Ebike Motor Power Curve Comparison Chart: BOSCH, Yamaha, Shimano, Brose

No problem. I usually just leave it in SPORT when off road (don't have eMTB update), because I see very small gains going up to TURBO mode from SPORT, and yet it's still quite powerful when I need it so I usually don't get stuck on easy and intermediate trails.

I can see where eMTB allowing up to 300% assist can come in handy at times if I was riding more difficult trails, while allowing the bike to throttle down the pedal assist when the rider isn't leaning hard on the pedals. When some people are riding difficult trails and leave it in TURBO the entire time, I can see where eMTB can save them some battery life when full power isn't needed.

On tarmac I'm usually okay with ECO and TOUR, unless I have a long steep climb ahead - then I bump it up for a short period until I can get my cadence up enough to maintain speed in TOUR if possible. If I had eMTB mode then I couldn't limit power assist to only 200% when 120% in TOUR isn't enough, so I'd worry about either losing range (vs SPORT) or about getting less of a workout because of the higher assist level.

I did notice recently that when I'm cruising on flat ground with the Powerfly 7, at 40rpm in top gear in ECO or TOUR mode, that I don't have to push very hard on the pedals - but then I encounter a mild climb where kicking the assist up to TURBO (from SPORT) can't always prevent me from having to shift down to a lower gear for a higher cadence.

Low-end power in top gear is pretty good, until I have to climb a grade that's more than 3-4%. Then I find it's required to down-shift from top gear to achieve a higher cadence to power through it, even in TURBO mode. The Bosche really sings at higher RPM (guessing 60-90RPM), regardless of assist level. But at 20MPH in top gear, sometimes the cadence isn't high enough to get max power for steep climbs.

For instance, at 20MPH in top gear I encounter many steep hills that are too difficult to maintain that speed in TURBO, even when standing and mashing on the pedals as hard as I can. I'll see my speed drop from 20 > 19 >18 > 16 and I'll have to downshift just to keep it going at 14-15MPH at 60-90RPM without losing even more speed.

On the other hand, in the lower gears the Bosche isn't very cadence sensitive, and it does fine with wide range of RPMS (maybe from 40-90 for me). In 1st gear I can still lug the bike down to 40rpm and crawl up a steep hill in TURBO like a tractor, because the lower gears are torque multipliers.

But in the highest gears it really seems to lose power when you climb a grade more than just a few degrees, no matter how much assist you select. I imagine these charts for measured power were in a lower torque multiplying gear ratio.
Here's a video comparing 2 ebikes with Yamaha pw-x (the supposedly high cadence version) and Bosch CX. Notice the yamaha has slower cadence compared to Bosch and is slower overall. This proves that human effort (higher cadence = higher output) contributes to the overall power and helps in the performance of the ebike.
The Bosch is faster and you also do more work on the Bosch.
Last edited:
Cadence lowest for the plate of more teeth in yamaha, yamaha is 250w vs 350w, 100w less in yamaha and more kilometers with the same battery watts.

you want more cadence in yamaha is to use cassette with more teeth. The yamaha assists from speed zero or second 0, you put higt step on upphill and assitsts from second 0.
250W divided 50km/h=5 w by 1 unit of kilometer hour.
350W divided 50km/h=7w by 1 unit of kilometer hour.350 w engine but reaches 550 w

That comparison could be the equivalent of a speedpedelec as the sduro trekking 45km / h 500W engine unlocked to 75km / h vs 350w unlocked to 75Km/h.
500w divided 50Km/h=10 w by unit of kilometer hour.
Minute 1.10 seconds Minute 1.34 seconds......../.led panel yamaha is green =mode eco /eco+/
Mode PWXE is yellow only in PWXE engine 80netow and 120rpm./BLue led is High and STD.
I see green led
Do not use the cassette in yamaha.Starts on 13 and 11 cassette teeth .....has a plate with more teeth so slow the start.
Last edited:
I think it is reasonable to assume the source was Bosch since their branding is clearly indicated on the chart while the other motor company trademarks are lacking.
They are all using their trademarked typeface. That being said, it is probably Bosch due to it showing the most power. But, I would want E8000 due to more compact motor and more battery integration.
The curves are a piece of data. Where the corresponding torque curves ? Either way, you start talking detailed specs like this, and 99% of the customer's eyes will simply glaze over. I guess its cute though that the techno geeks here talk about it. o_O
I'm waiting for the model that bypasses a motor altogether, and puts out 1.21 Jigawatts (gigawatts) via a flux capacitor. :)
One fundamental difference between Bosch and Yamaha is that you can pedal a Yamaha with the battery off. You cannot pedal a Bosch... well, you can, but it's extremely painful, the resistance from the motor is huge, the sprocket is tiny, so if you are out of juice you are well and truly doomed.
When looking at the power curves, it is not clear if this is peak or continuous power. My understanding is that in EU, continuous power is used to determine the power rating of the motor, not peak. I would want to know that all motors were tested at peak or continuous power. Perhaps the Brose is continuous power.
How does Wattage equate to Torque? It seems to me that the two are not exactly correlated

Power and torque are directly correlated at a known rpm: Power = Torque * angular velocity. Angular velocity is rpm in radians/second, so

Power [W] = Torque [Nm] * 2*pi * rpm / 60

Using that simple equation, I reverse engineered the original power curves to visualize the torque curves, the results look like this:

Power and torque are directly correlated at a known rpm: Power = Torque * angular velocity. Angular velocity is rpm in radians/second, so

Power [W] = Torque [Nm] * 2*pi * rpm / 60

Using that simple equation, I reverse engineered the original power curves to visualize the torque curves, the results look like this:

View attachment 21833
Great presentation! This will give you a more tangible idea of how much "pushing force" you get from the motors at different RPM's (or cadence). Note that these are at rated capacities, so the these are all in the highest assist levels.
The Power vs Revolutions curve is not a useful motor power curve , such a motor curve does not exist. It is only a matter of software control.
To compare motors, a graph of efficiency vs torque is required.
See here for more:

You have a point there but I think the graph shown were actual dynamometer performance of the motors at full output. These are "mechanical power" output as contrasted to "electrical power" input that you usually see on many ebike's powermeter display.

As you would see on this simulator below, at 20 mph, the electrical input is 406 watts (right yellow rectangle) , however, the actual power ourput from the motor is 337 watts (left yellow rectangle) with 83% efficiency (green figure).

If the grade is increased to 16 percent, the speed decreases to 5 mph and the electrical input increases to 719 watts, while the motor output is only 371 watts, with low efficiency of 51.6%.

One interesting thing to take note, the electrical input in amps is almost 1:1 correlation to the mechanical output in torque (Newton-meter), esp at the high efficiency rpm range. This is observed by Justin Lemire as he was dyno testing countless ebike motors and eventually came up with his ebike simulator.
Last edited:
Power and torque are directly correlated at a known rpm: Power = Torque * angular velocity. Angular velocity is rpm in radians/second, so

Power [W] = Torque [Nm] * 2*pi * rpm / 60

Using that simple equation, I reverse engineered the original power curves to visualize the torque curves, the results look like this:

View attachment 21833


By looking at the torque or the twisting force for each motor, it shows that 2 of the motors lose strength early in the cadence range.

The Brose motor's torque starts dropping above 30 RPM, so the highly touted 90 NM torque is only found at the lower end of the cadence range.

The Yamaha motor's torque drops steeply above 75 RPM. At that cadence, much of the torque to the rear wheel comes from your own legs.

This is most likely extracted from a dynamometer reading since the figures are about 15% lower than the advertised numbers. Dyno figures are usually lower since there are efficiency loses at the drivetrain/transmission and the wheels.

By the way, I saw the torque curve of the Bafang BBS and it is similar to the Brose.

Most hub motors also have similar torque profile as the Brose (at full output).
Last edited:
This graph is very manipulated from base in Newtown because to begin with it does not even show 90 nm, the graph of the Yamaha is a pw(100rpm), which does not pw-se (110 rpm)and pwx(120 rpm).The technical characteristics of the manufacturer must be respected, if it says that they are 90 nm are 90 nm in the case of the brose in the graph, if it says 80 nm is 80 and not 70.

The manufacturer respects the numerical values expressed and in some cases even give 10% more than what is expressed in the table of characteristics.

All these drive units are considered to be full
bosch cx 75 nm
yamaha pw/pw-se/pwx 80nm (100rpm) (110rpm) (120rpm)
Brose 90nm
shimano 70 nm
would not compare a 250 W drive unit with a 250 W hub because the drive unit carries a multiplier that allows you to climb mountain and multiply the torque
Last edited:
Yeah, the PW-X assists up to 120 rpm but at what torque? Instead of fading the torque steeply from 75 rpm (in PW), the torque fades away more gently so that the bottom can reach 120 rpm. But still, the torque starts to drop early on (even less than 75 rpm in pw-x)).


The Brose nominal and peak power are the same at roughly 350 watts.

The peak power of Bosch and Shimano (as well as the Yamaha) are way higher than the rated power or nominal power.

The performance curve of the shmano looks like the little brother (8/10th version) of the Bosch CX.

Last edited:
The Pwx has 5 modes, the PW / PW-SE has 4. The Pwx those 5 modes assists up to 120 rpm. The eco +, echo, standard and hight modes of the PW / PW-SE and the Expw only to PWX.
Each mode is assigned a maximum level of consumption in watts and that is the maximum torque delivered.
Eco + = 75w 30 nm?
ECo = 150w 45nm?
Standar = 250w 70nm?
The hight mode of the PW is = 440 real watts 80nm.
Expw = 440 ++ watts is more than 80nm .....¿550....600?

The graph you have in green is from a PW that is not PW-SE (110rpm) or PWX (120rpm).

The more watts the drive unit consumes, the more torque it delivers, for that are the controllers to avoid exceeding the mechanical electrical limit of the drive unit.
you with an eco mode you're never going to get 80nm.
Last edited:
Motorcycle engines in sport bikes tend to have peaky powerbands. That narrowness of energy creates a ride where you keep the engine revs boiling while you tap the shifter to maintain the rpm. Good fun with a motorcycle on a twisty backroad. Mid-drive PAS eBikes with the same narrow peak power range are not as fun because it’s your legs that keep the crank spinning at the required rpm. As terrain varies, so does your need to spin. Motorcycles best suited to everyday commuting have flatter powerbands. That applies to eBikes, too. A mid-drive that provides a useable level of assist over a wide rpm range is going to be less frenetic. Having a Yamaha or Bosch PL system, as reflected in that chart, would drive me nuts.

On the whole, I prefer mid-drive to hub motors, but the need to shift less frequently with the hub drive is nice - unless, of course, the rider enjoys shifting. The rising and falling whirring of the Bosch motor is reminiscent to shifting an internal combustion powered conveyance with lots of gears. A little less shifting and whirring while maintaining the balance and good torque sensing of the mid-drive would be nice.
Ebike Motor Power Curve Comparison Chart. This test result was leaked recently and has been a topic of conversation in the ebike world for the past few weeks. It shows the power curves of Bosch Performance, Bosch Percormance CX, Shimano, Yamaha and Brose motors.

I was at a dealer event for Moustache ebikes last month in France and this chart was referenced. Good data. I understand it was a leaked chart from one of the motor manufacturers.



View attachment 17593
The new Haibike have a 120nm TQ motor found here: