name brands (that may actually be small companies) are mid-driven." really only applies for more extreme grades and is more inline with DD hub drives than gear drives.

My DIY GMAC motor ebike is faster, weighs less with 650wh battery and consumes around the same wh per ride (typically 25 miles and 1500ft elevation) than my two brose bikes, The torque sensor is very good although not up to the brose standard and really doesnt matter much for the higher speeds I use it for. That being said, for the rides I do with it, its perfect (which is the whole reason I built it).

Hub drives and mid-drives each have their place. Lower speeds and steeper hills favor mid drives. Higher speeds favor hub drives.

 

[tomjasz]

Higher speeds favor hub drives.

Unless it’s a lightningrods mid.

 

[tomjasz]

My DIY GMAC

I missed that you built around this motor! I'd like to read more of your impressions!

 

[Johnny]

and to put these number into perspective at a cadence of 90rpm

Good computation, one minor detail though I doubt that Yamaha can output 80nm at 90rpm for a long time since it translates into 750W, that may even be over its stated peak value. It will not keep that up for a long time if at all and it may also be quite inefficient at that output.
That touted 250W nominal will give 80nm at a cadence of 30rpm at best.

 

[Manu]

I have a 36t/48t haibike trekking 2016 yamaha PW (70nm) and 11/36 cassette/cog.

PWX (2018) and PWX2(2020)( 80nm) more power,more power with high rpm.

unlocked 25km / h

48t x 11 x 68 rpm(turn) =40,4 km/h
36t x 11 x 90 rpm= 40,1 km/h

48t X 11 x 76 rpm=45,1km/h

500w battery discharges fast


68 turns per 60 seconds = 40.4 km / h

1 turn by second 60 turns per 60 seconds = 35,6 km/h

Lucky the owner of a trail or full fat haibike with double chainring 32/44 PW or PWX PWX "2..because Haibike has left these without a double chainring in this year 2021 ..... that double chainring unlocked with large teeth will give you allows you to go at a low cadence at a very good speed ........


My trekking haibike is not for sale

my trekking haibike 2016 25km / h is better than 2021

 

[linklemming]

Good computation, one minor detail though I doubt that Yamaha can output 80nm at 90rpm for a long time since it translates into 750W, that may even be over its stated peak value. It will not keep that up for a long time if at all and it may also be quite inefficient at that output.
That touted 250W nominal will give 80nm at a cadence of 30rpm at best.

Yes, very good points. Who knows under what conditions that 80nm applies.

No questions that mid-drives excel at going up hills. The steeper the better. For eMTBs its not even debatable. The only thing stopping me going up a steep hill with my Brose eMTB(34tFront/11-42tRear) is technical ability. The lower the gear the better it feels.

That being said, I have ever only experienced 'torque multiplication' once. My 11-42 slx cassette is 11-13-15-17-19-21-24-28-32-37-42. I have never used the largest gear and the 37t only once and I regularly go up stuff >30%.

I doubt someone with a Trance and a 75mm riser stem even has the ability to go up something steep enough to warrant a 51t rear cog (no weight on the front end). Perhaps ride data needs to be provided

To say that mid-drives are the best for everything and why all the big name manufacturers are using it is just absurd

I also have a mid-drive Brose speed pedelec bike (iZip Moda E3). Its a great bike but really starts to top out about 23mph and the battery really drains fast above 23mph (Moreso than hub drives I have used).

Select the best tool for the job

 

[linklemming]

I missed that you built around this motor! I'd like to read more of your impressions!

LOVE IT.

I have a thread in the DIY area

 

[Sierratim]

Sierratim, working from a practical perspective, something's broken.

I think you may be making a reading mistake in those curves..

Now I see, said the blind man.

Thank to you both for pointing out the error in my ways. I had used the Grin Motor Simulator extensively some years ago, but was obviously in need of a refresher before commenting on its results now.

I think this Simulator output is a correct comparison of the Bafang MDM and GRIN hub motor at a speed of 15mph with batteries delivering 206W & 197W for efficiencies of 75% & 79%, respectively.

I have added a note to my original post noting that my original conclusions are not correct.

 

[linklemming]

For those interested in playing with the grin motor simulator and trying mid drive vs hub drive, there is a really good thread on endless sphere about it.

How to use the motor simulator isnt always obvious (Im still pretty clueless)

https://endless-sphere.com/forums/viewtopic.php?t=100461

 

[Manu]

If I have a yamaha PW mid-drive motor and I'm telling you that it gives 40/45 km / h, a more advanced motor gives those same or higher cadences more torque and more rpm with more torque than my PW.


I have had to use my 11/36 completely and with the highest assistance mode, having a 10/50 ratio is very useful on a technical climb and when you are climbing without battery power.(25 kilograms)

 

[Sierratim]

Good computation, one minor detail though I doubt that Yamaha can output 80nm at 90rpm for a long time since it translates into 750W, that may even be over its stated peak value. It will not keep that up for a long time if at all and it may also be quite inefficient at that output.
That touted 250W nominal will give 80nm at a cadence of 30rpm at best.

This comparison chart is a couple of years old and doesn't include the latest motor updates such as Yamaha's PW-X2 but it's still useful in comparing mid-drive torque vs cadence. My understanding is that the PW-X2 motpr extends its peak torque curve out past a cadence of 100rpm.

When pushed to their limits these mid-drives can eventually overheat. The manufacturers don't publish this data but the Grin Simulator estimates that the Bafang BBS02 will take 28 minutes to overheat at full throttle and no speed limiter (28.6mph). With a 28mph speed limiter (90% throttle) for a Class 3 ebike the motor does not overheat.

I can't speak to the Yamaha motor, but I have ridden my Brose motor (Specialized 1.3) at its peak torque/power output for relatively long periods of time with no over heating issues. The bike's torque sensor reports a peak power output of 800W under these conditions.

 

[linklemming]

If I have a yamaha PW mid-drive motor and I'm telling you that it gives 40/45 km / h, a more advanced motor gives those same or higher cadences more torque and more rpm with more torque than my PW.


I have had to use my 11/36 completely and with the highest assistance mode, having a 10/50 ratio is very useful on a technical climb and when you are climbing without battery power.(25 kilograms)

No arguments there, my Brose Speed Pedelec will do 28mph, just takes ALOT more effort than my higher wattage hub drives.

I also use a 50t on my acoustic FS MTB, no debate there. To require it while under assistance with an eMTB would be quite a steep hill indeed and would likely exceed my MTB technical abilities.

Is a unpowered hub drive less efficient than an unpowered mid-drive

 

[AHicks]

Now I see, said the blind man.

Thank to you both for pointing out the error in my ways. I had used the Grin Motor Simulator extensively some years ago, but was obviously in need of a refresher before commenting on its results now.

I think this Simulator output is a correct comparison of the Bafang MDM and GRIN hub motor at a speed of 15mph with batteries delivering 206W & 197W for efficiencies of 75% & 79%, respectively.

I have added a note to my original post noting that my original conclusions are not correct.

Now 'yer talkin! Both the BBSHD and the the Grin/MAC hub motors do a nice job without being a complete pig when it comes to battery life - as long as you aren't trying to make them go fast...

 

[FlatSix911]

For those who want to see the torque multiplication/division for the pw-x and 10-51t 12 speed shimano cassette, here goesView attachment 76047
and to put these number into perspective at a cadence of 90rpmView attachment 76048
Im curious how much you use that 51t gear? I use a 50t on my acoustic FS MTB (32t front) only on one or two hills and its so steep that once I stop, I have to walk up the rest of the way.
I have only encountered torque multiplication once on my brose eMTB in 2 years(it was so steep I couldnt continue either going up or down and my feet would come out of my stiff sole cycling shoes when I attempted to go up). I have ridden this bike over 6k miles and it easily sees inclines >30% Please dont 'over generalize' hub motors. You are attributing traits to hub motors that really only apply to cadence based systems. Should I over generalize all mid drives based on bbs02 or bbshd? Of course not, that would be silly. Your 'over generalization' of "P.S. That's why a hub-drive motor e-bike requires far more power for the same climbing capability than the mid-drive one. More powerful motor is heavier and requires far larger (heavier) battery for the same range. That's why all e-MTB from big name brands (that may actually be small companies) are mid-driven." really only applies for more extreme grades and is more inline with DD hub drives than gear drives.
My DIY GMAC motor ebike is faster, weighs less with 650wh battery and consumes around the same wh per ride (typically 25 miles and 1500ft elevation) than my two brose bikes, The torque sensor is very good although not up to the brose standard and really doesnt matter much for the higher speeds I use it for. That being said, for the rides I do with it, its perfect (which is the whole reason I built it).
Hub drives and mid-drives each have their place. Lower speeds and steeper hills favor mid drives. Higher speeds favor hub drives.

I think this chart sums it up quite nicely.

 

[linklemming]

I think this chart sums it up quite nicely.
View attachment 76068

Thats awesome, where did you get that graph from.

That being said, the efficiency of the hub drives(geared and DD) would depend on the KV rating of the motors(rpm/volt). The mid-drive, notsomuch as you can control the motor speed thru gearing which is a HUGE advantage of mid-drives (gearing keeping the motor in its most efficient rpm range).