E bikes and the need for an electronic gear box

[John Heath]

I recently bought an E bike , 1000 dollars new Canadian , as insurance against the car breaking down. Turns out these fears were well founded as the car did break down leaving nothing but the E bike to navigate traffic. As most in this group can relate to one can not have a new toy to play with without taking it apart or worse redesigning. I guess it is in human nature despite the fact that it voids the warranty. The way I see it what is the point in having an E bike if you can not have fun changing this and that.

This bike is a ecoped pulse grey with 4 20 amp hour 12 volt batteries and direct drive hub motor. Regenerative braking , that's cool , and top speed 32 KMH limited by firmware. 

http://www.ecoped.com/en/pulse

The overall design of this bike is excellent for what it is. However it has one Achilles heel. It can not climb a hill that is better than 12 degrees.  The bike screams for a lower gear as the raw power to climb is there , 500 watts , almost 1 horse power. The problem is the hub motor is optimized for 32 KMH not 10 KMH. The bike screams for a lower gear ratio to climb hills without over heating the hub motor and mos output transistors. As the car is now fix attention can be payed to the E bike's missing lower gear to climb hills.

Physical planetary gears in the hub motor is out of the question as it is beyond my skills. There should be a way to accomplish this electrically.

I have a few ideas on the table. Switch from a 48 volt system to a 24 volt  system and hope the controller board can live with 24 volts. With only 24 volts the counter EMF from the hub should bottom out at about 15 or 20 KMH. This would save battery power going up hill. However this is avoids the big problem " where to find higher torque " to push the bike uphill without over heating the hub and controller output transistors? Any thoughts on this would be welcome.   

[jitter]

How are you going to introduce an extra gear without adding a mechanical ratio?
I'm guessing the current going to the motor is limited by the electronics to prevent overheating, because as you say, the power is there.

So dropping the battery to 24 V and doubling the current capability IMHO isn't going to be the solution.
BTW, contrary to internal combustion engines, electric motors can deliver huge torque at very low revs, so it's not going to stall. The manufacturer says 16 degrees slope is the max.

I guess what you need is a hub with a different ratio. Perhaps the manufacturer sells different configurations to different parts of the world?
I'm wondering what they would sell in e.g. the Alps...
Edit... 16 degrees equals 28.7%. That's huge, you won't find those even in the Alps, maybe on some forestry roads but not for normal traffic. More than 12% (7 degrees) can be found but are probably few and far between.

Hehe... 25%... (probably in Wales) no wonder the biker's walking...

[voltsandjolts]

I modified by Kona PHD commuter with a Bafang mid drive motor setup. I've never ridden a hub motor bike but I specifically went for mid drive so the motor can make use of the existing gear ratios on steep climbs - works great. Search keywords: Bafang BBS01 (250W and 350W) or BBS02 (500W and 750W) or BBSHD (1kW). Not sure if there is much you can do to change your hub motor gearing - build it into a smaller wheel?!

[Pjotr]

I have a few ideas on the table. Switch from a 48 volt system to a 24 volt  system and hope the controller board can live with 24 volts. With only 24 volts the counter EMF from the hub should bottom out at about 15 or 20 KMH. This would save battery power going up hill. However this is avoids the big problem " where to find higher torque " to push the bike uphill without over heating the hub and controller output transistors? Any thoughts on this would be welcome.

There is an old saying that still holds: "Voltages are cheaper than amps". This has to do simply with I2R losses, cable/wire diameter etc.

[John Heath]

How are you going to introduce an extra gear without adding a mechanical ratio?
I'm guessing the current going to the motor is limited by the electronics to prevent overheating, because as you say, the power is there.

So dropping the battery to 24 V and doubling the current capability IMHO isn't going to be the solution.
BTW, contrary to internal combustion engines, electric motors can deliver huge torque at very low revs, so it's not going to stall. The manufacturer says 16 degrees slope is the max.

I guess what you need is a hub with a different ratio. Perhaps the manufacturer sells different configurations to different parts of the world?
I'm wondering what they would sell in e.g. the Alps...
Edit... 16 degrees equals 28.7%. That's huge, you won't find those even in the Alps, maybe on some forestry roads but not for normal traffic. More than 12% (7 degrees) can be found but are probably few and far between.

Hehe... 25%... (probably in Wales) no wonder the biker's walking...

Great picture for the subject at hand. I can not say if the controller has current limiting on the output transistors. I hope so . As with most electronics today a diagram is hard to find as they do not troubleshoot down to a component level. They just sell you a new unit. I did some calculations after posting and maybe it does not have the power to climb hills faster.

1 KMH = 1 foot per second
1 horse power = 550 pounds lifted 1 foot per second
1 horse power = 750 watts
The hub motor says 350 watts peak 500.

Lets call the hub motor 1/2 horse power for 375 watts so it should lift 550 pounds 1/2 foot per second. The bike weighs 160 pounds and I weigh 160 pounds . Total weight of bike with myself is 320 pounds. Considering heat loss lets say the hub motor should lift myself and the bike 1/2 foot per second straight up. At 10 KMH the up hill sloop would have to be 1/2 foot up per 10 feet forward for the 1/2 horse power hub. My head is starting to hurt as I have no idea what that slope would be in degrees. Less than 16 degrees for sure. However the peak surge power of the hub motor could be in the 1 horse power range making 16 degrees a possibility.

I can say from experience that this bike is okay on moderate slopes but on a steep slope it is not going to happen. For that reason this bike would not be practical in a hilly city. Most slopes where I am are moderate so it is fine. It should be noted that my bike sells for 1000 dollars retail at Walmart while the competition starts at 2500 dollars with their fancy lithium battery pack that cost as much as my bike new including lead acid batteries.

     

[John Heath]

I modified by Kona PHD commuter with a Bafang mid drive motor setup. I've never ridden a hub motor bike but I specifically went for mid drive so the motor can make use of the existing gear ratios on steep climbs - works great. Search keywords: Bafang BBS01 (250W and 350W) or BBS02 (500W and 750W) or BBSHD (1kW). Not sure if there is much you can do to change your hub motor gearing - build it into a smaller wheel?!

Found it. there advantages to standard bike kits as the weight is less and as you say you can fiddle with the gear ratios.

[John Heath]

I have a few ideas on the table. Switch from a 48 volt system to a 24 volt  system and hope the controller board can live with 24 volts. With only 24 volts the counter EMF from the hub should bottom out at about 15 or 20 KMH. This would save battery power going up hill. However this is avoids the big problem " where to find higher torque " to push the bike uphill without over heating the hub and controller output transistors? Any thoughts on this would be welcome.

There is an old saying that still holds: "Voltages are cheaper than amps". This has to do simply with I2R losses, cable/wire diameter etc.

Yes and good point. A 48 volt system can achieve 1 horse power at 750 watts with just 15 amps. The wires can be 18 gauge without losing too much to heat.

[amyk]

Lets call the hub motor 1/2 horse power for 375 watts so it should lift 550 pounds 1/2 foot per second. The bike weighs 160 pounds and I weigh 160 pounds . Total weight of bike with myself is 320 pounds. Considering heat loss lets say the hub motor should lift myself and the bike 1/2 foot per second straight up. At 10 KMH the up hill sloop would have to be 1/2 foot up per 10 feet forward for the 1/2 horse power hub. My head is starting to hurt as I have no idea what that slope would be in degrees. Less than 16 degrees for sure. However the peak surge power of the hub motor could be in the 1 horse power range making 16 degrees a possibility.

320lb = 145.5kg = 1425 N

If you want to go up a 12 degree slope at 10km/h, i.e. 2.78m/s, the vertical component of the velocity will be 2.78 sin 12deg = 0.58m/s. 1425 N * 0.58m/s = 827W, and this is neglecting friction and other losses. With 375W you will achieve on the 12 degree slope a theoretical maximum of 0.26m/s or 0.95km/h. 500W gives 1.26km/h. Even if you have the gearing to allow the motor to run in its maximum power range, you won't be going very fast.

At 10 KMH the up hill sloop would have to be 1/2 foot up per 10 feet forward for the 1/2 horse power hub. My head is starting to hurt as I have no idea what that slope would be in degrees.

arcsin(1/20), or around 2.87 degrees.

[John Heath]

Now that I think of it just draw a line 20 inches then lift one end 1 inch and there is your angle old school slide rule style.  Back to more torque. If the hub 3 phase winding are wound in parallel from opposite sides , 50 / 50 chance , then switching them to series should produce 2 times the torque. Not sure if my meaning is clear. 10 amps is split to 5 amps and 5 amps for a given phase at opposite sides. If they are rewired to be in series then it is 10 amps on both sides and the hub counter EMF would be 2 times for a lower RPS counter EMF equilibrium. Sounds like energy is conserved so this should work providing they wound the opposite coils in parallel. This would be a true electronic transmission with 2 gears for low and high speeds. Hmmmm. 

[Berni]

You probably want to be careful with squeezing more torque out of the existing motor. As you send larger currents trough it or change the coil configuration you are increasing the magnetic field strength inside the electromagnets. If that field becomes too strong it can start to re-magnetise the permanent magnets in the opposite direction, causing them to loose strength and making the motor develop less torque.

[John Heath]

I see your point. The natural tendency of a permanent magnet is to demagnetize over time for a lower energy domain orientation. Too much current could hurry that process along. Old school magnets could be demagnetized just by dropping them. From what I googled the magnets are on the outside of the hub glued to a ring that is part of the wheel. This way the coils and electronics do not rotate.

[Berni]

Actually magnetizing things is a bit like springs.

When you bring a ferromagnetic material in to a magnetic field the field rotates the internal "tiny magnet" structures so that they all point in one direction, this makes non magnetized iron attracted to a magnet since it becomes a magnet it self, but if the field is strong enough it reorients the tiny magnets inside so far that they don't spring back in to there initial randomized direction, just like overstretching a spring to the point where it doesn't return to the initial state anymore. Because of this the iron stays magnetized after the magnet is removed.

Now a permanent magnet is just the same, except its starts with those tiny magnets in a very oriented state in order to give it a very strong magnetic field. Hitting them or making them hot can randomize that structure yes, but if you apply a sufficiently strong magnetic field in the opposite direction it will pull over the tiny magnets far enough that they don't return. Permanent magnets are designed to resist this as much as possible in order to keep being a magnet under various conditions, but with a strong enough field you can re-magnetize any magnet. When the field is pointing in the opposite direction its re magnetizing it towards zero so it basically demagnetizes it.

This is the reason why motors with permanent magnets are limited in terms of peak torque (long therm they tend to be thermally limited). In order to get more torque you need a stronger field and at some point the field becomes large enough to demagnetize its permanent magnets. There are also other things to worry about like the magnetic core getting saturated with the field, causing the electromagnets to become less effective at producing the field where its needed.

The way designers tend to get around this problem is by running permanent motors at ridiculously high RPM and then gearing them down to the speed they want. Thats the reason why those brushless motors for RC planes can produce like 1kW in a tiny motor.

[amyk]

Alternatively, this is why large motors all exclusively use wound fields and armatures; large permanent magnets are also extremely difficult to manufacture and handle.

[Delta]

Have you thought about, erm, pedalling?

[tggzzz]

Hehe... 25%... (probably in Wales) no wonder the biker's walking...

Pah! Side roads don't count. Why not use the A39 in Somerset as an example? (In the UK, a lower number denotes a more important and/or earlier road; 39 is not high!)
https://www.google.co.uk/maps/@51.2047417,-3.6063398,3a,75y,59.19h,74.19t/data=!3m6!1e1!3m4!1sEx3wC-QtI5zucugc_s2kyg!2e0!7i13312!8i6656

There's even a (single track) toll road to bypass it, should you feel the need.

[John Heath]

Actually magnetizing things is a bit like springs.

When you bring a ferromagnetic material in to a magnetic field the field rotates the internal "tiny magnet" structures so that they all point in one direction, this makes non magnetized iron attracted to a magnet since it becomes a magnet it self, but if the field is strong enough it reorients the tiny magnets inside so far that they don't spring back in to there initial randomized direction, just like overstretching a spring to the point where it doesn't return to the initial state anymore. Because of this the iron stays magnetized after the magnet is removed.

Now a permanent magnet is just the same, except its starts with those tiny magnets in a very oriented state in order to give it a very strong magnetic field. Hitting them or making them hot can randomize that structure yes, but if you apply a sufficiently strong magnetic field in the opposite direction it will pull over the tiny magnets far enough that they don't return. Permanent magnets are designed to resist this as much as possible in order to keep being a magnet under various conditions, but with a strong enough field you can re-magnetize any magnet. When the field is pointing in the opposite direction its re magnetizing it towards zero so it basically demagnetizes it.

This is the reason why motors with permanent magnets are limited in terms of peak torque (long therm they tend to be thermally limited). In order to get more torque you need a stronger field and at some point the field becomes large enough to demagnetize its permanent magnets. There are also other things to worry about like the magnetic core getting saturated with the field, causing the electromagnets to become less effective at producing the field where its needed.

The way designers tend to get around this problem is by running permanent motors at ridiculously high RPM and then gearing them down to the speed they want. Thats the reason why those brushless motors for RC planes can produce like 1kW in a tiny motor.

You have made some good points. In the case of magnetizing the iron it of less concern in this case as it is AC driven with polarity changing every phase transition. It would heat up the iron but magnetization would be a net 0 with an AC drive. However saturation could be a problem. I imagine they used as little iron as possible to keep the weight down. If I increase current by 2 it could be well into the saturation point of the iron. I made an interesting observation from just using the e bike. Torque is high when batteries are charged. After about 2 hours torque is less. Wires are the same , temp is the same and voltage with no load is within 3% so why less torque? My only conclusion is the battery impedance changes form 90 % charged to 40 % charged.  If this is true then the batteries must be getting warm to reflect this increase in impedance. I put my hand on the batteries and they were warm. The batteries are 4 x 20 amp hour 12 volt gel lead acid type.

[johansen]

short answer is it will be very difficult to get significantly more torque from the motor without also burning up the windings. i can't tell you off the top of my head what it takes to demagnetize them, but its on the order of 2T minimum, this is achievable in a motor. 3T is probably what it takes and you cannot generate those fields without burning up your motor.. literally the windings would melt in just a few seconds.

if you can get your lead acid batteries warm during discharge, they are not going to last very long.

also the battery terminal voltage will drop under load, significantly as the battery discharges. given that you are discharging them fast enough to get them warm, i would estimate you have 11.5 volts available under full load initially, at 40% capacity remaining, your battery terminal volts are probably around 9.5. this will account for part of your torque drop. also at lower battery terminal volts, the maximum speed of your bike will drop proportinally.


it is possible to run a synchronous (aka outrunner, bldc or brushless) motor above its nominal "speed limit" but you have to recirculate current in the windings to demagnetize the motor. this can be done, up to about double the rpm, but it costs you a lot in terms of copper loss in the motor, and power loss in the controller.

i dunno, perhaps 95% of ESC or E bike motor controllers cannot run the motor faster than the battery  voltage at unity power factor. basically spin the motor with a drill, rectify the ac power, measure the dc voltage, that dc voltage can't exceed the battery. --that's where the speed limit comes in.


---so it might be possible to wire in a Wye/Delta switch.. the problem is the phase of the hall effect sensors would be thrown off. 

such a winding change can be done with relays and would get your maximum speed up to 1.73 times as fast. if its already delta then you could cut the windings in half and then reconnect for wye, it just gets more complicated but can still be done.

[nuno]

Apparently those "bikes" have a good tolerance on the controller/motor, and I've seen a few people modding the controller for more current (= torque), by slightly decreasing the controller's current shunt (usually a piece of tick "copper" on the board, add some solder at one end to reduce the piece's length).
To help you not to go too far, if you try this you should pick a ridding course with a few Km and steep up hills; always start from a "cold" state and take your bike there before any change, take note of controller motor temperature at the end. Then make a small shunt decrease and redo the test, repeat until you're happy or the end temperatures change significantly.