EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: KTP on June 28, 2010, 05:49:24 pm
-
This is probably a quite silly idea, but I am wondering if it would be possible to replace the very long chain and gearing in my recumbent tadpole trike with a brushless hubmotor in the rear wheel, a smallish LiMn or LiFePO4 battery (maybe 400 watt-hour) and a fairly small brushless generator in the front pedal assembly. Since a human can only generate about 100 to 200 watts sustained, it could be a pretty small generator. The idea here is the cyclist would *see* a constant pedal force and cadence while the battery and rear hub motor would fill in the gaps on hills and during acceleration. A cyclist could even pedal while stopped and waiting for a traffic light or crosswalk.
I realize there are tons of reasons not to do this ;D including being totally screwed if the electronics fail or you don't manage the battery life, but one big issue to solve is transforming the irregular power stroke of the legs to the pedal into a continuous force at the generator, without using a heavy, bulky flywheel. I have heard it is very very hard to pedal a generator without a large flywheel because of the drag of the generator at the bottom end of the powerstroke until the other leg/pedal has turned enough to provide the next powerstroke.
So now I am trying to decide if there would be a way to control the generation of electricity such that the drag is reduced at these points and a natural cadence can be developed without this type of stuttering. Some sort of active rectification scheme?
-
Look up Hybrid Synergy Drive, which is the transmission used in the Prius. It uses two motor/generators as a CVT. One is directly coupled to the output shaft and the other goes to a planetary gearset (PSD) that goes between the output shaft and input shaft. Here's a good reference: http://www.ecrostech.com/prius/original/Understanding/Contents.htm
I actually had an idea for a hybrid bicycle using HSD. The main advantage is that operation is defined in software, allowing it to be changed very easily. Power measurement can be very accurate since power flow is managed electronically. And regeneration all the way to a standstill is possible by pedaling backwards, which spins the split motor/generator faster.
-
What a strange idea :) Anyway, it sounds like a cool thing!
I really like this field, and soon I probably will be involved in a research on regeneration-boost for motorbikes.
Let me fantasize a little.
If you plan to manage only a couple of hundred watts, motors could be small, especially if you think that usually they can be even overloaded for a short time!
I think this could be done with two motors only, without any gears. If control on the generator is done properly, you could be sure that you never (or mostly) do not give power back from the battery to the pedals, or you could think to control pedals speed such that it would become nearly constant, following average power given from your legs and supplying little pulsating torque to act like a flywheel... or even as a nonlinear flywheel, so that user could easily reach a certain optimal speed, and then produce his force at quasi-constant speed... At the same time, wheel motor could be commanded by pedals speed, and then start to brake and regenerate power when user pedals backward (as the brake command found in some bikes)...
A very complex control, really challenging, as I imagine it!
Congratulations for the idea, if you will work on it, and if I will be able to, I would be very happy to help you.
-
Obviously I'm thinking of using two complete electric drives, DC or (much better) brushless, with their position sensors or sensor-less control! ::)
-
I think a flywheel-based KERS design would be a more effective way of doing things; if you're going to add weight and complexity, might as well give you back what's usually wasted.
-
I am not familiar with a KERS design..will have to google that.
I guess most bicycles have a flywheel, it is the rear wheel whose inertia keeps it turning even when the pedal power stroke has reached an endpoint. My idea was the front brushless generator would be hooked to some black box which 3 phase rectifies the power during the power stroke, but senses when the legs are *almost* fully extended and either quits rectifying for a small time or maybe it would even very weakly energize the coils to provide a fake inertia in such a way that to the pedaler it feels like they are driving a normal flywheel. It would be a very bizarre circuit perhaps.
Edit: Hmm, now that I think more about it, I don't think it would need to provide any driving force to simulate inertia because in a normal bike the pedals are hooked to a freewheel device which does not allow them to be driven (such that you can coast without pedaling). I guess the intertia of the pedal arms and the motion of the cyclists legs is enough to carry over the powerstroke endpoint to the start of the next powerstroke. In this case the 3 phase rectifying circuit would just need to sense two endpoints of the powerstroke and taper off the current draw such that the cyclist gets a smooth motion. That seems pretty easy.
Scrat: I think it would be pretty cool to have a "digital gear" that would just be two pushbuttons on the bike handgrips. When you start pedaling from a stop you would be in the "lowest" gear and the rear motor would be energized according to how fast you pedal. You could click up and down on the pushbuttons to cause the pedal generator to provide more current to the battery which would increase the work done by your pedaling. Anyway, the idea was just to get rid of the chain and also to have a uniform cadence and aerobic exercise no matter what the terrain or how many stoplights there are.
-
Why not simply add a front hub motor ?
And why do you want to get rid of the chain ? This stuff is ridiculously cheap, reliable and efficient.
Also. Man output is just too slow for any generation if you won't be using multi-pole generator or anything else to speed it up.
-
I can't see it being more efficient than a conventional mechanical drive, look at the efficiency of each part. A hybrid car is only more efficient because of the efficiency vs speed and torque characteristics of an internal combustion engine which don't apply to human leg muscles This is not helped by the fact that larger motors and generators tend to be more efficient than smaller units. I don't see how you'll be able to get anything better than 80% efficiency and that's being optimistic.
-
I can't see it being more efficient than a conventional mechanical drive, look at the efficiency of each part. A hybrid car is only more efficient because of the efficiency vs speed and torque characteristics of an internal combustion engine which don't apply to human leg muscles This is not helped by the fact that larger motors and generators tend to be more efficient than smaller units. I don't see how you'll be able to get anything better than 80% efficiency and that's being optimistic.
The beauty of Hybrid Synergy Drive is that it is something like 75% mechanical drive and 25% electrical drive. It is actually an IVT (Infinitely Variable Transmission) since the input shaft can be turning while the output shaft is stationary (battery charge or start-off) or the output shaft can be turning while the input shaft is stationary (EV mode or normal regeneration). And while normal hybrid configurations need a minimum speed to regen, the bicycle implementation can regen all the way to zero because pedaling backwards creates the minimum speed for the split motor/generator. It would therefore be of great benefit if the payload is heavy and/or the trip is a lot of start-stop cycles.
And the idea of flattening the peaks helps a lot. Not in terms of technical efficiency, but in runtime.
-
I can't see it being more efficient than a conventional mechanical drive, look at the efficiency of each part. A hybrid car is only more efficient because of the efficiency vs speed and torque characteristics of an internal combustion engine which don't apply to human leg muscles This is not helped by the fact that larger motors and generators tend to be more efficient than smaller units. I don't see how you'll be able to get anything better than 80% efficiency and that's being optimistic.
I think that also human leg muscles have a speed range in which they are more efficient, otherwise there won't be any gears on normal bikes. However, efficiency comes from regeneration too (which is impossible to achieve using only pedals or combustion motor).
Why not simply add a front hub motor ?
And why do you want to get rid of the chain ? This stuff is ridiculously cheap, reliable and efficient.
Also. Man output is just too slow for any generation if you won't be using multi-pole generator or anything else to speed it up.
At a first look, even a single motor and normal chain for transmission would work well and be much more reliable, on the front or rear wheel, as Pyr0Beast says...
-
How much is this all going to weigh?
I can't see how this is going to be better than a chain and sprocket set which is much more efficient than a CVT.
I suppose it would be nice not to loose all the energy when you stop but surely the advantages are cancelled by the weight?
I think the best way to improve bicycle performance is to reduce the weight as much as possible and have a good fairing to reduce drag.
-
Hmm...well a bicycle chain isn't 100% efficient either, but it is pretty dang high..probably 98% or something for a normal bike....however a tadpole recumbent trike is not normal. The chain is about 3 times longer than a normal bike and also goes through plastic chain tubes and around a center delrin roller (albeit with good ball bearings). If I were to take a guess, I would say all of that reduces the efficiency to maybe 90%.
How efficient is a good multipole brushless hubmotor? I am going to guess 85% to 90% at speed....much less when accelerating.
There are other downsides to chain though....needs lubrication, which is very messy (all of my pant legs have black grease all over the lower 3rd, even though I use a light oil that is supposed to not pick up road grime). The chain can also come off in certain high torque situations.
With a regular bike, there is zero option of doing anything useful while waiting for a 3 min traffic light...also there is no ability to recapture energy from a long downhill ride...most will just be consumed as heat in the brake disks.
If you just want to smooth the aerobic exercise of a hilly ride, probably just adding a hub motor and keeping the existing drive chain would be the best plan. I was just daydreaming something a little more wacky.
-
Hi,
I'm a keen cyclist and honestly don't understand why you would want to complicate such an elegant and simple machine. I understand that the long chain is awkward to maintain and adjust but I'm not sure it can be beaten for efficiency. Recumbents have a complicated chain line so a shaft drive (which would be better) is probably out of the question. It would be great if the drive could be built into the frame somehow and a shorter chain used for the final drive.
I just think that your proposing a solution to a problem that does not exist. Please prove me wrong.
David.
P.S Hydraulic drive may be a much better idea (no disrespect intended).
http://en.wikipedia.org/wiki/Hydraulic_bicycle
-
Hi,
I'm a keen cyclist and honestly don't understand why you would want to complicate such an elegant and simple machine. I understand that the long chain is awkward to maintain and adjust but I'm not sure it can be beaten for efficiency. Recumbents have a complicated chain line so a shaft drive (which would be better) is probably out of the question. It would be great if the drive could be built into the frame somehow and a shorter chain used for the final drive.
I just think that your proposing a solution to a problem that does not exist. Please prove me wrong.
David.
If my guess is right, he idea was to smoothen the stroke-like motion of pedals on a bike, which could optimize performance. Moreover, recovering energy while braking, or simply allow to produce energy even when stopped (which can be done by electric drives) could be an amazing feature, in my opinion.
However, I agree with you that bicycle is "an elegant and simple machine", which I think will remain the same for many years yet!
-
Yes, mostly it is because I want to provide a constant pedal force of no more than say 100 watts no matter what the terrain. This is sort of like a hybrid car where the engine can operate in an idea range and the battery and electric drive assist during times where higher power is needed. (greatly simplified).
Recumbents can be very hard on the knees, and I already have pretty shoddy knees. When I am going up a fair sized hill for a moderate distance it can be very very hard unless I gear to the absolute lowest where I proceed at about 75% of walking speed. If I wanted to go at 7 or 8mph up these hills I would need to output 300 to 400 watts, which exhausts me for the next 15 minutes and makes my knees hurt for the rest of the ride. On the downside of the same hill I end up applying the brakes and not pedaling at all so that I don't exceed 20mph or so.
The brushless drive and small battery (400wh LiMn or LiFEPO4) would be about 20 pounds total added to the bike weight and I am guessing would be able to smooth out my ride. I will probably just try a hub motor first and see how it feels before I go crazy with anything advanced. There is a place that has a special going on for a 500 watt brushless hub motor laced to a 28 inch wheel for $100 and I may play with that a bit.
-
How much is this all going to weigh?
I can't see how this is going to be better than a chain and sprocket set which is much more efficient than a CVT.
I suppose it would be nice not to loose all the energy when you stop but surely the advantages are cancelled by the weight?
I think the best way to improve bicycle performance is to reduce the weight as much as possible and have a good fairing to reduce drag.
A bike doesn't need to be fast. You simply don't push pedals and run for a while on pure kinetic energy before you stop. :)
Also. Extra weight is annoying. With E-bike, I'm even slower into hills than with regular bike since the 350W motor is too weak to pull even the bike itself, leave alone my 80kg. However it does come handy on top of the hill to pick up the speed.
Hmm...well a bicycle chain isn't 100% efficient either, but it is pretty dang high..probably 98% or something for a normal bike....however a tadpole recumbent trike is not normal. The chain is about 3 times longer than a normal bike and also goes through plastic chain tubes and around a center delrin roller (albeit with good ball bearings). If I were to take a guess, I would say all of that reduces the efficiency to maybe 90%.
How efficient is a good multipole brushless hubmotor? I am going to guess 85% to 90% at speed....much less when accelerating.
There are other downsides to chain though....needs lubrication, which is very messy (all of my pant legs have black grease all over the lower 3rd, even though I use a light oil that is supposed to not pick up road grime). The chain can also come off in certain high torque situations.
With a regular bike, there is zero option of doing anything useful while waiting for a 3 min traffic light...also there is no ability to recapture energy from a long downhill ride...most will just be consumed as heat in the brake disks.
If you just want to smooth the aerobic exercise of a hilly ride, probably just adding a hub motor and keeping the existing drive chain would be the best plan. I was just daydreaming something a little more wacky.
Well, motors on bike would also require maintenance as well as batteries and electronic.
Properly set chain should never slip. Even when putting your whole strength into it, like just standing on a pedal. I wonder how will the generator pull this one out ?
With regular bike, we don't have a thing called regeneration but relaxation :)
3min at the traffic light is good enough to finally catch a breath, so you can push harder :)
Downhill. Hmm, I simply don't brake :)
Smoothing out the pedal force is done with elliptic gears fairly simple and it works.
-
A bike doesn't need to be fast. You simply don't push pedals and run for a while on pure kinetic energy before you stop. :)
Also. Extra weight is annoying. With E-bike, I'm even slower into hills than with regular bike since the 350W motor is too weak to pull even the bike itself, leave alone my 80kg. However it does come handy on top of the hill to pick up the speed.
ns.
Hmmm...is your real name Lance? No way would I be able to put out 350W sustained on a regular bike into a long hill.
I think the resistance to "improving" the bicycle stems from a "if it ain't broke, don't fix it" mentality.
I saw similar things in the archived postings from a 19th century forum about these new fangled *cars* replacing the tried and true horse...
-
A fixed gear is hard to beat for simplicity and efficiency, but it can only be optimized for a narrow range of speeds. A CVT or IVT (whether electrical or mechanical) is optimum over a wide range of speeds. HSD is just a clever way to integrate an IVT into a hybrid drivetrain.
HSD gives some other advantages, such as very accurate power measurements. (Some cyclists want to know just how much energy was used on the ride, such as for weight loss calculations.) By setting software to draw power from the split motor/generator while leaving the direct motor/generator off, the bicycle can be used as a generator. And by having the software check for a security token or similar before starting up, it can add an additional level of theft protection since the HSD will not operate unless the inverter/rectifiers are running.
HSD is not going to be a replacement for the fixed gear but there are some places where it makes sense.
-
I dont have any resistance to improving bicycles at all, as long as its real improvement and the reward is worth the effort. And there isn't any thing wrong with the "if it aint broken" mentality. First you have to figure out if it's really broken or just designed that way. Many a time I've found something that appears to be wrong on my motorcycle (looseness for instance) which later has turned out to be a design feature.
I dont own a recumbent due to lack of space but I'd love one. I'd even like to build my own but again no space. As a result I don't have direct experience of climbing hills on such a bike. This post seems to claim that sore knees are down to pedal choice, I dont know.
http://www.ransbikes.com/ITRCliplessPedals.htm
I saw similar things in the archived postings from a 19th century forum about these new fangled *cars* replacing the tried and true horse...
Where is this forum? I didn't know they had the internet in the 19th century? ;)
Finally I would like to offer my support if you do go ahead with any efforts to improve your bike and I dont want to dampen your enthusiasm. I'm just asking the same questions I'd ask myself if I was going to do the same.
David.
-
Well I think the only real way to know is to experiment with a low cost hub motor and see how it performs. I happen to have some LiMn 4.1V 4000mA Sony cells which can be discharged at 10C so the inital cost of everything would just be the $100 motor + $70 controller + a few bike bits and sprockets. I have read numerous posts that show people traveling 50 or 60 miles on a larger type lithium battery (say 1000 WH) with only light pedaling so this gives me hope for a much smaller battery assisting a person doing aerobic but not heroic pedaling for, say a 25 mile trip. 25 miles is about the limit of what I can do right now before my knees start to get sore, and that is on railroad grade bike trails. There are some round trip 50 mile trails I would *really* enjoy being able to complete in one outing though.
-
Hmm...well a bicycle chain isn't 100% efficient either, but it is pretty dang high..probably 98% or something for a normal bike....however a tadpole recumbent trike is not normal. The chain is about 3 times longer than a normal bike and also goes through plastic chain tubes and around a center delrin roller (albeit with good ball bearings). If I were to take a guess, I would say all of that reduces the efficiency to maybe 90%.
How efficient is a good multipole brushless hubmotor? I am going to guess 85% to 90% at speed....much less when accelerating.
The generator will have a similar efficiency, about 85%, then the electronics used to rectify and boost/drop the voltage won't be 100% efficient, maybe 85% to 95% depending on the operating conditions, which makes it inferior to a recumbent bike transmission.
There are other downsides to chain though....needs lubrication, which is very messy (all of my pant legs have black grease all over the lower 3rd, even though I use a light oil that is supposed to not pick up road grime). The chain can also come off in certain high torque situations.
What's cheaper a oiling and buying a new chain every few thousand miles or a new Li-ion battery?
With a regular bike, there is zero option of doing anything useful while waiting for a 3 min traffic light...also there is no ability to recapture energy from a long downhill ride...most will just be consumed as heat in the brake disks.
Those are the only advantages which I think will be outweighed by the huge number of disadvantages. Also, it's good to have a break every now and then, suppose you're Mr Joe average who can output 100W continuously but 300W peak. If you rest for five minutes you'll recover and be able to output nearly your full peak power when accelerating but if you're still tired from outputting 100W, the extra 200W from the motor won't gain you much, especially when you balance it against the extra weight and that the boost won't last for long.
-
The brushless drive and small battery (400wh LiMn or LiFEPO4) would be about 20 pounds total added to the bike weight and I am guessing would be able to smooth out my ride. I will probably just try a hub motor first and see how it feels before I go crazy with anything advanced. There is a place that has a special going on for a 500 watt brushless hub motor laced to a 28 inch wheel for $100 and I may play with that a bit.
Very interesting to find that object at only $100! It could be interesting also for many other things... Let us know what you will do when you play with it!
I expect (looking at the price) those are brushless DC motors, with integrated electronics for drive, so the first thing that comes to my mind is if they could be reversed (used as generators) without smashing them out...
-
The inverter would probably have a signal to turn it into a rectifier for regeneration. Or maybe it's one of the ones that doesn't which is why it is so cheap...
-
Perhaps this isn't such a silly idea as I first thought.
If a traditional chain and pedal transmission were used along with a battery and a regenerative break, it might actually be worth it.
I think a recumbent bike is probably the best design for this, especially for speed, the world record for a human powered vehicle was broken on a recumbent - 82.33mph (132.47kmph).
http://en.wikipedia.org/wiki/Recumbent_bicycle
-
Depending on your local terrain, I would guess that most braking is only over 1-5 seconds, so a supercapacitor array would be a better idea, or at least as a supplement to the battery. Supercaps can charge and discharge at much higher rates without losing efficiency or generating internal pressure and temperature problems.
-
With current technology, capacitors are simply not energy dense enough to be used as primary energy storage for traction motors. Here's a discussion about using an array of capacitors in a hybrid car:
http://www.diyaudio.com/forums/parts/148221-capacitor-array-starting-engine-2.html
Based on quick calculations, 200A*12v*1s=2400J
That amount of energy is attainable with just 6 4700uF, 450v capacitors charged to 430v and discharged to 100v. And they'll very likely outlast the car since capacitors do not degrade much when used within specifications.
However, traditional starters are very inefficient (high resistive losses), so the actual energy needed is lower. High voltages reduce resistive losses. Add some more tricks like unload solenoids to deactivate some of the cylinders and electric oil pumps to prime the bearings and the energy needed is even lower.
Where capacitors really excel is regenerative braking. The problem is the sheer size of the capacitor array needed. A 3000lb car going at 45MPH would need almost 3F of 450v capacitors to store all of the energy from regenerative braking. (As a joke, one of my friends would say that's a "Peggy-sized array"!) So the capacitive engine starter is already practical, but not the capacitive regenerative brakes.
-
Hmmm...is your real name Lance? No way would I be able to put out 350W sustained on a regular bike into a long hill.
I think the resistance to "improving" the bicycle stems from a "if it ain't broke, don't fix it" mentality.
I saw similar things in the archived postings from a 19th century forum about these new fangled *cars* replacing the tried and true horse...
Neither does a hub motor :)
We have quite steep hills however, they aren't long, but as I've said, it is harder to do it with 'help' than without it.
Well, it ain't broke, but it will be when you are trying to fix it :)
I can definitely pull out about 200W for about 5 minutes. Even 1kW (measured (actually it was 999W because meter couldn't go higher)) for a few seconds. But you have to recover for a few minutes after you reach the top (That is cycling, not sitting down).
I however do get out of breath at hills that are steep and long, don't really have the capacity to do it even in the slowest gear.
I have thought of building a device to measure biking power. It is a simple as hell, it measures the number of chain rings per seconds and tension on the chain.
However, best motor would be chain assisted and not hub. Hub motors have no power at low speeds.
IF you are searching for good and cheap batteries.
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=9174&Product_Name=Turnigy_5000mAh_5S_20C_Lipo_Pack (http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=9174&Product_Name=Turnigy_5000mAh_5S_20C_Lipo_Pack)
I'm thinking about replacing my lead ones with this (Cyclon 36V 5Ah)
18V and 5Ah (90Wh) for less than 50 bucks is awesome :)
With current technology, capacitors are simply not energy dense enough to be used as primary energy storage for traction motors. Here's a discussion about using an array of capacitors in a hybrid car:
http://www.diyaudio.com/forums/parts/148221-capacitor-array-starting-engine-2.html
Yes. And they cost more per Wh.
And you have two useless systems instead of one.
Why don't you build a mechanical regeneration ? ;)
-
I think super capacitors have too poorer energy density to store a significant amount of power.
How larger capacitor do you need?
What's cost effective?
Suppose you go for a 50V 50F capacitor and the DC-DC converter works down to 10V, E = 0.5C(V12-V22) = 0.5*50(502-102) = 60kJ which is equivalent to a 12V 1.39Ah battery.
You just need a batter which is capable or absorbing the energy released going down a steep hill for a few minutes.
-
Why don't you build a mechanical regeneration ? ;)
Flywheels and pressure tanks don't have the energy density of a modern battery, although they can easily do better for peak power. The flywheel would need some sort of CVT or IVT, while the pressure tank would just need some sort of configurable pump/engine. An electrical hybrid just sounds like a better choice for now.
However, best motor would be chain assisted and not hub. Hub motors have no power at low speeds.
http://endless-sphere.com/forums/viewtopic.php?f=6&t=10301&start=0
It has enough torque to tow a bus and the race demonstrates the top speed quite well. (And note that it raced in EV mode, which is why it's called electric when it technically is a hybrid.)
-
I would really like to see the effect of a high speed flywheel on a lightweight bike as you try to go around a corner! Good luck steering. :D
The reason I suggest supercaps (for the regeneration aspect) is that most braking done in commuter journies is fairly short and sharp. Traffic lights, kerbs, junctions etc. Where you need to go 20mph to 0mph, and then back to 20mph in maybe 30 seconds (that is 2 cycles of approx 500W for 15 seconds). That would be a huge challenge with batteries, and the power only needs to be stored for a short period of time.
You know how a circuit transfers energy most efficiently when the load impedance matches the source impedance? Well it is the same here. Batteries have a relatively high ESR, but to get maximum torque you need a motor which is very low resistance. Think of it as putting capacitors on a power supply, to smooth out all those nasty load spikes. This makes the source look much "stiffer" and robust to the load. Incidentally, remember the EEVBlog on calculator design, and how the HP20B calculator makes inefficient use of the batteries? Appropriate caps in the power supply there could hugely extend battery life, without knocking back the clock speed.
It isn't easy to find decent size supercaps, but ones I have seen that are in development for cars could easily hold 7500J (150kg 0-36kmh) without presenting huge packaging problems on a bike. One of these alone stores nearly 11kJ: They measure only 138 x 61mm http://www.maxwell.com/ultracapacitors/products/large-cell/bcap0650.asp
They aren't cheap right now (about £100 GBP for that one), but you can bet that prices will drop very rapidly indeed, and density will go up at the same time. There is big money out there investing in this sort of technology.
http://uk.mouser.com/ProductDetail/Maxwell-Technologies/BCAP3000-P270-K05/?qs=sGAEpiMZZMuos25lhNoLamJr2MSVxS1C
-
Note that the ESR of supercaps is pretty bad too, and they dislike high current draw, though possibly still better than batteries.
-
Mixed supercap-battery energy storage systems are the best you can do at the moment.
This is because it doesn't exists a single device which has the highest power density and the highest energy density: batteries have the highest energy density, but they have lower power density than supercaps, (since these last can manage very very high currents without getting damaged). By combining (with some intelligently controlled converters in between) it is possible to exploit the full power+energy capabilities of both device types.
Take a look at this figure, which also treats flywheels:
http://physics.technion.ac.il/~rutman/jeremy%27s%20seminar_files/powerdensityvsenergydensity.jpg
(flywheels need mass to store energy, while supercaps theoretically don't, so we can expect their density will improve and, more important, they don't oppose to you when trying to steer your bike :D )
-
Note that the ESR of supercaps is pretty bad too, and they dislike high current draw, though possibly still better than batteries.
From what I understand it's a trade off between leakage current and ESR. A memory backup super capacitor will have a very high ESR and a low leakage current, a large super capacitor for pulse power will have a very low ESR and a high leakage current.
-
I wasn't aware that there are supercaps designed for high-power applications, I assumed people were abusing the supercaps designed for memory backup. Self discharge is obviously not an issue at all if you use it for short-term energy storage, as long as it doesn't lose a significant part of the charge within minutes, it should be fine, unless you're in mountainous terrain with long descents.
-
But with modern LiFePO4 batteries able to handle very high charge and discharge rates, is there a point in using capacitors?
-
It seems that those batteries (LiFePO4) have power density near to supercaps, but they still have shorter lifetime (2000 vs 1 million cycles, just looking at one random Maxwell's datasheet and wiki for LiFePO4 batteries).
So supercaps could still be much better peaks management (e.g. in KERS systems).
-
For large peaks, what about hydraulics?
-
I don't know much about mechanical energy storage devices, but my first thought is that there are no faster systems than electric ones. Isn't it so? For managing of the quick charge/discharge required by a regenerative hybrid system for vehicles, I think electric storage is the best thing. The presence of hybrid cars on the market and KERS in competitions confirms me this opinion...
-
For large peaks, what about hydraulics?
How can hydraulics store power?
Liquids aren't very compressible. The only way I can imagine is to use a piston with a spring but I think it's more sensible to use a mechanical spring.
-
How can hydraulics store power?
Liquids aren't very compressible. The only way I can imagine is to use a piston with a spring but I think it's more sensible to use a mechanical spring.
http://en.wikipedia.org/wiki/Hydraulic_hybrid
-
That must be the accumulator I was describing.
The only way I can imagine is to use a piston with a spring but I think it's more sensible to use a mechanical spring.
http://en.wikipedia.org/wiki/Hydraulic_accumulator
Still, looks pretty bulky.
-
Li batteries work well for regeneration.
At 20C or more of discharge and charge, that isn't really an issue.
And they hold steady voltage unlike supercaps.
-
Li batteries work well for regeneration.
At 20C or more of discharge and charge, that isn't really an issue.
And they hold steady voltage unlike supercaps.
Of course it's a trade-off between power, energy, weight or volume and... system complexity. It is possible for a simple battery charger to be enough, but this isn't so easy to determinate. I think that this can be a discussion with no end if we don't try to list some hypothetical specifications.
Supercaps can hold very much higher currents, and I know for sure they are used in systems like hybrid vehicles (together with batteries, generally), and they are the preferential choice when a KERS-like (short time braking, short time torque boost) behaviour is required. Since battery's life is shorter (in number of cycles and time), the two have to be combined to achieve the best performances and reliability/duration (this is done even in small energy scavengers, at least for research). Supercaps' charge law (as for normal caps, charge=C*Voltage) and their strict voltage limit (they don't tolerate overvoltages) impose the use of an intelligent controller and a couple of DC-DC converters, making the whole system quite complex.
-
Li batteries work well for regeneration.
At 20C or more of discharge and charge, that isn't really an issue.
And they hold steady voltage unlike supercaps.
Of course it's a trade-off between power, energy, weight or volume and... system complexity. It is possible for a simple battery charger to be enough, but this isn't so easy to determinate. I think that this can be a discussion with no end if we don't try to list some hypothetical specifications.
Supercaps can hold very much higher currents, and I know for sure they are used in systems like hybrid vehicles (together with batteries, generally), and they are the preferential choice when a KERS-like (short time braking, short time torque boost) behaviour is required. Since battery's life is shorter (in number of cycles and time), the two have to be combined to achieve the best performances and reliability/duration (this is done even in small energy scavengers, at least for research). Supercaps' charge law (as for normal caps, charge=C*Voltage) and their strict voltage limit (they don't tolerate overvoltages) impose the use of an intelligent controller and a couple of DC-DC converters, making the whole system quite complex.
A tradeoff that is not worth it. If a battery can power that 1kW bike motor it can also suck power from a 0.5kW generator or you blow excess like heat.
Supercaps will die with high current and they aren't much cheaper than batteries anyway.
Cycles don't matter. One incomplete charge and discharge does not equal a full cycle.
Li batteries can do 300 FULL cycles a year and their capacity drops nearly the same as if they haven't any. They die regardless of their usage.