Regenerating electricity on an e-bike

[Low-Q]

Hi there :-)
I got that crazy idea to make regeneration of electricity on my e-bike when the bike runs downhill, by using its motor as a generator when the throttle is off.
My e-bike is custom made by myself. The motor is located in the front wheel hub, and it provides 7 Volt 3-phase voltage increase for each 10km/h increased - when used as a generator.
I plan to use a powerful step up converter that runs from 10 - 60 VDC input. So I want to connect a 3-phase rectifier to the three wires that goes to the motor, and feed the step up converter.

The step up converter can deliver a steady 45VDC output as long as the input voltage is 10VDC or more input.

What my concerns is about, are the possible "short circuit" scenario when that converter is connected to the BMS and the motor. Because the converter wants to deliver 45V out, while the same wires from the battery, via the BMS delivers 42V to the ESC when the battery is not connected to a charger.
A BMS is a circuit that controls charge and discharge current, while it is balancing each of the 10 cells.
The drawback, is that the converter delivers 45V also when the bike is using its motor, forcing 45VDC input to the ESC.
Should I have a switch on the breaks or something that disconnect the converter when I do not need charging, or will the circuit work fine without a switch?

If you are an electronics engineer, I would be very happy if you could provide some information on how this is suppose to work.

 

[phinds]

I got that crazy idea to make regeneration of electricity on my e-bike when the bike runs downhill,

I don't understand why you would think that's crazy. Regenerative braking is old technology.
https://en.wikipedia.org/wiki/Regenerative_brake

 

[sophiecentaur]

Should I have a switch on the breaks or something that disconnect the converter when I do not need charging, or will the circuit work fine without a switch?

It would be easy to determine whether there is Power going to the wheels or from them because of the direction of flow of the current for a given polarity of Voltage. This problem has been solved by all successful regenerative braking systems. You should read the link from @phinds and as many other links (or even printed sources) about regenerative braking. You will not get far by just asking questions on PF because the topic has many facets. Best to come back with very specific questions, once you know a bit more or you will find that individual PF members will want you to do it 'their way' (mea culpa with that). There are many ways of skinning a cat.

 

[lekh2003]

I got that crazy idea to make regeneration of electricity on my e-bike when the bike runs downhill, by using its motor as a generator when the throttle is off.

Basically every electric vehicle uses this concept. I've seen few bikes which have done this, though several cars.

 

[sophiecentaur]

using its motor as a generator when the throttle is off.

Are you sure that this what you actually want in practice? For efficiency, the best use you can make of the Potential Energy at the top of a hill is to maximise your Kinetic Energy at the bottom, That will take you a long way up the next slope. The energy storage with a battery charging cycle is a lot less efficient than this so you would really need some control and choice of what to do when going downhill. Yes, regenerative braking is probably always a good idea but an automatic system cannot foresee your requirements in the same way as you can when cycling.
Also, an automatic system could probably reduce your average speed; not too important, perhaps but could be frustrating if your late arrival at the bottom of the hill could mean you missed a chance at traffic lights etc. A switch would be easy to use,

 

[Low-Q]

Thanks for the replies folks! My major concern was wether to use a switch or not to prevent possible shortcircuit or maybe a damaging feedback loop - depending on how you see it. The converter will feed the ESC input with 45V which possibly will conflict with the battery output. That means (possibly) the battery will never deliver anything and I must use the pedals all the time without any help from the motor - because the BMS is then in charging mode. Therefor I think a switch on the break handle will solve this.
The idea of regeneration might not be crazy, it's common on larger vehicles, but not on e-bikes as far as I know. Specially not the way I want it to work - all the time while just rolling down a hill.
I have taken your advice and listen to some "experts" on the field. I'll use a switch on the break handle, so it activates charging when activating the break handle a little - before the break pads touches the breaks disc.
The remaining discovery is wether or not I will take a salto over the steering wheel when the charging kics in...

 

[sophiecentaur]

I'll use a switch on the break handle, so it activates charging when activating the break handle a little

That sounds the way to do it - a bit of skill and 'prediction' could eliminate actual use of the brake almost entirely.

 

[Merlin3189]

I don't think this is done with commercial products, because the gain is too small to be worth it.

However, if you want to try, I wonder whether it would be a good idea to have a separate power storage cell to be charged during braking and use the power from that to drive a properly balanced charger while running or stationary. (Just my off the cuff reaction to your worries.)
This would seem to separate the two processes and avoid any unfortunate interaction. Switching between modes could be automatic.
The storage cell could be of lower voltage, not requiring upconversion while being charged. This would compensate for the need to upconvert to charge the main battery. There would of course be an extra 25% or so loss due to the extra charging cycle. Maybe a large capacitor bank could be used and be more efficient?

 

[sophiecentaur]

a separate power storage cell to be charged during braking

That would mean extra weight at all times and would not be necessary. It is quite possible to charge the battery with an intelligent switch mode charging system that would always avoid 're-circulating' energy, which you would want to avoid at all costs. The charge regulator could be current-based which would not involve any contention.
I have already made the point that an energy storage cycle with a battery involves far more loss than sticking with the KE you have with your moving bike, in the first place, so maintaining a good speed would be best unless you can foresee a reason to slow down in the near future. But that's just regenerative braking. If you feel like cycling whilst the regen process is going on then that would be 'money in the bank' and, importantly, it would be under the cyclist's control. A good charge / discharge display could let the cyclist manage effort and, for instance, make life easy towards the end of a trip when it would be possible to plan to arrive with no charge left, so no more pedalling needed. But that wouldn't involve machine intelligence - just a bit of learning on the cyclist's part,
PS Merlin makes a good point about the marginal advantage - the market is often right. (Not always though, or there would never be innovations.)

 

[Low-Q]

The parts I ordered for this experiment did cost almost nothing. I do this for the cause of the experiment and do not have high hopes. Just wanted some input from skilled people to prevent possible hazards :-). I will test my bike with and without regen. on my way to work - which is 29km one way. No charging before I'm home in the afternoon. Then I can monitor the difference in capacity after a couple of trips.
Large battery pack. 36V nominal at 42Ah, so no worries about capacity :-)
Vidar

 

[sophiecentaur]

Large battery pack. 36V nominal at 42Ah, so no worries about capacity :-)

What does that weigh? Up tp 30kg for lead acid? That represents about one third of my weight. It will certainly need to earn its keep. Did you consider a much lower capacity - just to get you up the awkward hills?
Reality check . . . . did you mean 4.2Ah capacity?

 

[rumborak]

This has probably popped into every cycling engineer's head at some point (inckuduin mine). The problem is that most energy is lost during riding by friction and wind resistance, none of which can be recouped. Because any device will add weight to the bike, thus requiring additional energy by the rider, there is essentially no way of ending up with a net positive.

 

[sophiecentaur]

This has probably popped into every cycling engineer's head at some point (inckuduin mine). The problem is that most energy is lost during riding by friction and wind resistance, none of which can be recouped. Because any device will add weight to the bike, thus requiring additional energy by the rider, there is essentially no way of ending up with a net positive.

But it can end up with spreading the worst loads from hill climbing over a bigger distance and also you can carry some electrical energy with you when you start your journey.

 

[Low-Q]

What does that weigh? Up tp 30kg for lead acid? That represents about one third of my weight. It will certainly need to earn its keep. Did you consider a much lower capacity - just to get you up the awkward hills?
Reality check . . . . did you mean 4.2Ah capacity?

It is really 42Ah, forty two :-)
One LiFePo at 20Ah on the luggage carrier behind the seat and a stack of LiPos at 22Ah in between the frame. Both with separate BMS. I have approx 130km range if I sweat a little.
The total weight is less than 30kg. The 20Ah battery is about 7kg. The Lipo is much lighter.

Edit: LiFePo weights 7.2kg. LiPo weights 4.5kg. Total weight of the e-bike is 30.5kg.

 

[Low-Q]

This has probably popped into every cycling engineer's head at some point (inckuduin mine). The problem is that most energy is lost during riding by friction and wind resistance, none of which can be recouped. Because any device will add weight to the bike, thus requiring additional energy by the rider, there is essentially no way of ending up with a net positive.

One thing is sure. Air resistance increase rapidly with velocity. It is the dominating factor as velocity pass 7'ish mph. So any hope of gaining back som of the energy on a horizontal ride is little. However, when climbing a hill, and not take the advantage of a "free ride" downhill will ofcourse shorten the batteryrange. Some of it can be charged back to the battery at moderate velocities. I'm not a racer, so taking it easy downhill will hopefully reduce loss from air resistance, and instead feed the battery with as much of the energy my body plus bicycle (Me 100kg + e-bike 30.5kg) can put back.

 

[sophiecentaur]

It is really 42Ah, forty two :-)
One LiFePo at 20Ah on the luggage carrier behind the seat and a stack of LiPos at 22Ah in between the frame. Both with separate BMS. I have approx 130km range if I sweat a little.
The total weight is less than 30kg. The 20Ah battery is about 7kg. The Lipo is much lighter.

Edit: LiFePo weights 7.2kg. LiPo weights 4.5kg. Total weight of the e-bike is 30.5kg.

Imressive! Quite a chunky little devil and it must take a lot of the effort of a journey that's within its range. After that . . . . .