Regenerative braking, without a battery?

Hi everyone!

I am undertaking quite a large project for my final year at university, and need some pointers as to where to go.

Basically I will need to design a quite specific controller that is capable of feeding power to a motor (on the scale of 24v) and then receiving power in the form of regenerative braking, and storing it in an Ultra cap. The ultra cap will be an array of 2.5v 300 farad polymer ultra caps, set up in grid formation so as to keep the 300 farad capacitance, but raise the voltage output to 25v.

The controller should be able to know when the ultra cap has reached it's maximum charge holding capacity (albiet rarely) and instead switch the current to a heat dissapating resistor array, and inform the user (via a small LED?) that the ultracap is fully charged.

The user should be able to 'switch' on the power flow to the motor, but the switch back to regenerative braking should be automatic.

Another key requirement is that the controller itself be powered by the ultracap, and require no replaceable battery, thus making it maintenace free. Or have a power usage so low that a small Li-ion battery (or something similar) will be able to power it non stop for at least 10 years without replacing it.

Obviously the user will have to regeneratively brake before using the controller and thus it should not require initial processing power to charge the ultracap

I would like to try and come up with this on my own, so if you could give me topics of research and places to find them, then please do! However, I would also love some contructive feedback and your own ideas as to how to achieve this.... I'm pretty stumped as to where to start so anything would be fully appreciated!

Thanks for your time,

James

How tiny is the motor?, 300F doesn't sound like much of an energy store?, what sort of current can they supply?.

Where does the motor get its power?

Okay, I guess I'll have to spill a little more

The motor will be a pancake 'hub' motor for a hybrid electric bike I am designing. My personal specifications are to run the motor at 24V as I am quite limited on funds and the ultra caps are pretty expensive going into the range of 1000s of Farads.

Basically to power the motor, the user will have to pedal the bike for a short distance, after which sufficient power will have been stored in the ultracap to assist with the pedal drive when the user wants, (on the uphill sections). Whilst the user is pedalling electricity will be generated from a seperate typical magnetic flux source (from the natural rotation of the wheels) which will constantly top up the ultra cap. That is the motors power source.

So whilst moving I'm only expecting the motor to be able to power the wheel for 60 seconds continuously on its own. But with regenerative braking and a seperate generation from the rear wheel, I hope to be able to extend the time to a lot more.

I can then switch around the ultracaps when I get more funding to increase the distance it can go without pedal power!

Was that any help?

Edit----
I think ultracaps can supply as much current as you want, though I am thinking that I need about 7 or 8 amps?
/edit---

James1986 said:

I think ultracaps can supply as much current as you want, though I am thinking that I need about 7 or 8 amps?

Click to expand...

I would suggest you check on that?, I was under the impression that they could only supply fairly low currents? - but check how thick the wires on them are.

I also think your estimated current requirements are FAR too low - radio controlled model cars take many times that current (100A speed controllers are quite common).

The ultra cap will be an array of 2.5v 300 farad polymer ultra caps, set up in grid formation so as to keep the 300 farad capacitance, but raise the voltage output to 25v.

Click to expand...

hi James,
I would suggest you do some research regarding connecting low voltage super caps in series/parallel.

I've worked with AVX supercaps and they are quite fine with delivering lots of amps.

Have you considered mechanical methods, such as a flywheel, as a supplementary energy store? Or pressurized vessel full of gas or liquid?

James,

With no initial back emf out of the motor or stored power..you will have to apply an extremely low resistance short across the armature coils to get braking...assuming permanent magnet field you should get drag if motor attempts to turn.

jorcioppi

RadioRon said:

Have you considered mechanical methods, such as a flywheel, as a supplementary energy store? Or pressurized vessel full of gas or liquid?

Click to expand...

I've read that they have developed a hydraulic (oil/air) hybrid storage system for trucks that has significantly better braking energy recovery than electric hybrids (something like 50% versus 30%). A hydraulic motor uses oil to compress air in a high pressure tank during braking. The energy is then sent back through the motor during acceleration.

The super cap might power the bike for only a few seconds. Why not use a battery?
Charge the battery at home overnight.

I would think the extra weight from the motor would offset any energy gain found in a batteryless regenerative power system.

Wow guys,

Thanks for all the feed back, I'll try to answer as many points as I can.

Nigel, I think that one of the main benefits of ultra caps is their ability to capture quick bursts of energy, whether in high voltage or high current, which is the main advantage over a battery alternative. Also, don't those rc cars go at much higher speeds?

Eric, the research I have done tells me that when I connect a capacitor in series the voltage adds up but the capacitance exponentially decreases, (something like 1/c1 + 1/c2 etc.) To counter this effect you need to place the capacitors in parallel. This website explains it pretty well: Capacitors in series and parallel

Ron and Cruts, your ideas are very interesting. I would be keen to learn more about the hydraulic system, but I don't know how it could be adapted to something as small as a pedal bike. The flywheel idea probably wouldnt be feasible, as I want to keep the bike as light weight as possible, and the space restrictions would be quite challenging. Also I would have to place it very low to the ground to ensure the gyroscopic effect that follows would do no harm.

Joe, thats the idea I think, but I'm not sure how to do that within the controller. I only have limited circuit board design. Do you have any tutorial websites or DIY websites that could teach me about regenerative circuits. The places I have found so far are quite vauge, and difficult for me to understand.

Audio, currently pretty much all e-bikes use batterys, which gives them the same fundamental flaw. They all have to replace their batterys after 6months - a year, depending on the quality, type, and usage. They also typically only have less than 1000 charge cycles. The great thing about the ultracaps is that they have nearly 500,000 + charge cycles with no memory or storage degredation. As I also said earlier the batteries are terrible at receiving quick bursts of energy, and will only take about 30% of whats thrown at them short term; ie when you brake. The ultracaps can take pretty much all of it! Although I may have to use a combination of both if the ultra cap route doesn't provide enough power. This, however I am trying to avoid as I want it to be as cheap as possible (cost wise) and as simple as possible.

Blueroom, the motor itself is quite lightweight. The idea is that it shouldnt be a complete takeover from the joy of riding your bike, but it should be able to help you along your way. For example going up hill. If I can manage to store roughly 2 minutes worth of motor power time generated by the users motion and braking, then the user can simply switch this on when they get to an uphil climb, and it can begin to discharge thus powering the motor. I may need to incorporate an array of dynamos on the V brakes instead for better decceleration, as the user will be constantly generating power when moving anyways. (unless I can freewheel it when its off).

Thanks for all your feedback guys, keep it coming. I really need help with circuit design though, so if you know of good websites with help on these kinds of inputs and outputs then I would be most grateful!

James

Hi Guys,

Thanks for all getting back to me!

Nigel, according to the power requirements of the motor all I need to acheive maximum efficiency at the desired rpm of between 2000-2500 is 7A. Don't the model cars require a much higher RPM than this due to their size? That is probably why they need so many amps. (I know nothing about model cars btw) And I think that one of the main features of an ultra cap that makes it so attractive is the fact that it can absorb and give out huge bursts of energy, whether its voltage or current? (provided as you say, that the wires can handle it.)

Eric, I think I probably didn't make it clear enough My bad, what I meant was that the formation should be kind of like prison bars? Capacitors in series to up the voltage, then capacitors in parallel to keep the capacitance? Not quite a grid!

RadioRon, you have me interested with your pressure vessel idea, but a flywheel will be too big and heavy for the space constraints of a pedal bike. Still, I like your out of the box thinking, keep it up! Could you perhaps explain a little bit more on how you would expect to capture the braking energy with the pressure vessel and then reuse it to accelerate the bike?

Joe, thats what I am hoping for. Do you have any reseources on typical circuit design I can look into for this purpose? Im not very well versed on circuits, so any tutorials you can point me to would be very helpful! Thanks!

Cruts, that is also an interesting idea; when you say oil do you mean petrol, as in gasoline? Or just some medium that they use for storage that happens to be oil; as this is a pedal bike, and there will be no combustion engine =(.
Still, these ideas are pretty innovative, and it would be great to adapt something like that to the common bicycle.

Audioguru, currently the market is flooded with bikes that use batteries, but the main reason I want to develop one with ultracapacitors is that batteries are terrible. Lets face it. They loose their ability to hold charge after only 1000 charges and they do all sorts of horrible leaking things, they are also (on this scale) heavy. As well as this, they only absorb something like 30% of energy given in a regen situation. An ultra cap can recharge and discharge with a much higher absorbsion rate (close to 90%) and carry on doing so with no degredation for 500,000+ charges! The beauty of the design will be that it does not power the bike completely, but assists with the difficulties of going uphill. Or thats the idea.... If it can power itself then that would be a bonus. Of course if it doesnt work then I might need to add a battery to supplement the power.

Blueroom, I would like to hope not The motor only weighs something like 4-6kgs and if a person can pedal just a little bit, then they should already get some help from the system!

Thanks once again for all your help so far. Keep the ideas coming, and also any places you have found on the net that are particularly helpful with controller design and regen circuit design, please post!

James

James1986 said:

Hi Guys,

Thanks for all getting back to me!

Nigel, according to the power requirements of the motor all I need to acheive maximum efficiency at the desired rpm of between 2000-2500 is 7A. Don't the model cars require a much higher RPM than this due to their size? That is probably why they need so many amps.

Click to expand...

Model cars are also only a tiny fraction of the size and weight, and the power required is probably much too small to power a human sized cycle.

I would suggest you try some experiments for a start, which would add to your overall marks anyway, use a battery (charged off the bike) and monitor the current under various different load conditions. Then repeat with super caps, and compare the results you get.

Under heavy load, or starting, conditions - I suspect you will be amazed at the current required.

James,

An approach might be fast recharge NiCD cells...Sinclair did it years ago with cells stuffed in the frame of an electric moped.

The simplest DC controller could be ganged DFETs connected back-to-back in a bipolar switch driven by variable duty cycle gate pulses. Look at DFET applications in power data sheets. Current sensing in both directions could control current in battery discharge direction and load from batteries in back EMF direction while braking. Throttle control could be three position switch ( spring return to neutral ) with acceleration clockwise...braking counter clockwise. Power pulse rate or braking pulse rate could ramp up slowly at suitable ramp rate.

As far as I know, supercaps with high current capabilities are still under development and are very expensive.

joecioppi

James1986 said:

Cruts, that is also an interesting idea; when you say oil do you mean petrol, as in gasoline? Or just some medium that they use for storage that happens to be oil; as this is a pedal bike, and there will be no combustion engine =(.

Click to expand...

The oil can be any liquid but it is typically hydraulic oil as used in hydraulic control systems. It's just used as an incompressable working fluid to transfer the energy between the compressed air and the motor. The air provides the energy storage since it's compressable, rather like a perfect spring.

Nice one guys, thanks!

Seems like my earlier post actually appeared after all!

Sorry to answer everyone twice

I will do as you suggest Nigel, that is a perfect way to start! Just trying to get hold of the motor at the moment, all the way from Shanghai!

Thanks for the advice Joe, I will look into it!

Cruts, thats a great idea, but i'll have to find some example mechanisms to show me how I can adapt it!

Keep those suggestions coming!

James1986 said:

Nice one guys, thanks!

Seems like my earlier post actually appeared after all!

Sorry to answer everyone twice

I will do as you suggest Nigel, that is a perfect way to start! Just trying to get hold of the motor at the moment, all the way from Shanghai!

Click to expand...

It's a pretty standard way to develop a project from scratch, do it in small pieces - and for a uni project it generates lots of nice documentation with graphs

hi

i have nothing really advantageous to tell you, was just posting to say that this sounds like a really good idea i am very interested in what yo are doing here...sounds impressive so good luck to you, plus i'm going to look into ultra caps for my final year poject although i nee them to be very very small physically lol

good luck with your project mate hope it goes well for you

Thanks Will, I will post updates as I go!

Check out maxwell technologies on google for some ultracaps, they do a huge range, from the very small, to the very large