Running a 96v motor (HPM3000B) at 68v

any downsides?

i need a low-kv motor so i can get away w/ using a 1-stage reduction w/o the sprockets getting ridiculously big. the latter detail being important when riding in mud. large sprockets tend to get dipped in mud and cause problems

Overclocker said:

any downsides?

Click to expand...

If it's a 3kw motor at 96V, then the same motor is a 1.5kw motor at 48V, and for thick mud is simply not going to be enough motor except maybe for a kids' bike.

John in CR said:

Overclocker said:

any downsides?

Click to expand...

If it's a 3kw motor at 96V, then the same motor is a 1.5kw motor at 48V, and for thick mud is simply not going to be enough motor except maybe for a kids' bike.

Click to expand...

Actually It's a square relationship (e.g. Benjamin Vedder explained it in this post), so a 96V motor run at 48V, that's geared the same top speed, only is a 750W motor.


Or in your case @Overclocker, the HPM3000B will only be half as powerful @68V as when run @96V. If that's enough for you or not is something you must decide.
If I was you I'd probably look for a highher pole count/e.g. lower k/v motor...

twikis said:

John in CR said:

Overclocker said:

any downsides?

Click to expand...

If it's a 3kw motor at 96V, then the same motor is a 1.5kw motor at 48V, and for thick mud is simply not going to be enough motor except maybe for a kids' bike.

Click to expand...

Actually It's a square relationship (e.g. Benjamin Vedder explained it in this post), so a 96V motor run at 48V, that's geared the same top speed, only is a 750W motor.


Or in your case @Overclocker, the HPM3000B will only be half as powerful @68V as when run @96V. If that's enough for you or not is something you must decide.
If I was you I'd probably look for a highher pole count/e.g. lower k/v motor...

Click to expand...

While I didn't read the full detail of Vedder's article, nothing blatantly wrong stood out, so you're pulling his info out of context. For a given motor, when you run it at lower than intended voltage, you get proportionately less power from it. Power = Torque X rpm . Rpm is directly proportional to voltage, since Kv (rpm/volt) is a constant for a given motor. It also has a torque constant (torque/amp), and current handling remains the same, so torque is the same. The voltage constant (Kv) and torque constant (Kt) for our BLDC motors are inversely related such that Kv in rpm/volt = 9.549/Kt in Nm of torque per amp of motor current. Pole count really has no bearing on either.

Since he wants to keep the chain out of the mud with a smaller rear sprocket, he needs to go with a significantly larger motor to run at low rpm in order to approach anything resembling the performance of the Perm.

Personally I'd go with a hubmotor run as a mid-drive, but that's because I already have quite a few smaller diameter longer stator high efficiency hubbies with a Kv of 18-19rpm/volt that handle plenty of power that would perfect for the implementation. In fact I just started a build using 21" moto wheels that I reasonably expect to be more silent than typical belt drives despite using 08B chain, because I'm using a 28t front and 57t rear sprocket. I've done similar before with bicycle chain, and with good alignment and a chain tensioner and it was essentially as quiet as a DD hubmotor.

John in CR said:

While I didn't read the full detail of Vedder's article, nothing blatantly wrong stood out, so you're pulling his info out of context. For a given motor, when you run it at lower than intended voltage, you get proportionately less power from it. Power = Torque X rpm . Rpm is directly proportional to voltage, since Kv (rpm/volt) is a constant for a given motor. It also has a torque constant (torque/amp), and current handling remains the same, so torque is the same. The voltage constant (Kv) and torque constant (Kt) for our BLDC motors are inversely related such that Kv in rpm/volt = 9.549/Kt in Nm of torque per amp of motor current. Pole count really has no bearing on either.

Click to expand...

Whoops. Sorry and yes, you're right. No idea what got into my head there regarding the square relation lol

Still, pole count is one of the most important factors, rpm wise. Any given motor with double the pole count, all other factors being the same, will have about twice the torque and half the rpm at similar losses. Unfortunately, high pole count motors are hard to find, and hub motors are usually big and heavy, compared to their copper and iron mass...

In my experience (06B-2 chain in the Cityel), industrial chains simply don't last any reasonable amount of time in motorized vehicles, and are usually as cheap and low quality as can be. A 28t front sprocket should be very easy on the chain though, and you might be happy with it

of course i'm going to overclock the 96v motor by feeding in more amps

hpm3000b is 3kw, 6kw max

but the higher resistance of the 96v winding will produce more heat, right?

since that motor isn't in the grin tech simulator, i used the mxus 4503 hubmotor. set up with essentially unlimited battery and controller capability (999999a, 0.0000001ohms resistance), so the motor and the voltage applied to it should be the limiting factors. system a is 96v, and system b is 48v, just so you can see the comparison of half the voltage, to see how all the other things relate.

don't know if it's an artifact of some other setting, or a real phenomenon, but the motor watts at 96v are 4653 vs 1001 at 48v; close to 5:1 with no slope but full throttle. the relationship is closer to 2:1 if the slope is 90% instead of zero.

https://www.ebikes.ca/tools/simulator.html?motor=MX4504&batt=cust_96_0.0001_9999&mass=170&hp=0&cont=cust_999999_99999_0.0000001_V&cont_b=cust_999999_99999_0.0000001_V&motor_b=MX4504&batt_b=cust_48_0.0001_9999&mass_b=170&hp_b=0&bopen=true

Graph Syst A Syst B
Wheel Torq 67.5Nm 25.2Nm
Mtr Power 4653W 1001W
Load 4715W 1015W
Efficiency 85.7% 88.2%
RPM 657.9 rpm 378.6 rpm
Electrical Syst A Syst B
Mtr Amps 66.2A 25.2A
Batt Power 5429W 1134W
Batt Amps 56.6A 23.6A
Batt Volts 96.0V 48.0V
Performance Syst A Syst B
Acceleration -0.06 kph/s -0.02 kph/s
Consumption 66.3 Wh/km 24.1 Wh/km
Range 14479 km 19936 km
Overheat In 9.3 minutes never
Final Temp >250 °C 68 °C



then the same comparison but with 68v for how it would be for your battery voltage. the ratio is closer to 2:1 for the same settings that get almost 5:1 at 48v. again, dunno if it's an artifact or not.

https://www.ebikes.ca/tools/simulator.html?motor=MX4504&batt=cust_96_0.0001_9999&mass=170&hp=0&cont=cust_999999_99999_0.0000001_V&cont_b=cust_999999_99999_0.0000001_V&motor_b=MX4504&batt_b=cust_68_0.0001_9999&mass_b=170&hp_b=0&bopen=true

Graph Syst A Syst B
Wheel Torq 67.5Nm 41.1Nm
Mtr Power 4653W 2179W
Load 4715W 2246W
Efficiency 85.7% 87.4%
RPM 657.9 rpm 506.5 rpm
Electrical Syst A Syst B
Mtr Amps 66.2A 40.6A
Batt Power 5429W 2493W
Batt Amps 56.6A 36.7A
Batt Volts 96.0V 68.0V
Performance Syst A Syst B
Acceleration -0.06 kph/s -0.08 kph/s
Consumption 66.3 Wh/km 39.5 Wh/km
Range 14479 km 17195 km
Overheat In 9.3 minutes never
Final Temp >250 °C 126 °C

Awesome idea @amberwolf!

When you double the voltage and keep the same gear ratio, you end up with about twice the speed. Since speed and air resistance are in a square relationship, you need roughly 4 times the power to overcome air resistance for twice the speed. Incl. higher losses in the motor those numbers look good :thumb: (it is true though that this simulator has some unnecessarily high rounding errors in its algorithm, while all else about it is perfect...)



Buuuut: I just observed something interesting, which actually (/unfortunately) proves the theory I've had in the back of my mind, that voltage and output efficiency (and therefore load capability) are connected in a square relationship, to be right.

Have a look at the green curves (which represent efficiency) in this simulation set (system A and system B are identical, except that system B is running at half the voltage and half the current, and has a 2:1 reduction instead of 4:1):
https://www.ebikes.ca/tools/simulator.html?motor=MX4504&batt=cust_96_0.0001_10&mass=170&hp=0&cont=cust_999999_40_0.0000001_V&cont_b=cust_999999_20_0.0000001_V&motor_b=MX4504&batt_b=cust_48_0.0001_10&mass_b=170&hp_b=0&bopen=true&mid=true&mid_b=true&gear=4&gear_b=2&throt=100&throt_b=100

They're very close to each other, so the motors have got about the same efficiency at any given speed during acceleration (actually system B's efficiency is a bit lower). The issue is though that compared to system A, volts and amps are halfed, meaning it can only run at 1/4 the power (in this case it's 1/4 the wheel torque) for the same efficiency as it would at 96V

When you turn the amps for system B up to 40A, like for system A, you get this graph:
https://www.ebikes.ca/tools/simulator.html?motor=MX4504&batt=cust_96_0.0001_10&mass=170&hp=0&cont=cust_999999_40_0.0000001_V&cont_b=cust_999999_40_0.0000001_V&motor_b=MX4504&batt_b=cust_48_0.0001_10&mass_b=170&hp_b=0&bopen=true&mid=true&mid_b=true&gear=4&gear_b=2&throt=100&throt_b=100

As you can see, the efficiency curve of system B is much lower, and therefore the losses at any given speed are much higher at half the power of the 96V system.

Actually I remembered the reason why that is now, but my head is already bursting hahah. If nobody else explains it soon, I'll make another post trying to explain it. :thumb: