it's you who misread the situation.

The Bafang G06 motor is robust enough, it's not a big problem to put 1kW though it for 10 minutes to climb a hill. I have said so many times. It's not kind to the motor but it'll live. It's the rest of the electronics.

You need a proper 48V controller and LCD. You can't use the existing Lishui HL controller and the C200 LCD.

If you use KT kit, you need a new wiring loom.

I have explained all this yesterday. To convert the Rio FB to 48V, you need a budget of around £400.

The real issue is those who want to use it as a motorbike, running it at maximum speed to save time. Overvolting gives them a normal looking bike with a brand and an EN15194 sticker while breaking law.

And add to that, my friend helps him to do that to me.

 

 

I get your concerns about your brand etc, which is fine for selling your products. But just because someone over volts their bike does not mean they want to break the law. That only happens if the user deliberately misrepresents the wheel diameter in the LCD set up.

 

As VFR said, it's much better to get that 1kW from high volts/low amps than the other way around. Heat is your enemy, which is proportional to the current squared.

it's you who misread the situation.

The Bafang G06 motor is robust enough, it's not a big problem to put 1kW though it for 10 minutes to climb a hill. I have said so many times. It's not kind to the motor but it'll live. It's the rest of the electronics.

You need a proper 48V controller and LCD. You can't use the existing Lishui HL controller and the C200 LCD.

If you use KT kit, you need a new wiring loom.

I have explained all this yesterday. To convert the Rio FB to 48V, you need a budget of around £400.

The real issue is those who want to use it as a motorbike, running it at maximum speed to save time. Overvolting gives them a normal looking bike with a brand and an EN15194 sticker while breaking law.

And add to that, my friend helps him to do that to me.

That's what I said in post #3!

wheezy, if a customer calls us about a bike, we always, always ask for his/her weight and height and where he/she lives. We make bikes with 40NM motors (85SX on the Faro) to Bafang BPM 65NM motor on the Big Bears. Horses for courses, the bike must be able to do the job.

There is no need for more than 36V 20A controller to make a pedal assist bike. Only those who want to go beyond 25mph need more than that.

On exceptional circumstances, too tall or 30st, I suggest our 48V kits.

It's not I don't understand people have different neeeds, I cannot sell a bike or a kit to someone who tells me upfront he is going to use the bike 'off road'. My reply to them has always been that we don't make off road bikes.

Edited July 5, 20205 yr by Woosh

wheezy, if a customer calls us about a bike, we always, always ask for his/her weight and height and where he/she lives. We make bikes with 40NM motors (85SX on the Faro) to Bafang BPM 65NM motor on the Big Bears. Horses for courses, the bike must be able to do the job.

There is no need for more than 36V 20A controller to make a pedal assist bike. Only those who want to go beyond 25mph needs more than that.

 

 

The thing is, I would not want to put 20A into a "250W" motor. It's a bad idea. Much better to use 15A and a higher voltage to get the peak power needed for hill climbing etc. With voltage adjusted accordingly, that extra 5A means almost double the heating of the motor for the same power output.

wheezy, that's not correct. The phase of the current and voltage in/on the coils are at 90 degrees angle current versus voltage. The coils are not resistors unless when it stalls, then the controlller will stop sending power to it.

The amount of heat wasted through the motor is shown as = 1- yield. So, low yield = more wasted heat. The yield changes with the RPM and Kv, a motor constant. The voltage has no real effect on the motor itself, as I said, the yield changes with the RPM. You want it optimal at 200 RPM for full size bikes. The motor is wound so that you get optimal yield for the voltage you are going to use.

You are going to shed more heat proportionally unless you ride always at the optimal RPM.

When you overvolt one of my motors, you move the optimal RPM to 260. So instead of enjoying best speed at 15mph, it's going to be 20mph.

I know it's fine for some, but not what I am supposed to do.

I think there is some things people are forgetting.

Firstly, to get more torque, (other than changing gearing) you always have to put more current through the motor, as torque is proportional to current.

 

Battery current is not the same as phase (motor) current.

 

The extra torque that comes from overvolting is caused by extra motor current over and above the battery current, as motor system steps down voltage and steps up current. (acts like buck converter)

This can be demonstrated below with motor simulator, as motor current of 20.5 amp at 36v X 18.2amp controller is same as 48v X 14 amp, both representing a 30% increase in normal 36v X 14amp controller.

There is no magic way (in this case) of getting more torque without more heat in motor, without changing gearing. e.g. Volt up,Gear down.

 

https://www.ebikes.ca/tools/simulator.html?motor=MXUS_XF07&cont=cust_18.2_70_0.03_A&wheel=700c&mass=100&hp=50&bopen=true&cont_b=cust_14_70_0.03_A&motor_b=MXUS_XF07&wheel_b=700c&mass_b=100&hp_b=50&batt_b=B4816_GA&grade=8&grade_b=8

Edited July 5, 20205 yr by Sturmey

you need to think of electricity like waves, the voltage at the storage capacitor is not same as battery's voltage. It is wavy, although connected to the battery with two short fat wires, black and red.

the increase in motor current that is needed for more torque is partially supplied by the 2.2mF storage capacitor.

Higher voltage increases the energy stored, thus gives higher discharge current through the coils.

Oscillations caused by the back EMF in an overvolted motor create also 30%-50% higher voltages when riding at high speed because the magnets in an overvolted motor are working at over saturated state. When the increase current cannot create any more magnetic flux, its wave is reflected back like a sea wave hitting a rock. Same thing as regen, spontaneous turbulent voltage is much higher than your battery's voltage.

The thing you should worry about is the width of the tracks on your controller's PCB. Wave travels only on the surface of the conductor. Higher wave creates more heat on the tracks.

If your storage capacitor is not good enough, the back EMF may kill this capacitor at set fire to the controller.

Rule of thumb: lower voltage = better reliability.

Edited July 6, 20205 yr by Woosh

wheezy, that's not correct. The phase of the current and voltage in/on the coils are at 90 degrees angle current versus voltage. The coils are not resistors unless when it stalls, then the controlller will stop sending power to it.

The amount of heat wasted through the motor is shown as = 1- yield. So, low yield = more wasted heat. The yield changes with the RPM and Kv, a motor constant. The voltage has no real effect on the motor itself, as I said, the yield changes with the RPM. You want it optimal at 200 RPM for full size bikes. The motor is wound so that you get optimal yield for the voltage you are going to use.

You are going to shed more heat proportionally unless you ride always at the optimal RPM.

When you overvolt one of my motors, you move the optimal RPM to 260. So instead of enjoying best speed at 15mph, it's going to be 20mph.

I know it's fine for some, but not what I am supposed to do.

 

I should have made myself clearer. I'm mainly talking about near stall conditions, eg starting off, slowly climbing a steep hill, heading into a strong wind.

I think there is some things people are forgetting.

Firstly, to get more torque, (other than changing gearing) you always have to put more current through the motor, as torque is proportional to current.

 

Battery current is not the same as phase (motor) current.

 

The extra torque that comes from overvolting is caused by extra motor current over and above the battery current, as motor system steps down voltage and steps up current. (acts like buck converter)

This can be demonstrated below with motor simulator, as motor current of 20.5 amp at 36v X 18.2amp controller is same as 48v X 14 amp, both representing a 30% increase in normal 36v X 14amp controller.

There is no magic way (in this case) of getting more torque without more heat in motor, without changing gearing. e.g. Volt up,Gear down.

 

https://www.ebikes.ca/tools/simulator.html?motor=MXUS_XF07&cont=cust_18.2_70_0.03_A&wheel=700c&mass=100&hp=50&bopen=true&cont_b=cust_14_70_0.03_A&motor_b=MXUS_XF07&wheel_b=700c&mass_b=100&hp_b=50&batt_b=B4816_GA&grade=8&grade_b=8

 

 

Can you explain how the controller is acting like a Buck converter? To do this requires a significantly sized inductor, which I have not seen in any controller I have taken apart.

 

There is a small Buck converter for the control electronics, but not as far as I am aware for the main drive.

 

I would like to see an oscilloscope trace of controller output, demonstrating that the controller is actively reducing voltage (other than being unable to supply enough current and having it's voltage pulled down).

you need to think of electricity like waves, the voltage at the storage capacitor is not same as battery's voltage. It is wavy, although connected to the battery with two short fat wires, black and red.

the increase in motor current that is needed for more torque is partially supplied by the 2.2mF storage capacitor.

Higher voltage increases the energy stored, thus gives higher discharge current through the coils.

Oscillations caused by the back EMF in an overvolted motor create also 30%-50% higher voltages when riding at high speed because the magnets in an overvolted motor are working at over saturated state. When the increase current cannot create any more magnetic flux, its wave is reflected back like a sea wave hitting a rock. Same thing as regen, spontaneous turbulent voltage is much higher than your battery's voltage.

The thing you should worry about is the width of the tracks on your controller's PCB. Wave travels only on the surface of the conductor. Higher wave creates more heat on the tracks.

If your storage capacitor is not good enough, the back EMF may kill this capacitor at set fire to the controller.

Rule of thumb: lower voltage = better reliability.

 

 

The 2,200 microF capacitor is for filtering and smoothing, not for driving the motor when output from the controller drops. The energy in a capacitor is 0.5*C*V squared. At 40V, this is less than 2 Joules of energy, that's not going to drive a "250W" motor for very long!

 

These days there is not much point for controller manufacturers to make 36V specific systems and 48V specific systems. It's much cheaper to have one PCB, one production line, one source of standard conponents and just solder a bridge in for the desired LCV setting. So in most cases these days, "36V" controllers have no problem handling 48V and this is well within their spec and not stressing them significantly.

 

In terms of shorter life, (which is debatable) I'd much rather increase the life of the motor than the controller as it's much easier and cheaper to replace the controller.

At 40V, this is less than 2 Joules of energy, that's not going to drive a "250W" motor for very long!

you got it wrong. The pulses going through the coils are about 22kHz-40kHz.

the pulse going through the coils lasts only about 0.04ms.

 

what is the power of 2J delivered in 0.04ms?

 

It's not the battery that can deliver that power. It's the job for the storage capacitor.

I think there is some things people are forgetting.

Firstly, to get more torque, (other than changing gearing) you always have to put more current through the motor, as torque is proportional to current.

 

Battery current is not the same as phase (motor) current.

 

The extra torque that comes from overvolting is caused by extra motor current over and above the battery current, as motor system steps down voltage and steps up current. (acts like buck converter)

This can be demonstrated below with motor simulator, as motor current of 20.5 amp at 36v X 18.2amp controller is same as 48v X 14 amp, both representing a 30% increase in normal 36v X 14amp controller.

There is no magic way (in this case) of getting more torque without more heat in motor, without changing gearing. e.g. Volt up,Gear down.

 

https://www.ebikes.ca/tools/simulator.html?motor=MXUS_XF07&cont=cust_18.2_70_0.03_A&wheel=700c&mass=100&hp=50&bopen=true&cont_b=cust_14_70_0.03_A&motor_b=MXUS_XF07&wheel_b=700c&mass_b=100&hp_b=50&batt_b=B4816_GA&grade=8&grade_b=8

 

 

If voltage is irrelevant, why don't we just use 12V systems and wind our motors accordingly to spin at the right RPM? Voltage does have a role to play, it improves efficiency and reduces current demand for a given Wattage.

 

There seem to be a lot of people giving opinions (myself included!) without the necessary understanding of motor electronic engineering and control. I'd like to hear from someone who really understands these things and has studied this matter in depth. Otherwise this is just going to generate more heat than light.

 

As time goes on, I will make measurements, keep experimenting etc and gain more knowledge. But in the mean time I will continue to use my bike at higher voltages and enjoy the better ride that it gives.

If voltage is irrelevant, why don't we just use 12V systems and wind our motors accordingly to spin at the right RPM? Voltage does have a role to play, it improves efficiency and reduces current demand for a given Wattage.

 

There seem to be a lot of people giving opinions (myself included!) without the necessary understanding of motor electronic engineering and control. I'd like to hear from someone who really understands these things and has studied this matter in depth. Otherwise this is just going to generate more heat than light.

 

As time goes on, I will make measurements, keep experimenting etc and gain more knowledge. But in the mean time I will continue to use my bike at higher voltages and enjoy the better ride that it gives.

high voltage is important to reduce transport loss but bad for the longevity of the components and incurs bigger loss through EM radiations when you use in an oscillating circuitry like e-bike motor.

I only use higher 48V instead of 36V to reduce the stress on the battery cells.

 

In my days, CPUs run at 5V when I built my first computer using Motorola 6800 CPU, TTL circuits and 1kilobit static memory chips. You can't have the density of modern intel CPUs if they still runs on 5V. Less voltage gives better reliability.

high voltage is important to reduce transport loss but bad for the longevity of the components and incurs bigger loss through EM radiations when you use in an oscillating circuitry like e-bike motor.

I only use higher 48V instead of 36V to reduce the stress on the battery cells.

 

In my days, CPUs run at 5V when I built my first computer using Motorola 6800 CPU, TTL circuits and 1kilobit static memory chips. You can't have the density of modern intel CPUs if they still runs on 5V. Less voltage gives better reliability.

 

 

 

This just proves my point.

 

Sometimes being on this forum is like listening to a load of old men sat in the corner of the pub, telling each other more and more fantastic tales. Each of the others has enough knowledge to cast doubt on what the speaker is saying, but not enough to disprove it out right...while the speaker has enough knowledge to make their story have some credence, but not enough to prove its truth out right.

 

It's entertaining but right now I need to get on with some work!

 

you still have not answered my question:

you said, "At 40V, this is less than 2 Joules of energy, that's not going to drive a "250W" motor for very long! "

 

My question is:

 

so knowing that the pulse width is 0.04ms, what is the power delivered through the coils?

 

Answering this question may lead you to understand the motor better than I could explain.

you got it wrong. The pulses going through the coils are about 22kHz-40kHz.

the pulse going through the coils lasts only about 0.04ms.

 

what is the power of 2J delivered in 0.04ms?

 

It's not the battery that can deliver that power. It's the job for the storage capacitor.

2J per 0.04secs is 0.08W!

2J per 0.04secs is 0.08W!

Very funny.

You know where I am going with this.

 

Just to remind ourselves:

 

Power = energy / time, a 22KHz signal has a wavelength of 1s/22,000 = 0.045ms = 0.000045 second.

If your 48V battery is fully charged, the battery voltage is 53.5V.

Your storage capacitor has 2.2mF capacitance, it stores this much energy:

 

E = 0.5*V*V*C = 0.5 * 0.0022 * 53.5 * 53.5 = 3.14 Joules

 

If it was all discharged, the average power in the pulse would be 3.14/0.000045= 70KW.

Of course the average pulse does not have 53.5V amplitude but is still quite considerable.

 

Here is a picture of a very SLOW trace, the motor is on startup where the duty cycle is set to 5%. I chose it because the trace is clean, it illustrates the principle of BLDC motor without suffering back EMF interference.

 

http://vedder.se/wp-content/uploads/2014/08/start_d5_almost_auto_voltage-1024x408.png

The capacitor's job isn't to power the motor. It's there to smooth out the battery current. As you rightly say, the demand from the controller is high frequency pulses at an average current of say 17 amps., when going slowly up a hill. To achieve that average, the individual pulses would have to be much higher, which would most likely trip the BMS maximum current, especially at low rpm, where you can hear the individual pulses. The capacitor smooths that out to make a continuous flat demand at 17 amps.

 

If you look in a range of controllers, you'll see that there's a substantial difference in the size of that/those capacitors. If they drove the motor, you'd have to have different sizes depending on the power you used, but the sizes you see in the controller have no relation to the power output.

you need to remember electricity is wave.

If it was flat, the signal across the Atlantic from the earliest telegraph cable (1858) would not have been scrambled.

The early engineers did not undesrtand this, they increased the voltage until the cable burned out.

The cable functioned for only three weeks, but it was the first such project to yield practical results. The first official telegram to pass between two continents was a letter of congratulations from Queen Victoria of the United Kingdom to President of the United States James Buchanan on August 16. Signal quality declined rapidly, slowing transmission to an almost unusable speed. The cable was destroyed the following month when Wildman Whitehouse applied excessive voltage to it while trying to achieve faster operation.

 

Look at the wave forms on the oscilloscope trace. The FET connects directly the coils to your battery. If the wave were flat, how do you explain that trace?

 

Now follow the signal along the path from the battery to the coils. The world around us is oscillating in nature. Remember the most fundamental law in the universe: when a particle changes direction, it emits electromagnetic waves.

You must think beyond the simplification of averages.

 

The easiest way to understand electrical conductivity in an oscillating circuitry is to understand how current flows in a rectilinear aerial.

 

This is a half wave dipole:

 

If they drove the motor, you'd have to have different sizes depending on the power you used, but the sizes you see in the controller have no relation to the power output.

the controller can use the pulse width and frequency modulation to adjust how much power is sent to the motor. As I've shown in the previous post, the duty cycle of that trace was 5%. A 2.2mF capacitor can transmit a lot of power, 1kW is easy peasy. If you need more power, you increase the frequency. A stroboscope is a simple example of high discharge current of a capacitor about 0.1 microfarrad.

The writing on that chart you showed is too small to see what the chart represents. This is an actual commutation trace of an ebike motor:

that's a better pictures. You can see clearly how fast the pulses are.

If you can program the controller yourself, you can probably make the motor more flexible without changing the hardware. A bit like the free open source firmware stuff for the TSDZ2.

By the way, I took the Brompton kit for a long-ish ride yesterday. Set the wheel diameter to 16" - maintained 18mph both outbound and inbound. The assistance seems to keep up with my cadence, never hit the treacle. I still have to find an app for GPS test. Quite a few people took an interest at the Brompton. Maybe because I rode it derestricted (on French roads).

Edited July 6, 20205 yr by Woosh

All very interesting to us non technical people, but I don't want to go faster, I don't want to go further.

All that I want for my folding hub motor bike is to climb steep hills.

Don't tell me about soldering shunts or even about soldering.

If the manufacturer set the controller just so, obviously it's not meant to give a higher amperage.

Can you explain how the controller is acting like a Buck converter? To do this requires a significantly sized inductor, which I have not seen in any controller I have taken apart.

 

There is a small Buck converter for the control electronics, but not as far as I am aware for the main drive.

 

I would like to see an oscilloscope trace of controller output, demonstrating that the controller is actively reducing voltage (other than being unable to supply enough current and having it's voltage pulled down).

Hi. The 'significantly sized inductor' is the motor winding itself. I have attached a good educational booklet explaining how these motors behave during pulse width modulation. Page 14 & 15 in particular gives some explanation of how this extra current occurs during the 'diode freewheeling periods'. The controller acts in conjunction with the inductance of the motor. When the mosfet turns off, there is a circuit for the stored induced magnetic energy in the motor coils to continue through the 'body diodes'. This extra current gives extra motor torque and both extra current (in table) and extra torque (in graph) are shown on the motor simulator link I posted earlier and reposted below.

https://www.ebikes.ca/tools/simulator.html?motor=MXUS_XF07&cont=cust_18.2_70_0.03_A&wheel=700c&mass=100&hp=50&bopen=true&cont_b=cust_14_70_0.03_A&motor_b=MXUS_XF07&wheel_b=700c&mass_b=100&hp_b=50&batt_b=B4816_GA&grade=8&grade_b=8

Toshiba.pdf

Edited July 6, 20205 yr by Sturmey

Hi. The 'significantly sized inductor' is the motor winding itself. I have attached a good educational booklet explaining how these motors behave during pulse width modulation. Page 14 & 15 in particular gives some explanation of how this extra current occurs during the 'diode freewheeling periods'. The controller acts in conjunction with the inductance of the motor. When the mosfet turns off, there is a circuit for the stored induced magnetic energy in the motor coils to continue through the 'body diodes'. This extra current gives extra motor torque and both extra current (in table) and extra torque (in graph) are shown on the motor simulator link I posted earlier and reposted below.

https://www.ebikes.ca/tools/simulator.html?motor=MXUS_XF07&cont=cust_18.2_70_0.03_A&wheel=700c&mass=100&hp=50&bopen=true&cont_b=cust_14_70_0.03_A&motor_b=MXUS_XF07&wheel_b=700c&mass_b=100&hp_b=50&batt_b=B4816_GA&grade=8&grade_b=8

 

Thanks Sturmey, I'll have a read and try and make sense of it. Hopefully you are the "wisest old man" in this particular pub

 

I'm not sure if I'm getting more confused or less after looking at the simulator

 

If the motor current is vastly different from the battery current, are controllers rated in terms of what they take from the battery rather than they supply to the motor?

 

What is interesting from the simulator is that when you look at the torque with the higher current controller, it drops off quite quickly well before the 25kph limit, whereas the 48V lower current version still has plenty of torque. So if you set the gradient to 4.5%, no human input, then the 36V system can only sustain 21.5 kph max, whereas the 48V system is still able to sustain 24.6 kph. It looks like there is a sweet spot, where so long as you can sustain more than 20 kph on gradients, the 48V system will perform better in terms of speed between 20 and 25 kph than the 36V system.

 

This may be why I'm getting good results, I only need PAS setting 1 and it gives nice and smooth acceleration all the way up to 25 kph and on hills around me, climbing is better. On extremely steep hills, according to the simulator, it shouldn't make any difference.

 

It's worth noting on the site, they don't specify any particular voltage for their motors.