The higher voltage reduces the current draw slightly current (A) = Power/ Voltage
... so slightly less demanding on the individual cells (if the battery has the same number of cells per parallel group)
Ebike Battery Pack Energy Density Talk
- Thread starter emtbeast
- Start date
... so slightly less demanding on the individual cells (if the battery has the same number of cells per parallel group)
- Thread starter
- #32
Yes they did a power/cadence curve graph also if I remember correctly, I would rather see a torque/cadence curve, think that tells a better picture, asked why they didn't include that(as it's definitely been measured, they did measurments in a the same lab that also does test for a paper magazine that posted such graphs) got no answer unfortunately.
Yes the weight difference is negligible, but it's there, my point was mainly heat, going of my Reign and buddy's Levo, the bikes motors in extreme cases(full power, hot outside temperatures, steep inclines) heat up almost to the point where you could fry eggs, I do think that even if a small difference in these extremes it's definitely noticeable. Motor magnets exposed to high heat(over 80°C) loose some magnetism momentarily and if exposed longer can also lead to a permanent loos of some magnetism. All this leads to decreased efficiency. I think that any step towards better efficiency is good.
That's interesting EU regs separate the voltage considering AC is measured in RMS, which is the DC equivalent. If you measure it without the RMS calculation you'd be surprised how high it is! On 460vac system a non RMS measurement will be over 600volts.
harrysmalls
Member
To your point about Magnesium battery enclosure. Have you considered different manufacturing/machining methods used compared to Aluminum? As far as I know, that aspect alone is significantly more expensive. Also, increased fire hazard, as Magnesium would burn in a battery fire.
- Thread starter
- #35
Hey, yes I did, but as said I don't work in the industry that's why I also wrote, maybe someone from the industry can elaborate further about what magnesium isn't used more.
I am also aware of fire hazards present with magnesium, however, magnesium is only flammable in their pure form when molten or in powder or in ribbon form.
The magnesium used for everyday use is usually an alloy, not pure.
If I'm not mistaken Bosch, Giant(PW-XM), Brose use magnesium alloys for ebike motor housings, so it's doable I guess.
I am also aware of fire hazards present with magnesium, however, magnesium is only flammable in their pure form when molten or in powder or in ribbon form.
The magnesium used for everyday use is usually an alloy, not pure.
If I'm not mistaken Bosch, Giant(PW-XM), Brose use magnesium alloys for ebike motor housings, so it's doable I guess.
In the 60s and 70s Porsche made their engine crank cases (and gearbox housings) out of magnesium for weight saving purposes. With 50 years of aging as proof (the engine cases not me!!!) the magnesium cases corrode and the screw threads can pull out. For repair purposes magnesium is VERY difficult to weld repair. In summary aluminium is heavier but tougher. For ebike motors, I would have thought magnesium would be the preferred material.Hey, yes I did, but as said I don't work in the industry that's why I also wrote, maybe someone from the industry can elaborate further about what magnesium isn't used more.
I am also aware of fire hazards present with magnesium, however, magnesium is only flammable in their pure form when molten or in powder or in ribbon form.
The magnesium used for everyday use is usually an alloy, not pure.
If I'm not mistaken Bosch, Giant(PW-XM), Brose use magnesium alloys for ebike motor housings, so it's doable I guess.
- Thread starter
- #37
So for a simple ebike battery housing it's totally acceptable to use a magnesium alloy.
Yes, I would think for a simple [battery?, or motor?] housing it would be fine. I'd imagine that the dimensions of the bearing bores would have to be very slightly modified (compared to an aluminium equivalent) because of the different thermal expansion. The only other consideration for a motor housing is the hammering it can potentially take- if bottomed out on rocks etc... for this purpose aluminium is tougher.
- Thread starter
- #39
That's why I have probably seen some pics of broken cooling fins on the Bosh Magnesium motor housing, haven't seen a full broken one from impacts. I guess it's totally fine to use a magnesium alloy for ebike parts, otherwise they wouldn't do it.
I guess from the manufacturer's perspective it's a weight saving of a few grams which makes it easier to market!
- Thread starter
- #41
An example:
Yamaha PW-X3(Alu) 2,75kg vs PW-XM(Mag) 2,6kg. That's a weight saving of 150g( cca 5,5%), I wouldn't call that a few grams.
Also magnesium alloys have better cooling properties, one other advantage and point for marketing.
TheKaiser
Active member
I know this is a thread about batteries, but I saw you mentioned DJI trying to keep "under the regulatory EU (250W average)" and you guys clearly are super technically well versed, so I was wondering if you know/could explain to me how these motor specs work? For example, Bosch is touting motors with 600w peak output, but is there a finite time limit imposed by the law on how long that peak power can be sustained? If they are 250w nominal, or average, how is that average determined? It would seem like they would have to have some software tracking that in the motor, but I am not aware of people complaining about sudden power cuts to meet regs. I ask because it seems that end user behavior and chosen settings vary so widely, I can't see how they would ever know what ended up being the "average" power used for a given user, so I'm assuming there is some standardized formula or system. I also know some motors do actually offer a finite time period "Boost" mode, which you have to manually activate and which only lasts for 30 sec, but Bosch 600w figures are not in that category.
- Thread starter
- #43
Hey, no worries,
I will try even though it's a bit complex...
For starters a bit about electric motor power, as I see there seem to be some misconceptions about how electric DC motor power is determined/measured.
The power output of an electric motor is a complex measurment:
P(Watt) = V(Volt) * I(Ampere)is only true in special circumstances eg when the load is a resistor.
It is not true in cases of an electric motor, whose real output depends on its Q factor(motor quality factor), switching control, load and also with ebikes how the energy source, battery in this case and its BMS(battery management system) is set up.
As the battery is not a stable source of energy delivery, it's a lot on the BMS setup, how the motor works throughout the whole battery SOC(state of charge).
A DC electric motor gets pushed around its axle by solenoids, whose voltage across their terminals varies with time, the intensity of the current flowing through it is also a function of time, the output power is also a function of time with more variables than just V (voltage) and I (current Intensity).
Each function has also its harmonics, the resultant P is a complex function and it can be only measured in a lab environment.
The motor power is usually measured while the motor reaches its thermal equilibrium(same temp as the environment) in a room at 25°C.
Then with ebike motors there's also the factor of internal transmission that needs to be taken into consideration when looking at the final crank output numbers, quite a few lab tests show a few % lower torque figures than the actual factory specs say.
The 250W EU limit 》》》
Like u and I was confused for some time(especially with rising ebike power numbers) I see a lot of people are also confused about the 250W Power limit in EU. Not sure this will be a good enough explanation as it's questionable, considering all the EU regulations etc...but for me it was good enough(cha bu duo).
So I did a little bit of online research about the European legislation EN15194, it is the standard that is most commonly applied. I didn't go read the whole documentation as it's a handful but found a recap of it online. Maybe the bottom description will clears things up why such differences between brands.>>>>
The recap says, there's effectively no limit in the European legislation EN15194, which It's all about the words and test methods, neither of which pin it down. The rules are that the manufacturer must not label the motor with more than 250w and a motor with a 250w label mustn't overheat when running at 250w continuous. The manufacturers have to be careful about terminology and calculations.
Hope this helps.
I will try even though it's a bit complex...
For starters a bit about electric motor power, as I see there seem to be some misconceptions about how electric DC motor power is determined/measured.
The power output of an electric motor is a complex measurment:
P(Watt) = V(Volt) * I(Ampere)is only true in special circumstances eg when the load is a resistor.
It is not true in cases of an electric motor, whose real output depends on its Q factor(motor quality factor), switching control, load and also with ebikes how the energy source, battery in this case and its BMS(battery management system) is set up.
As the battery is not a stable source of energy delivery, it's a lot on the BMS setup, how the motor works throughout the whole battery SOC(state of charge).
A DC electric motor gets pushed around its axle by solenoids, whose voltage across their terminals varies with time, the intensity of the current flowing through it is also a function of time, the output power is also a function of time with more variables than just V (voltage) and I (current Intensity).
Each function has also its harmonics, the resultant P is a complex function and it can be only measured in a lab environment.
The motor power is usually measured while the motor reaches its thermal equilibrium(same temp as the environment) in a room at 25°C.
Then with ebike motors there's also the factor of internal transmission that needs to be taken into consideration when looking at the final crank output numbers, quite a few lab tests show a few % lower torque figures than the actual factory specs say.
The 250W EU limit 》》》
Like u and I was confused for some time(especially with rising ebike power numbers) I see a lot of people are also confused about the 250W Power limit in EU. Not sure this will be a good enough explanation as it's questionable, considering all the EU regulations etc...but for me it was good enough(cha bu duo).
So I did a little bit of online research about the European legislation EN15194, it is the standard that is most commonly applied. I didn't go read the whole documentation as it's a handful but found a recap of it online. Maybe the bottom description will clears things up why such differences between brands.>>>>
The recap says, there's effectively no limit in the European legislation EN15194, which It's all about the words and test methods, neither of which pin it down. The rules are that the manufacturer must not label the motor with more than 250w and a motor with a 250w label mustn't overheat when running at 250w continuous. The manufacturers have to be careful about terminology and calculations.
Hope this helps.
- Thread starter
- #45
I get that oftenTheKaiser
Active member
That is very helpful, thanks! So, if I understand correctly from your last paragraph, in the EU there is no limit on the motor's output. The only requirement is that it not say more than 250w on the label, and it has to work correctly (not overheat) at 250w. That is pretty wild, but it explains why, while I often see people using the term "average" to refer to the 250w figure, I also see people refer to it as 250w "nominal". I always thought that nominal was an interesting word to use, and now I see why!Hey, no worries,
I will try even though it's a bit complex...
For starters a bit about electric motor power, as I see there seem to be some misconceptions about how electric DC motor power is determined/measured.
The power output of an electric motor is a complex measurment:
P(Watt) = V(Volt) * I(Ampere)is only true in special circumstances eg when the load is a resistor.
It is not true in cases of an electric motor, whose real output depends on its Q factor(motor quality factor), switching control, load and also with ebikes how the energy source, battery in this case and its BMS(battery management system) is set up.
As the battery is not a stable source of energy delivery, it's a lot on the BMS setup, how the motor works throughout the whole battery SOC(state of charge).
A DC electric motor gets pushed around its axle by solenoids, whose voltage across their terminals varies with time, the intensity of the current flowing through it is also a function of time, the output power is also a function of time with more variables than just V (voltage) and I (current Intensity).
Each function has also its harmonics, the resultant P is a complex function and it can be only measured in a lab environment.
The motor power is usually measured while the motor reaches its thermal equilibrium(same temp as the environment) in a room at 25°C.
Then with ebike motors there's also the factor of internal transmission that needs to be taken into consideration when looking at the final crank output numbers, quite a few lab tests show a few % lower torque figures than the actual factory specs say.
The 250W EU limit 》》》
Like u and I was confused for some time(especially with rising ebike power numbers) I see a lot of people are also confused about the 250W Power limit in EU. Not sure this will be a good enough explanation as it's questionable, considering all the EU regulations etc...but for me it was good enough(cha bu duo).
So I did a little bit of online research about the European legislation EN15194, it is the standard that is most commonly applied. I didn't go read the whole documentation as it's a handful but found a recap of it online. Maybe the bottom description will clears things up why such differences between brands.>>>>
The recap says, there's effectively no limit in the European legislation EN15194, which It's all about the words and test methods, neither of which pin it down. The rules are that the manufacturer must not label the motor with more than 250w and a motor with a 250w label mustn't overheat when running at 250w continuous. The manufacturers have to be careful about terminology and calculations.
Hope this helps.
You also brought up something else, which has been a longstanding question of mine, although I was not sure what term to use for it. It's this "Q" (Quality) factor (not to be confused with the other Q-factor that refers to pedal stance width), which I always found myself calling "motor efficiency" or something like that. and perhaps you can help clarify how much it matters, and if there are any standardized measures of it.
If we have, just for the sake of argument, a 500wh battery, and we have a motor, running at 250w continuous, it seems at first glance that one would get 2hr of use before the battery is drained. But is that 250w typically measured as the input to the motor from the battery, or downstream of the motor as the output from the motor? I suspect it's the input to the motor, and I think that's what you were getting at when you wrote "...the factor of internal transmission that needs to be taken into consideration when looking at the final crank output numbers, quite a few lab tests show a few % lower torque figures than the actual factory specs say." Those few % lower torque figures (and I'm assuming power too) represent the frictional losses (both mechanical and electrical) of the motor system.
So I guess the question I've been puzzling over is, how much do these motors vary in terms of their internal losses? I see some people refer to one motor as being more "thirsty" for battery juice than another, but is there any clear objective measure that you are aware of that would tell us which motors are more efficient (have a higher Q factor), in terms of their conversion of battery juice to forward motion?
I know that end user riding behavior, and the motor's settings, and elevation change on the ride, and the rider weight, etc... will make the biggest difference in actual real world milage, which is why I specified the motor running at 250w continuous. I'm looking for an "all else being equal" sort of measurement of energy input to the motor vs. energy output from the motor.
- Thread starter
- #47
Hey, I will try to answer to your 5 paragraphs in 5 points. I will try to use some online simplified explanations.
1. 250W Continuous
Well yes, mostly you understood correctly, the manufacturers must not label the motor more than 250W and the motor during testing must not overheat while continuously working at 250W.
The regulations on which the testing is done say how long the period of testing is, did not go into detail, but from memory I recall I read that the testing is done in 30min period.
Terminology 》》》this is a tricky one, not only here in ebike talk, but all over the modern tech world.
So yes I would avoid the term "average power", as it's completely wrong here. The term "nominal power" comes closer but imo also still not quite correct and my preferred term is "continuous power"
Nominal versus peak power. A motor’s nominal wattage is the maximum amount of power it can sustain for long periods of time, while the peak output is the wattage a motor is capable of for a very short burst. Peak output wattage is still a useful rating that gives you an idea of how hard a bike might accelerate or how it might feel on short bursts on a very steep hill.
Motors have thermal limits before they fail. Continuous and peak power are based upon those thermal limits. Continuous power can be used non stop. The cooling for the motor, whether it is a fan, natural convection, water cooling, is sufficient to keep the motor at a happy temperature. Peak power is driven more by the thermal mass and magnetic limitations of the motor and because the temperature here is rising faster than we can cool the motor, this can only last a limited amount of time.
2. Motor Q factor
This one is a bit technical...I tried to find a simplified explanation, hope this is clear enough. Q factor(also known as Quality Factor) is defined as a dimensionless parameter and it measures the performance of a coil(for our case this is it as the stator of the motor is a coil), a capacitor, or an inductor in terms of its losses and resonator bandwidth. The Q factor indicates energy loss in a design. Also, the Quality Factor (Q) may be defined as the ratio between stored energy and energy dissipated per cycle in a circuit.
While all electric motors probably have a Q factor label and it's efficiency is driven by it, this is a label only the engineers in the production use and for the end user it's a non existent spec. The end result of a system with a good q factor motor is seen in good efficiency at high loads.
3./4. System Losses
Yes you are correct with the 500Wh battery, 250W continuous power draw for 2h, If that is measured between the battery and the motor(the electrical power). Even here we have thermal losses(battery cells, wires, motor electronics,...) The BMS is the one the tries to control the energy supply and the motor controller distributes it accordingly to demand. How and where the numbers the labs get from testing is more a question for them, the important thing is when comparing numbers it should be from the same lab and same environment(several factors).
Yes the internals of the motor definitely slightly affect the end numbers, although most torqe difference probably comes from the max load at wich the testing is done, as at high power and rpm, the torque is dropping lower(P = F x Rpm).
The motors are tuned to a specific rpm for best efficiency, I they did the testing slightly out(over) of this window, the outcome can so be lower torque figures. This is why I always said, the manufacturers should focus on a wider torque band, especially for emtb motors. Unfortunately no, it's not possible from specs to read which motor will be more efficient, but there are tests, and from my experience, they tell a pretty accurate real life situation:
Screenshot - YT Velomotion Magazin
5. System Overall
Exactly, the system on it's own is just a part of the story when talking about efficiency of an ebike, two biggest enemies are rider/system weight, inclines, followed by tyre choice and pressure, terrain and environment temperatures.
Last edited:
TheKaiser
Active member
Wow, that is really great info, thank you so much for such a thorough reply! Also, thank you for posting that screenshot of the chart, as I have never seen a comparison of motor power consumption that objective before. It very perfectly illustrates how much power consumption goes up with the grade. Even more intriguingly, assuming they controlled for wattage, are these other trends that jumped out at me:Hey, I will try to answer to your 5 paragraphs in 5 points. I will try to use some online simplified explanations.
1. 250W Continuous
Well yes, mostly you understood correctly, the manufacturers must not label the motor more than 250W and the motor during testing must not overheat while continuously working at 250W.
The regulations on which the testing is done say how long the period of testing is, did not go into detail, but from memory I recall I read that the testing is done in 30min period.
Terminology 》》》this is a tricky one, not only here in ebike talk, but all over the modern tech world.
So yes I would avoid the term "average power", as it's completely wrong here. The term "nominal power" comes closer but imo also still not quite correct and my preferred term is "continuous power"
Nominal versus peak power. A motor’s nominal wattage is the maximum amount of power it can sustain for long periods of time, while the peak output is the wattage a motor is capable of for a very short burst. Peak output wattage is still a useful rating that gives you an idea of how hard a bike might accelerate or how it might feel on short bursts on a very steep hill.
Motors have thermal limits before they fail. Continuous and peak power are based upon those thermal limits. Continuous power can be used non stop. The cooling for the motor, whether it is a fan, natural convection, water cooling, is sufficient to keep the motor at a happy temperature. Peak power is driven more by the thermal mass and magnetic limitations of the motor and because the temperature here is rising faster than we can cool the motor, this can only last a limited amount of time.
2. Motor Q factor
This one is a bit technical...I tried to find a simplified explanation, hope this is clear enough. Q factor(also known as Quality Factor) is defined as a dimensionless parameter and it measures the performance of a coil(for our case this is it as the stator of the motor is a coil), a capacitor, or an inductor in terms of its losses and resonator bandwidth. The Q factor indicates energy loss in a design. Also, the Quality Factor (Q) may be defined as the ratio between stored energy and energy dissipated per cycle in a circuit.
While all electric motors probably have a Q factor label and it's efficiency is driven by it, this is a label only the engineers in the production use and for the end user it's a non existent spec. The end result of a system with a good q factor motor is seen in good efficiency at high loads.
3./4. System Losses
Yes you are correct with the 500Wh battery, 250W continuous power draw for 2h, If that is measured between the battery and the motor(the electrical power). Even here we have thermal losses(battery cells, wires, motor electronics,...) The BMS is the one the tries to control the energy supply and the motor controller distributes it accordingly to demand. How and where the numbers the labs get from testing is more a question for them, the important thing is when comparing numbers it should be from the same lab and same environment(several factors).
Yes the internals of the motor definitely slightly affect the end numbers, although most torqe difference probably comes from the max load at wich the testing is done, as at high power and rpm, the torque is dropping lower(P = F x Rpm).
The motors are tuned to a specific rpm for best efficiency, I they did the testing slightly out(over) of this window, the outcome can so be lower torque figures. This is why I always said, the manufacturers should focus on a wider torque band, especially for emtb motors. Unfortunately no, it's not possible from specs to read which motor will be more efficient, but there are tests, and from my experience, they tell a pretty accurate real life situation:
View attachment 151888
Screenshot - YT Velomotion Magazin
5. System Overall
Exactly, the system on it's own is just a part of the story when talking about efficiency of an ebike, two biggest enemies are rider/system weight, inclines, followed by tyre choice and pressure, terrain and environment temperatures.
1. At 0% grade, it seems that Bosch and Shimano are most efficient, at 4.8.-4.9Wh/km, and it is interesting to see how much more some of the other motors consume. For example, the DJI motor is using over 20% more power to go the same distance. That is a clear example of what one might consider to be a "thirsty" motor, at least on flat terrain.
2. At 10% grade, the picture can be somewhat different. For example, while all versions of the Bosch CX have the same power consumption at 0% grade, the new Gen5 version is about 12% more efficient than the other versions on a 10% grade. Similarly, the Giant and DJI motors both seem to be more efficient, relatively speaking, on a 10% grade, than their 0% grade power consumption would suggest. Even still, on the 10% grade, the DJI motor has fallen further behind the Bosch Gen 5, now using about 30% more power, so it's gotten even more "thirsty" as the grade went up.
That is really interesting data, as it could have a pretty dramatic real world effect on range. For example, I like the idea of doing big rides in the backcountry, and so range anxiety is a very real thing for me. I was bummed out to see that the new Santa Cruz Vala (Bosch Gen5) only comes equipped with a 600wh battery, as I like the bike otherwise, whereas the Specialized Turbo Levo (Custom Brose) uses a 700wh, and the Amflow DJI bike is available with an 800wh. As it turns out, all 3 of those bikes may have very similar range, given the above chart. And conversely, a bike with the new Bosch Gen5 that has their 800wh battery would blow all the other options away in terms of range. And, as luck would have it, my thread derailment still lead us back to the original topic of battery capacity, just like I'd planned!
- Thread starter
- #49
Wow, that is really great info, thank you so much for such a thorough reply! Also, thank you for posting that screenshot of the chart, as I have never seen a comparison of motor power consumption that objective before. It very perfectly illustrates how much power consumption goes up with the grade. Even more intriguingly, assuming they controlled for wattage, are these other trends that jumped out at me:
1. At 0% grade, it seems that Bosch and Shimano are most efficient, at 4.8.-4.9Wh/km, and it is interesting to see how much more some of the other motors consume. For example, the DJI motor is using over 20% more power to go the same distance. That is a clear example of what one might consider to be a "thirsty" motor, at least on flat terrain.
2. At 10% grade, the picture can be somewhat different. For example, while all versions of the Bosch CX have the same power consumption at 0% grade, the new Gen5 version is about 12% more efficient than the other versions on a 10% grade. Similarly, the Giant and DJI motors both seem to be more efficient, relatively speaking, on a 10% grade, than their 0% grade power consumption would suggest. Even still, on the 10% grade, the DJI motor has fallen further behind the Bosch Gen 5, now using about 30% more power, so it's gotten even more "thirsty" as the grade went up.
That is really interesting data, as it could have a pretty dramatic real world effect on range. For example, I like the idea of doing big rides in the backcountry, and so range anxiety is a very real thing for me. I was bummed out to see that the new Santa Cruz Vala (Bosch Gen5) only comes equipped with a 600wh battery, as I like the bike otherwise, whereas the Specialized Turbo Levo (Custom Brose) uses a 700wh, and the Amflow DJI bike is available with an 800wh. As it turns out, all 3 of those bikes may have very similar range, given the above chart. And conversely, a bike with the new Bosch Gen5 that has their 800wh battery would blow all the other options away in terms of range. And, as luck would have it, my thread derailment still lead us back to the original topic of battery capacity, just like I'd planned!
good work, yes I would also say the Bosch CX gen5 is most probably the most efficient system overall and combined with the 800Wh battery and 29er wheels it's a range monster.If you are fairly fit and relatively light(70-80kg) or less then the Giant's Syncdrive Pro 2 with the 800Wh is also pretty impressive IMO.
1. Attached is a screenshot of my longest ride. My bike is a mullet Reign E+1 800Wh 2022(26,5kg bike), 92kg(ride ready).
I had 24% battery remaining at the end.
I think it is possible to push 2400m of Altitude gain from the system with proper settings(lowering the torque to a max of 70Nm).
2. Here are screenshots from a yt test of the new Bosch CX gen5 on the 600Wh and the 800Wh everything in Turbo mode, only climbing is measured. It's pretty darn efficient, I dare to say it's the most efficient system on the market today(here the motor and system Q factor is really high).
600Wh(20,7km, 1605m Altitude gain)
800Wh(26,1km, 2132m Altitude gain)
3. Then here is are few test charts from lab tests. In these charts it's possible to see motor behavior(power/torque relationship)at different cadences, and due to losses we can see smaller torque figures than the actual motor specs.
The numbers are measured at the crank(so all the internal transmission losses and electrical losses including) torque/cadence curve(left power, right torque, bottom cadence)for the:
Sram Eagle(Brose)
Bosch CX gen4
Giant Syncdrive Pro 2
4. Then here is a power/cadence curve comparison of several motors.
Now I am derailing the thread
Well in a way it's linked to battery tech...
Last edited:
TheKaiser
Active member
No worries on the derailment, good info and thought provoking conversation are always welcome as far as I am concerned, plus as the OP/thread starter, I think international posting law grants you special privileges!
If you are fairly fit and relatively light(70-80kg) or less then the Giant's Syncdrive Pro 2 with the 800Wh is also pretty impressive IMO.
1. Attached is a screenshot of my longest ride. My bike is a mullet Reign E+1 800Wh 2022(26,5kg bike), 92kg(ride ready).
I had 24% battery remaining at the end.
View attachment 151953View attachment 151954
I think it is possible to push 2400m of Altitude gain from the system with proper settings(lowering the torque to a max of 70Nm).
2. Here are screenshots from a yt test of the new Bosch CX gen5 on the 600Wh and the 800Wh everything in Turbo mode, only climbing is measured. It's pretty darn efficient, I dare to say it's the most efficient system on the market today(here the motor and system Q factor is really high).
600Wh(20,7km, 1605m Altitude gain)
View attachment 151973
800Wh(26,1km, 2132m Altitude gain)
View attachment 151959
3. Then here is are few test charts from lab tests. In these charts it's possible to see motor behavior(power/torque relationship)at different cadences, and due to losses we can see smaller torque figures than the actual motor specs.
The numbers are measured at the crank(so all the internal transmission losses and electrical losses including) torque/cadence curve(left power, right torque, bottom cadence)for the:
Sram Eagle(Brose)
View attachment 151965
Bosch CX gen4
View attachment 151966
Giant Syncdrive Pro 2
View attachment 151967
4. Then here is a power/cadence curve comparison of several motors.
View attachment 151970
Now I am derailing the thread
That is a heck of a Strava ride...over 100km and 2000m climbing...and still with 24% battery remaining...wow! While I am no stranger to hard pedaling, and ride analog road/gravel all the time, when I get on an e-MTB I have a tendency to want to run it in Turbo/Boost most of the time, which may have given me a distorted idea of what is possible in terms of range. Having that extra power on tap, which can make flat terrain and even gentle uphills feel like downhills, is very addictive.
Those screenshots from the Bosch Gen5 test are cool, both for a real world test, and also because they seem very consistent with each other in terms of the watt/hour to elevation gain relationship. That gives me more confidence in the reproducibility and applicability of the results.
Regarding those torque and power curve charts, that is interesting info. The ebike-mtb.com motor tests were the first ones I had seen which demonstrated how varied the optimal cadence ranges and output behaviors were for the existing motors, even ones with similar specs, but I don't think they had any charts that showed the full spectrum. On those charts that you posted, which show power vs. torque at different cadences, it is interesting to see the relationship between the 2 and how it varies both for a given motor, and between motors. I don't have the technical knowledge that you do, so can you offer any additional insight into the significance of those power and torque curves? For example, is there any magic to riding at the cadence where the power and torque curves cross? Does that point (or some other point) represent the optimal point of efficiency of battery input to motor output?
I see the Sram/Brose has an admirable amount of low cadence torque, which can be nice when caught in a tricky climbing situation. I recently had the opportunity to put in several rides on the Bosch SX motor (low torque but potentially high power), which I'd thought I would get along with well given that I tend to be a fairly high cadence rider on analog bikes. It was fine, and had good power control/modulation, but then I did a ride right after it on a ep801 bike and the greater torque was instantly noticeable, even at moderate cadences, which for a turbo mode junky like me was very noticeable.
The issue I have with those charts are they don't tell you what the motors are set to in terms of assistance character or how much pedal force is being applied. I'm assuming that most e-bike motor apps have an adjustable max torque as well as the assist character. It makes a big difference in how fast a motor ramps up the torque vs input rpm and measured pedal force. You can set the Shimano to 85nm max, but if the assist character is turned down then the pedal force and rpm required to get there would be pretty high. This would also change the measured efficiency by a lot. So if you wanted a certain motor to rate higher on efficiency, you'd turn the assistance character way down. Now on a 10% grade it's wh usage looks really good. Then on the power test you turn it to the highest assist character and the torque vs cadence looks great at all rpms. On both tests the max torque is 85nm, but the assist is way different.
TheKaiser
Active member
That is an excellent point! I would have assumed there was some attempt to standardize the settings, but I can't read the source articles, so I have no way of confirming it, and am not even sure how possible it is to do so given each of these motors uses its own proprietary software. Feel free to share any info you have on that topic.
Same issue with dirt bike dyno "tests" nowadays. There's phone apps that you can change the timing and fueling at different rpms and engine loads (throttle position on most dirt bikes) so you don't know if they had it turned to max advance or retarded it to minimum advance.
On these e-bikes motors the output is variable and adjustable for max torque, cadence, incline, and rider input power. The Shimano ETube App is the one I have experience with and the UI looks like this:
The ghosted lines are where you can choose the assist character, all the way right is turned way down and to the left is most aggressive. So you can have the max torque turned up, but the assist character turned down. This would give you hella battery life, but it'd feel like no assistance until you really started pushing the rpm and human power input or got on a really steep grade. At the most aggressive character setting, you could turn the max torque down and still feel like your getting a lot of assistance at lower rpm and rider input. I ride in trail mode 95% of the time and I played with turning the assist character up two lines, max torque was already at 85nm from the factory. With just this small adjustment the assist was too much and I had to turn it back to level 6 which is stock.
On the power cadence curve graph that was posted the Shimano power level falls off a cliff well before the other motors. This tells me either they had the assist cutoff set too low, or it hit the max speed before the other motors. This would point to software or gearing discrepancies with the test since the max speed should be the same on all the motors tested. I can tell you from experience it's got plenty of assistance above 110rpm, I've hit 125rpm before playing with it and there was no sudden drop in power. To really test these motors they need to get all the apps and test them in "factory" mode, which might differ from bike manufacturer to manufacturer, and then do a test with max torque and max assist character, and then another with max torque and minimum assist character.
- Thread starter
- #54
Hey @timo2824 good points there, want provide some more info where I got the charts and all..
The charts I provided are mostly from 2 sources:
1. ---------》
www.bike-magazin.de
youtube.com
-----------《
2. ------------》
youtube.com
-------------《
From my knowledge both sources do their testing in the same lab(PT labs) in Germany. As far as I remember from their videos and articles and a magazine I also bought, they do the testing with all the settings on the motors set to their max output overall(assistance, torque and launch or whatever others call it). While Velomotion gives a lot of information how the testing is done bike magazin does not as it's a paid magazine. The charts I provided were a lucky onetime access to their test via some link don't remember how I got it.
If we look at Velomotion's info, they do two different tests one with 250W of input power and one with 100W and as previously said on max assist settings.
Your points that motor have different assistance characteristics are on point, that's why they say they do the testing on max settings, just to most possibly even out the ods between the motors.
Of course a perfect match can never be done, but it also shouldn't be as then then we miss the point of testing different machines and their settings out of the factory. Now, If we can see differences at max load, that's a pretty good indicator for even greater difference at a lower load.
@TheKaiser Yes your assumptions are correct, the motor will be most efficient at the cadence where the two lines(power/torque) meet.
The Bosch Gen 5 tests are from this yt channel...the guy has more tests in videos and all nicely explained and documented:
About battery consumption on an ebike.
There is one thing that eats the battery pretty quick. I see a lot of riders doing technical uphill over obstacles etc..., the way an electrical motor works is that it's efficiency peaks at around 85% of its maximum rpm.
At low rpm the torque and current(consumption) is really high, but as you can see it drops with increased rpm and so at a high enough rpm you have the highest efficiency(low current draw and low torque).
What I am trying to say, every time you stop, slow down and accelerate it eats up the battery much more, and so spirited technical uphill riding is a biiig battery eater. A real life example, starting from a stop in high assistance mode and a gear to high(hard on the pedals) is going to eat a loot of battery.
You achieve best efficiency with good shifting and a good consistent motor rpm/pedaling cadence(the motor rpm adapts to your cadence).
Happy New Year 
The charts I provided are mostly from 2 sources:
1. ---------》
E-MTB
EMTB testet neue E-Mountainbikes und Komponenten. Daraus entstehen unabhängige Tipps für Ihren E-Bike-Kauf. Plus: E-Bike Touren mit dem Mountainbike.
www.bike-magazin.de
BIKE Magazin
Der offizielle Youtube-Kanal von BIKE – Europas größtem Mountainbike Magazin. Wir versorgen Dich mit Videos und Clips rund ums Thema Mountainbike. Zu folgenden Themen: - BIKE-Tests: Mountainbikes, MTB-Komponenten und Zubehör - Fahrtechnik: How-to-Videos zu Wheelie, Bunnyhop & Co. -...
youtube.com
2. ------------》
Velomotion Magazin
Velomotion.de ist dein digitales Magazin für alle Themen rund um Fahrrad- und E-Bike. Du findest hier die neuesten Infos zu Bikes und Zubehör: vom Kinderrad bis zum Pedelec, vom Downhill-Boliden bis zur Triathlon-Maschine. Velomotion testet Fahrräder und E-Bikes ebenso wie technisches Zubehör...
youtube.com
From my knowledge both sources do their testing in the same lab(PT labs) in Germany. As far as I remember from their videos and articles and a magazine I also bought, they do the testing with all the settings on the motors set to their max output overall(assistance, torque and launch or whatever others call it). While Velomotion gives a lot of information how the testing is done bike magazin does not as it's a paid magazine. The charts I provided were a lucky onetime access to their test via some link don't remember how I got it.
If we look at Velomotion's info, they do two different tests one with 250W of input power and one with 100W and as previously said on max assist settings.
Your points that motor have different assistance characteristics are on point, that's why they say they do the testing on max settings, just to most possibly even out the ods between the motors.
Of course a perfect match can never be done, but it also shouldn't be as then then we miss the point of testing different machines and their settings out of the factory. Now, If we can see differences at max load, that's a pretty good indicator for even greater difference at a lower load.
@TheKaiser Yes your assumptions are correct, the motor will be most efficient at the cadence where the two lines(power/torque) meet.
The Bosch Gen 5 tests are from this yt channel...the guy has more tests in videos and all nicely explained and documented:
About battery consumption on an ebike.
There is one thing that eats the battery pretty quick. I see a lot of riders doing technical uphill over obstacles etc..., the way an electrical motor works is that it's efficiency peaks at around 85% of its maximum rpm.
At low rpm the torque and current(consumption) is really high, but as you can see it drops with increased rpm and so at a high enough rpm you have the highest efficiency(low current draw and low torque).
What I am trying to say, every time you stop, slow down and accelerate it eats up the battery much more, and so spirited technical uphill riding is a biiig battery eater. A real life example, starting from a stop in high assistance mode and a gear to high(hard on the pedals) is going to eat a loot of battery.
You achieve best efficiency with good shifting and a good consistent motor rpm/pedaling cadence(the motor rpm adapts to your cadence).
Happy New Year
Last edited:
The graph with the power to cadence curve really sticks out that something isn't right with their testing. The Shimano has assistance well above the 110rpm point that it drops off a cliff. It's still interesting info, I'd really like to be there during a test to see how they're doing it. Would be really cool to be a part of it.Hey @timo2824 good points there, want provide some more info where I got the charts and all..
The charts I provided are from mostly from 2 sources:
1. ---------》
E-MTB
EMTB testet neue E-Mountainbikes und Komponenten. Daraus entstehen unabhängige Tipps für Ihren E-Bike-Kauf. Plus: E-Bike Touren mit dem Mountainbike.
-----------《BIKE Magazin
Der offizielle Youtube-Kanal von BIKE – Europas größtem Mountainbike Magazin. Wir versorgen Dich mit Videos und Clips rund ums Thema Mountainbike. Zu folgenden Themen: - BIKE-Tests: Mountainbikes, MTB-Komponenten und Zubehör - Fahrtechnik: How-to-Videos zu Wheelie, Bunnyhop & Co. -...
2. ------------》
-------------《Velomotion Magazin
Velomotion.de ist dein digitales Magazin für alle Themen rund um Fahrrad- und E-Bike. Du findest hier die neuesten Infos zu Bikes und Zubehör: vom Kinderrad bis zum Pedelec, vom Downhill-Boliden bis zur Triathlon-Maschine. Velomotion testet Fahrräder und E-Bikes ebenso wie technisches Zubehör...
From my knowledge both sources do their testing in the same lab(PT labs) in Germany. As far as I remember from their videos and articles and a magazine I also bought, they do the testing with all the settings on the motors set to their max output overall(assistance, torque and launch or whatever others call it). While Velomotion gives a lot of information how the testing is done bike magazin does not as it's a paid magazine. The charts I provided were a lucky onetime access to their test via some link don't remember how I got it.
If we look at Velomotion's info, they do two different tests one with 250W of input power and one with 100W and as previously said on max assist settings.
View attachment 152616
Your points that motor have different assistance characteristics are on point, that's why they say they do the testing on max settings, just to most possibly even out the ods between the motors.
Of course a perfect match can never be done, but it also shouldn't be as then then we miss the point of testing different machines and their settings out of the factory. Now, If we can see differences at max load, that's a pretty good indicator for even greater difference at a lower load.
@TheKaiser Yes your assumptions are correct, the motor will be most efficient at the cadence where the two lines(power/torque) meet.
The Bosch Gen 5 tests are from this yt channel...the guy has more tests in videos and all nicely explained and documented:
About battery consumption on an ebike.
There is one thing that eats the battery pretty quick. I see a lot of riders doing technical uphill over obstacles etc..., the way an electrical motor works is that it's efficiency peaks at around 85% of its maximum rpm.
View attachment 152615
At low rpm the torque and current(consumption) is really high, but as you can see it drops with increased rpm and so at a high enough rpm you have the highest efficiency(low current draw and low torque).
What I am trying to say, every time you stop, slow down and accelerate it eats up the battery much more, and so spirited technical uphill riding is a biiig battery eater. A real life example, starting from a stop in high assistance mode and a gear to high(hard on the pedals) is going to eat a loot of battery.
You achieve best efficiency with good shifting and a good consistent motor rpm/pedaling cadence(the motor rpm adapts to your cadence).
Just for reference I work with industrial motors, mostly AC, but I find all this very interesting. Just for fun I changed my assistance character in trail mode up 5 points and rode a 7 mile trail that I ride several times a month. My normal time on this trail is 40 minutes, I finished it in 34 minutes with the more aggressive motor setting. I just use the 10 bar battery level indicator on the display and total battery usage was the same 2 bars I normally use, but with less time. The total work done was the same, but over a shorter duration. I tried to put the same effort in as normal, but the assistance was so much that it was much easier, especially on the climbs. It's really cool how much control and tuning we have with the apps.
- Thread starter
- #56
Yes to see the testing would be really interesting, agree, in some of the Velomotion's videos, the test bench can be seen at work.
Do you ride the EP8 or EP801, they surely have a different firmware, and at the end it's also a question on what firmware was on the motor at the time of testing, maybe it was one with a glitch, we don't know unfortunately, but it's a concern to think about.
This is a chart from Velomotion's testing:
They compare cadence/power in the first photo, don't know why, the normal thing would be cadence/torque, but OK.
This just the power curve at max:
This is thermal stability of the motor:
I ride an ep801 with the newest firmware. There's definitely something going on with the ep801 in this test because it should be able to get to 600 watts, but looks like it's limiting for some reason. So probably not setup to max correctly or there was an issue with it. I know it seems to really like higher cadence, like 100+rpm in real world riding, but according to this test it's falling fast after 85rpm. It might be because the torque sensor, during testing does it say what they were using as a dyno? If it wasn't enough resistance to make whatever was spoiling the crank read adequate input torque, then the assistance would be cut back. That's what I like about it is a very natural feeling assistance, the harder you pump the more it outputs. I know some like other motors because they feel like a lot of assistance, so they might put more bias into the cadence and less into the input torque on those motors.Yes to see the testing would be really interesting, agree, in some of the Velomotion's videos, the test bench can be seen at work.
Do you ride the EP8 or EP801, they surely have a different firmware, and at the end it's also a question on what firmware was on the motor at the time of testing, maybe it was one with a glitch, we don't know unfortunately, but it's a concern to think about.
This is a chart from Velomotion's testing:
They compare cadence/power in the first photo, don't know why, the normal thing would be cadence/torque, but OK.
View attachment 152621
This just the power curve at max:
View attachment 152622
This is thermal stability of the motor:
View attachment 152623
TheKaiser
Active member
I'm a bit confused by that last graph, with the power, current, torque, and efficiency curves, as I'm having trouble reconciling it with some of the earlier charts you posted. In the earlier charts you posted midway up the page, for the Giant, SRAM, and Bosch motors, we saw how with increasing cadence, torque trended down, and power trended up. At some point the two curves crossed each other, (at around an 80-90rpm cadence), which we'd discussed being the point of maximum efficiency, even though motor power increased a bit more at cadences beyond that point. On the other hand, in the above graph, the dotted green efficiency line seems to go up and up, well past the point of peak power generation, suggesting that maxing out the RPM is most efficient, even if it's well beyond the point of peak power. I know that the earlier charts were for pedal cadence, and this most recent chart is for motor RPM, but it still seems like the same principle should apply in terms of where a motor is most efficient, relative to it's power output. I would have thought that the optimal efficiency would be at around 1000rpm, when the power and torque curves crossed, not at 4000rpm when power and torque are both falling off the chart.Hey @timo2824 good points there, want provide some more info where I got the charts and all..
The charts I provided are mostly from 2 sources:
1. ---------》
E-MTB
EMTB testet neue E-Mountainbikes und Komponenten. Daraus entstehen unabhängige Tipps für Ihren E-Bike-Kauf. Plus: E-Bike Touren mit dem Mountainbike.
-----------《BIKE Magazin
Der offizielle Youtube-Kanal von BIKE – Europas größtem Mountainbike Magazin. Wir versorgen Dich mit Videos und Clips rund ums Thema Mountainbike. Zu folgenden Themen: - BIKE-Tests: Mountainbikes, MTB-Komponenten und Zubehör - Fahrtechnik: How-to-Videos zu Wheelie, Bunnyhop & Co. -...
2. ------------》
-------------《Velomotion Magazin
Velomotion.de ist dein digitales Magazin für alle Themen rund um Fahrrad- und E-Bike. Du findest hier die neuesten Infos zu Bikes und Zubehör: vom Kinderrad bis zum Pedelec, vom Downhill-Boliden bis zur Triathlon-Maschine. Velomotion testet Fahrräder und E-Bikes ebenso wie technisches Zubehör...
From my knowledge both sources do their testing in the same lab(PT labs) in Germany. As far as I remember from their videos and articles and a magazine I also bought, they do the testing with all the settings on the motors set to their max output overall(assistance, torque and launch or whatever others call it). While Velomotion gives a lot of information how the testing is done bike magazin does not as it's a paid magazine. The charts I provided were a lucky onetime access to their test via some link don't remember how I got it.
If we look at Velomotion's info, they do two different tests one with 250W of input power and one with 100W and as previously said on max assist settings.
View attachment 152616
Your points that motor have different assistance characteristics are on point, that's why they say they do the testing on max settings, just to most possibly even out the ods between the motors.
Of course a perfect match can never be done, but it also shouldn't be as then then we miss the point of testing different machines and their settings out of the factory. Now, If we can see differences at max load, that's a pretty good indicator for even greater difference at a lower load.
@TheKaiser Yes your assumptions are correct, the motor will be most efficient at the cadence where the two lines(power/torque) meet.
The Bosch Gen 5 tests are from this yt channel...the guy has more tests in videos and all nicely explained and documented:
About battery consumption on an ebike.
There is one thing that eats the battery pretty quick. I see a lot of riders doing technical uphill over obstacles etc..., the way an electrical motor works is that it's efficiency peaks at around 85% of its maximum rpm.
View attachment 152615
At low rpm the torque and current(consumption) is really high, but as you can see it drops with increased rpm and so at a high enough rpm you have the highest efficiency(low current draw and low torque).
What I am trying to say, every time you stop, slow down and accelerate it eats up the battery much more, and so spirited technical uphill riding is a biiig battery eater. A real life example, starting from a stop in high assistance mode and a gear to high(hard on the pedals) is going to eat a loot of battery.
You achieve best efficiency with good shifting and a good consistent motor rpm/pedaling cadence(the motor rpm adapts to your cadence).
- Thread starter
- #59
Hey I understand your dilemma. The last graph u are referring is a plain characteristic chart for a DC(direct current) electrical motor, don't take it for granted, the power curve is very simetric here, it can be quite different for a different voltage(36V) motor.
All ebike motors are DC motors, the graph shows the basic principles of how a DC electric motor works. The values presented here are the ones an electric motor has as it is. I just wanted to point out that a lot of stop and go creates a lot of current peaks that put the battery under a lot of stress and so drain it faster.
Please don't take these values as comparison for the previous graphs for different motors as those measurments are taken/measured on the crankset(internal transmission comes into place).
Yes I understand why you would think that the 1000rpm would be the most efficient point, but look at the current draw at that point it is really high, and you have some power, this means the motor at that point is extremely inefficient, but on the other side, the motor spins at a comfortable 80% and so has low torque, little current draw, low consumption at about the same power.
It all depends what you need/want from a system, at low rpm you have huge torque, low power and a lot of consumption(probably good when going up steep terrain-that's where u need torque), at high rpm you have low torque, low consumption(good for touring and flater terrain), the power figure is always a derivative of the other two.
Power(W) = Force(Nm) x Rpm
An ebike motor is a complex system to explain on a forum, honestly trying to understand it is mostly just a hobby for me, and even I don't understand everything into detail, but learn step by step.
Hey, I will try to answer to your 5 paragraphs in 5 points. I will try to use some online simplified explanations.
1. 250W Continuous
Well yes, mostly you understood correctly, the manufacturers must not label the motor more than 250W and the motor during testing must not overheat while continuously working at 250W.
The regulations on which the testing is done say how long the period of testing is, did not go into detail, but from memory I recall I read that the testing is done in 30min period.
Terminology 》》》this is a tricky one, not only here in ebike talk, but all over the modern tech world.
So yes I would avoid the term "average power", as it's completely wrong here. The term "nominal power" comes closer but imo also still not quite correct and my preferred term is "continuous power"
Nominal versus peak power. A motor’s nominal wattage is the maximum amount of power it can sustain for long periods of time, while the peak output is the wattage a motor is capable of for a very short burst. Peak output wattage is still a useful rating that gives you an idea of how hard a bike might accelerate or how it might feel on short bursts on a very steep hill.
Motors have thermal limits before they fail. Continuous and peak power are based upon those thermal limits. Continuous power can be used non stop. The cooling for the motor, whether it is a fan, natural convection, water cooling, is sufficient to keep the motor at a happy temperature. Peak power is driven more by the thermal mass and magnetic limitations of the motor and because the temperature here is rising faster than we can cool the motor, this can only last a limited amount of time.
2. Motor Q factor
This one is a bit technical...I tried to find a simplified explanation, hope this is clear enough. Q factor(also known as Quality Factor) is defined as a dimensionless parameter and it measures the performance of a coil(for our case this is it as the stator of the motor is a coil), a capacitor, or an inductor in terms of its losses and resonator bandwidth. The Q factor indicates energy loss in a design. Also, the Quality Factor (Q) may be defined as the ratio between stored energy and energy dissipated per cycle in a circuit.
While all electric motors probably have a Q factor label and it's efficiency is driven by it, this is a label only the engineers in the production use and for the end user it's a non existent spec. The end result of a system with a good q factor motor is seen in good efficiency at high loads.
3./4. System Losses
Yes you are correct with the 500Wh battery, 250W continuous power draw for 2h, If that is measured between the battery and the motor(the electrical power). Even here we have thermal losses(battery cells, wires, motor electronics,...) The BMS is the one the tries to control the energy supply and the motor controller distributes it accordingly to demand. How and where the numbers the labs get from testing is more a question for them, the important thing is when comparing numbers it should be from the same lab and same environment(several factors).
Yes the internals of the motor definitely slightly affect the end numbers, although most torqe difference probably comes from the max load at wich the testing is done, as at high power and rpm, the torque is dropping lower(P = F x Rpm).
The motors are tuned to a specific rpm for best efficiency, I they did the testing slightly out(over) of this window, the outcome can so be lower torque figures. This is why I always said, the manufacturers should focus on a wider torque band, especially for emtb motors. Unfortunately no, it's not possible from specs to read which motor will be more efficient, but there are tests, and from my experience, they tell a pretty accurate real life situation:
View attachment 151888
Screenshot - YT Velomotion Magazin
5. System Overall
Exactly, the system on it's own is just a part of the story when talking about efficiency of an ebike, two biggest enemies are rider/system weight, inclines, followed by tyre choice and pressure, terrain and environment temperatures.
The issue I'm seeing is that each of these Wh/ km charts show useage/ efficiency at max power output. Anytime you go faster, you burn more energy per km.
So, unless I am misunderstanding, for these range tests to be valid the testers would need to normalize for input pedaling assistance AND output power, to properly access efficiency.
Similar Threads
-
-
-
Specialized M2 M3 battery for Gen 2/3 Levo, improved capacity 860Wh, almost same weight :-)- Started by cream
- Replies: 13
-
Specialized M1 battery for Levo Gen1, improved capacity 576 Wh, same weight :-)
- Started by cream
- Replies: 15
-