Review of Argos £285 Folding Bike

[saneagle]

The capacitors in the controller and the inductors (phase windings) in the motor act together to some extent like a buck converter so both ways of seeing it are useful, especially 'average' voltage and current. In a basic hub motor system as the bike slows the back EMF from the motor falls, so out of the controller's average output voltage there's more left over to drive more average current through the motor, resulting in more torque. I'm not a 'wood wide web' person so I wouldn't call it communication between controller and motor, it's the physics happening.

I prefer simple science. Energy comes out of the battery and goes into the motor, regulated by the controller. If you put an oscilloscope on the controller output, you see a square wave of height corresponding to the battery voltage and each pulse length determined by the controller to get whatever simulated wave form and frequency it chooses. The battery voltage pushes the energy of each pulse into the motor. It matters not what happens to that energy when it arrives at the motor, and we know that energy can only be transferred.

 

[Ghost1951]

It's pulse width modulation isn't it. Full battery voltage, but the mosfets are turned on for either less or more time depending on the power required. More power means a greater proportion of the time the mosfets are on full, and a smaller proportion of the time for less power - this being determined by the power algorithm, the PAS setting and or throttle if applicable.

I went looking online for a diagram showing pulse width modulation and this is the only one I can find right now. Obviously, this one is a square wave. Some are sine wave but the same idea applies.



MOSFETs in class E bias are incredibly efficient at handling large currents and very little is lost as heat in the mosfet, which is really just a switch in that mode.

 

[saneagle]

It's a bit like pulse width modulation isn't it. Full battery voltage, but the mosfets are turned on for either less or more time depending on the power required. More power means a greater proportion of the time the mosfets are on full, and a smaller proportion of the time for less power - this being determined by the power algorithm, the PAS setting and or throttle if applicable.

I went looking online for a diagram showing pulse width modulation and this is the only one I can find right now. Obviously, this one is a square wave. Some are sine wave but the same idea applies.

View attachment 59650

That's it. The sinewave controllers have a different width for every pulse, with the width varying in a sinewave.

 

[Ghost1951]

That's it. The sinewave controllers have a different width for every pulse, with the width varying in a sinewave.

I am more familiar with this concept in digital radio transmitters. I have three small, home built weak signal transmitters which operate mosfets as power amplifiers, albeit the maximum power out is only about four watts. The mosfet switching speed is the radio frequency being employed in the radio transmitter and the load is not a motor but a filter circuit of coils wound on toroids and capacitors and the antenna which all has to be resonant so the power can leave the system and not come back and heat up the PA. If the load is not resonant the mosfet goes out of class E and heats up to destruction. It doesn't take long One key parameter in this particular application is the gate capacitance of the particular mosfet, because they are switching on and off at over 14 million times a second. If the mosfet proposed has too high agate capacitance it probably won't operate at that kind of frequency. The gate is driven by a switching transistor powered by a tiny programmable synthesiser chip which is controlled by a micro controller. The cheap and chearful little BS170 is used - three of them in parallel. They cost pennies per unit - about 10p.

 

[Bonzo Banana]

How is the current controlled with a hub motor and controller, so the duty cycle controls voltage by the length of the duty cycle and that is fed through capacitors to smooth it out but in a hill climbing situation where the motor is bogged down at a slow speed and using maximum power and we know that a lot of hill climbing leads to a much more limited range from the battery pack. It feels like when the duty cycle is low and there isn't much voltage going to the motor there isn't much current either. How is the controller drawing more current for hill climbing or is the torque created only related to voltage? So when the motor is bogged down climbing a steep hill is voltage just transformed from speed to torque somehow? Hill climbing takes a lot more power than just going fast on a flat surface at the maximum assistance speed. Is the duty cycle 100% for hill climbing but significantly less for providing the maximum assistance speed of 15.5mph?

How does it all work with a direct drive hub motor where they can work from about 200W to 2000W and voltage means much higher speeds but torque is often much more limited in its improvement, what can be 10x as much wattage might only mean a 50% improvement in torque i.e. 30Nm to 45Nm? Does this prove your point in that voltage provides speed control but current control is very limited.

I must admit I can't quite get my head around the overall picture because I'm struggling to think of voltage as torque I guess.

So if an ebike was going along at 15.5mph on the flats and then hits a huge hill and it now could only generate a speed of 5mph, what is the controller doing in those circumstances? Maximising the duty cycle at 100% and re-timing the synchronisation of the hub motor for torque?

 

[saneagle]

How is the current controlled with a hub motor and controller, so the duty cycle controls voltage by the length of the duty cycle and that is fed through capacitors to smooth it out but in a hill climbing situation where the motor is bogged down at a slow speed and using maximum power and we know that a lot of hill climbing leads to a much more limited range from the battery pack. It feels like when the duty cycle is low and there isn't much voltage going to the motor there isn't much current either. How is the controller drawing more current for hill climbing or is the torque created only related to voltage? So when the motor is bogged down climbing a steep hill is voltage just transformed from speed to torque somehow? Hill climbing takes a lot more power than just going fast on a flat surface at the maximum assistance speed. Is the duty cycle 100% for hill climbing but significantly less for providing the maximum assistance speed of 15.5mph?

How does it all work with a direct drive hub motor where they can work from about 200W to 2000W and voltage means much higher speeds but torque is often much more limited in its improvement, what can be 10x as much wattage might only mean a 50% improvement in torque i.e. 30Nm to 45Nm? Does this prove your point in that voltage provides speed control but current control is very limited.

I must admit I can't quite get my head around the overall picture because I'm struggling to think of voltage as torque I guess.

So if an ebike was going along at 15.5mph on the flats and then hits a huge hill and it now could only generate a speed of 5mph, what is the controller doing in those circumstances? Maximising the duty cycle at 100% and re-timing the synchronisation of the hub motor for torque?

I think you have it the wrong way round. The voltage is fixed by the battery, and the current is given by the duty. You talk about average values, but in that case it's better to see each pulse ASAP pulse of energy, height times width giving Joules. The battery voltage is very relevant because it works in conjuction with the back emf to have an effect on power and maximum speed.

Forget about the main capacitor. It's not so much for smoothing as giving the instant current needed for each pulse to aid the battery.

The more current when going slow and less current when going fast is a direct result of the back emf. It's the net voltage (battery voltage minus back emf) that pushes current into the motor. As the motor speeds up, the effective height of each pulse reduces.

When you increase the battery voltage, the height of each pulse increases directly, so there's more energy per pulse to increase the torque.

My explanation is not strictly scientifically correct, but gives a layman's view so that you can see cause and affect. What happens inside the motor is extremely complicated, but, as I said earlier, we can see how much energy comes out of the controller.

 

[Ghost1951]

Also, the algorithm of the controller is intervening as road speed rises and falls on hills and undulations all added to by your pas settings and also perhaps pedalling cadence. So you have a combination of physics in the back emf and current inside the motor and the intervention of the controllers reaction to its inputs.