cap on battery, influences charging?

[ProDigit]

It's been mentioned before that placing a large capacitor in parallel with the battery is a good idea, and I tend to agree!
However, does it influence battery charging?

I presume not, because the BMS is connected straight to the cells, and the cap is outside of that connection, but I could be wrong.

For a bike like the Xtreme XB-700Li, I presume the BMS is on the battery, and the charger is just a voltage controller (like a digital transfo)?


Thanks!

 

[ZOMGVTEK]

I don't see any real advantage to putting a capacitor on your bike battery. You want to try and get it as close to the load as possible, in or on the controller. Low ESR/L is key.

Regardless, it shouldn't make any difference at all in how the battery charges.

 

[neptronix]

It is like putting a tiny ( couple milliamp hours, maybe? ) battery in parallel with your battery.

What's the idea? if your battery is being stressed for a couple milliseconds, then you need a bigger battery, a battery with a higher C rating, or a lower load.

 

[ProDigit]

I was thinking in lines of a $60 5Farad cap, they use in car audio to provide the sub woofer amp with enough power not to overburden the fuses; not a regular capacitor.
a 2-4Farad cap should have enough energy to provide power for 2 seconds the least; but combined with the battery, perhaps have enough power to saturate the motor for the whole acceleration phase.

It could be nice for regenerative breaking, though my bike does not support that (yet).

 

[ProDigit]

You want to try and get it as close to the load as possible, in or on the controller. Low ESR/L is key.

Would you mind explaining what you mean here?

 

[neptronix]

http://www.physicsforums.com/showthread.php?t=256751

4 farads will buffer about a nanosecond-millisecond worth of your amp draw. Farads are more like amp-seconds rather than amp-hours.

I'm not seeing how this would benefit you at all. Why wouldn't you just add more amp hours of lithium battery? it'd be cheaper, smaller, lighter, and better in every way.

I think you need to get a solid idea of how microscopic the amount of energy these capacitors contain per volume and weight.

http://en.wikipedia.org/wiki/Electric_double-layer_capacitor#Specific_energy

The best case scenario is that you buy a capacitor that has high power output, but the energy density and volume of a half dead lead acid battery. Except you pay many multiples of what a lithium battery watt hour would actually cost you for the privilege.

 

[amberwolf]

There is a recent thread (last couple weeks or so) discussing the math and reasons why it's probably pointless to stick caps on a battery. Sorry I don't have a link.

 

[ProDigit]

Interesting!
Thanks for pointing that out; although many 3 to 4 farad caps are used in audio systems, where operation near to e-bike power levels are used (500W RMS amps and the likes). Under such circumstances, the caps provide sufficient energy to drive the peaks; but perhaps not as much as I had hoped (eg: provide power to the motor for the 10 seconds of acceleration time).
although they are not the only ones driving the motor,the battery is still connected in parallel with it, so it's just there to compensate the voltage difference/voltage drop of before and after load.

If I may turn topic of the thread around;
If a capacitor can not be used to help initial acceleration process due to having too low capacity, then my question would be the following:

When I drive my bike, I can hear a tone coming from my motor. That tone is a chopped voltage by the controller.
So in essence the batteries are not providing a continuous flow of energy, but a chopped one to the bike.
I wonder if it will affect the lifetime of the batteries (providing several hundred pulses of voltage per second, versus providing more of a constant flow).

A 2 Farad capacitor would contain more than enough energy to level out the pulses created by the controller/motor to the battery.
Would it make sense for this reason to install one?

 

[ZOMGVTEK]

It's nice you are observing the bikes operation and determining ways to best optimize it. However, you would need a VERY large capacitor to have any noticeable impact in the sense you are regarding. Car audio capacitors tend to be rated for 15V or so, and it's not a good idea to run a capacitor, especially a big cheap one, at its limits. When they blow, its a real mess. It's also worth noting that the VAST majority of car audio capacitors do not nearly meet the ratings. The 1F models tend to be the closest, but most 3F+ models can be something like 0.5-2F in real life, under ideal loading. It's not really the energy they store that can help a system, its how quickly it can dump the power. A battery and lots of wire typically can't deliver power EXTREMELY fast, due to things like inductance and resistance. In most e-bikes, you have substantially less length of wire, and you are often running closer to continuous power, rather than gigantic spikes like in a car audio system.

What battery are you using? 500W should be very easy to supply, unless you are using a very low capacity pack made out of poor cells with a poor BMS in the way.

energy = 1/2 * c * v^2

Assuming you purchase a 3F capacitor that actually measures at 3.0F...

You would need 2 of those car audio caps to work in a '24V' system, with peaks up to about 28-29V safely. If you have a 36V bike, you would need 3 in series.
Let's assume you have a 36V battery charged to 40V.

energy = 1/2 * 3.0 * 40^2
Thats 2,400 watt-seconds, or 2/3 watt-hours. That will run a 500W load for about 4 seconds. That's not all that bad, until you realize the size, weight, and cost of this 36V 3.0F capacitor is going to be about the same as 1kWh of LiPo that can run the same motor, at the same 500W, for nearly 2 hours.

The 'pulses' you speak of, are how the controller regulates speed. It has fairly poor rotor positioning off the hall sensors, so its going to be a bit crude at those super low few RPM speeds.

 

[liveforphysics]

The battery itself has a ton of capacitance inherently (a bunch of high surface area foil in layers close to each other).

If you installed something like 4 x 10F car audio caps in series to make a 2.5F 48v cap for your battery, you would be carrying around 20-30lbs and more bulk and volume in exchange for 2.8kJ of energy. That's 0.8watt-hours, or about 1/2 the energy that adding a single AA cell to your pack would provide.

If you wanted to filter controller induced voltage ripple at the battery for some reason (the controller is where you want to filter it, the battery doesn't care), you would want electrolytic caps, and perhaps 0.01F would be more than enough as long as they were low ESR.

The double-layer carbon caps used in those car audio caps to get huge capacity numbers typically have uselessly high ESR to help filter much of anything in the car audio or ebike application. I remember the last "6F" car audio cap I bought on ebay turned out to have an ESR of 350mOhm... In other words, completely useless for stabilizing anything, for every 10amps of ripple current, it's going to have a voltage ripple of 0.35v on it's own...

 

[ProDigit]

Thanks for your responses!
Indeed, I forgot that those capacitors are for 12V systems mainly (although one can always over-volt a capacitor somewhat).
The output of a car amp usually is in the 60V range, not the input, which is 12V; so I gave up on the idea for a capacitor (since my bike runs at 48V).

And using 2 (or 4) in series does not make them compatible with 4x the voltage.

Best is to use a battery then...
What kind of battery would you suggest?(one that can provide the extra juice at acceleration; is not too big or expensive, lightweight, and can be put in parallel with the internal battery (the battery of the bike is LiPo4 48V, 20Ah) (about the size of a large capacitor, or small controller would be nice, so I can install it underneath the saddle).
What I would need is a high current battery, at 48V; but also very affordable!
I suppose I could plug in a 48V10Ah LiFePo4 battery, but they cost $+300, great for range extending too, but too expensive for me!

It'd be nice if both batteries would have their own BMS,and I'd be able to charge both from a single charger. You think it's possible?

I presume when using a small battery,just for the sake of increasing amps to the motor at acceleration, that a diode bridge is needed?

If I'd have to install a battery simply as a range extender, I could use a fuse system as currently is used in my bike; disable one fuse, enable the second, and both batteries would be used separately, but I wanted to look into something that could boost my initial acceleration a bit at the same time (which I thought a capacitor would do, but seemingly I was mistaken there)!

Thank!

 

[liveforphysics]

Thanks for your responses!
Indeed, I forgot that those capacitors are for 12V systems mainly (although one can always over-volt a capacitor somewhat).

Ultracaps are electrochemical devices, the solvent disassociates and off-gasses with potentially explosive results in an over-voltage situation, they do not have the potential for margin that some electrolytics have (though in practice you should never design to even run at 80% of any caps maximum voltage rating).

And using 2 (or 4) in series does not make them compatible with 4x the voltage.

If you put them in series, it does add the combined voltages, so if you put 4x 12v caps in series, it does work as a 48v cap. Just remember your capacity would be 1/4th the individual capacity of a single cap. (but the same energy stored as the sum of the 4 individual caps).

Best is to use a battery then...
What kind of battery would you suggest?(one that can provide the extra juice at acceleration; is not too big or expensive, lightweight, and can be put in parallel with the internal battery (the battery of the bike is LiPo4 48V, 20Ah) (about the size of a large capacitor, or small controller would be nice, so I can install it underneath the saddle).

We have many threads on various battery choices with lots of options and advantages and disadvantages, and thousands of real-life-experiences feed back from people using them for ebikes. Bringing it up in this pointless thread full of incorrect BS seems like a waste of everyones time.

I presume when using a small battery,just for the sake of increasing amps to the motor at acceleration, that a diode bridge is needed?

Thank!

WTF? No. I recommend reading some newb guide threads to learn a few basic concepts before continuing to post this drivel.

 

[etriker]

Here is some info on using ultracaps on ebikes.

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20010069673_2001112722.pdf

They say they help but on page 16, 80lbs total for the caps ?

 

[neptronix]

The reason that capacitors are useful in a high power audio system is that the peaks of power are usually quite short ( like the sound of a bass drum hitting for a 1/4th-1/10th of a second, generating a spike of power that is a few hundred or thousand watts for very short period of time ).

When you are accelerating from a dead stop, you are using a pretty massive amount of power for many seconds. Depending on your setup, you would require thousands of farads to buffer that power.

It would be radically cheaper, smaller, and lighter to just use some extra amp hours of battery to achieve the effect of lowering the voltage sag you're trying to combat. Or using a battery with such high output that voltage sag isn't an issue at all.

Like when i hit my 20AH 36V 20C battery with a 42 amp load, my voltage drops from about 41v fully charged to 40v. I could double my battery capacity, hit it with the same load, and expect my voltage to only drop about 0.5v.

How bad is the voltage drop you're trying to combat?

 

[Punx0r]

Even in the car audio world those of a scientific disposition (as opposed to a liking for bling and snakeoil) regard power caps as pointless/useless.

 

[ProDigit]

How bad is the voltage drop you're trying to combat?

When the main battery is driven ~30 minutes, I'm talking about 25% voltage drop. Right now, 50% voltage drop at acceleration, because the battery is nearly empty.

*edit: Correction on the voltage drop: I'm mainly reading from the supplied fuel gauge, I haven't actually measured the voltage. When fully charged, the meter remains at 100%, acceleration or not. When driving ~30-60 minutes, the meter remains at 100%, but drops to 75% at acceleration. When driven few hours, (like right now), the meter remains at 75%, meter drops to 50% when accelerating. So basically a battery that would just add that 25%? (probably a 5Ah battery could do).

To liveforphysics: Why do you advocate against using power diodes? For the 3V voltage drop, or for other reasons?

 

[neptronix]

Even in the car audio world those of a scientific disposition (as opposed to a liking for bling and snakeoil) regard power caps as pointless/useless.

Oh good! glad to hear there are some actually intelligent people in the car audio world.
Those are the guys who have a giant array of SLA batteries in the back of their car, right?

Prodigit: you need to measure the voltage drop you are getting and stop looking at a meter that isn't really giving you any useful data. A simple multimeter hooked up to the battery will do. Something like a GT watt meter, turnigy watt meter, or cycle analyst would be even better.

If it turns out that you have an awfully saggy battery, then all a capacitor will do for you is make it not saggy for about a millisecond or so.

 

[Punx0r]

Those are the guys who have a giant array of SLA batteries in the back of their car, right?

I imagine so. I don't know if anyone has used Lipo yet

 

[ProDigit]

Even in the car audio world those of a scientific disposition (as opposed to a liking for bling and snakeoil) regard power caps as pointless/useless.

Oh good! glad to hear there are some actually intelligent people in the car audio world.
Those are the guys who have a giant array of SLA batteries in the back of their car, right?

Prodigit: you need to measure the voltage drop you are getting and stop looking at a meter that isn't really giving you any useful data. A simple multimeter hooked up to the battery will do. Something like a GT watt meter, turnigy watt meter, or cycle analyst would be even better.

If it turns out that you have an awfully saggy battery, then all a capacitor will do for you is make it not saggy for about a millisecond or so.

Caps are very useful in car audio world for a few reasons.

Probably the voltage meter on my bike is made in such a way that at the cutoff voltage the gauge will be in red (that is 10%), and at fully charged voltage, the meter is at 100%. If that be the case, 55V at fully charged, 32 at empty, when the meter shows 50-75% that would mean that the battery is somewhere between 50 and 43V.

I'm already determined to use a battery in parallel with my current battery, just to provide some extra juice. I need a lifepo4 that's preferably around $200 or less (shipping included). Most batteries I find of 10Ah are sold for over $300, so if I could find a 5Ah for ~ half the price I'd be happy!

I just don't get why some people act like they know everything, and come with conflicting information (as to why not use diodes on both batteries?), without explaining why.

 

[etriker]

I just don't get why some people act like they know everything, and come with conflicting information (as to why not use diodes on both batteries?), without explaining why.[/quote]

Right ? Not very scientific.

Diodes can keep you out of trouble when doing parallel battery testing.

Get some of these. One on each battery. That will tell you what is going on.


http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=10080

 

[neptronix]

I'm already determined to use a battery in parallel with my current battery, just to provide some extra juice. I need a lifepo4 that's preferably around $200 or less (shipping included). Most batteries I find of 10Ah are sold for over $300, so if I could find a 5Ah for ~ half the price I'd be happy!

I just don't get why some people act like they know everything, and come with conflicting information (as to why not use diodes on both batteries?), without explaining why.

I'm glad you got it

I for one can't give you a solid answer on the diodes part. It may have to do with the complication of running two BMSed packs together. You will not see a bunch of diodes, fuses, capacitors, or anything else in the non-BMS RC Lipo and A123 packs that people are building on here.

When people can't seem to agree with something on the forum here, the answer is usually in the middle. Or has just not been properly tested.

Oh and punxor, a RC Lipo powered sound system sound be badass. Imagine how much power just a few 4S / 5 amp hour nanotech packs in parallel could put out. I've thought about building some portable noise weapons using some, for various nefarious reasons . Some of the higher end nanotech could put out a 2-3kW constant per 5amp pack.

 

[etriker]

I'm already determined to use a battery in parallel with my current battery, just to provide some extra juice. I need a lifepo4 that's preferably around $200 or less (shipping included). Most batteries I find of 10Ah are sold for over $300, so if I could find a 5Ah for ~ half the price I'd be happy!

I just don't get why some people act like they know everything, and come with conflicting information (as to why not use diodes on both batteries?), without explaining why.

I'm glad you got it

I for one can't give you a solid answer on the diodes part. It may have to do with the complication of running two BMSed packs together. You will not see a bunch of diodes, fuses, capacitors, or anything else in the non-BMS RC Lipo and A123 packs that people are building on here.

When people can't seem to agree with something on the forum here, the answer is usually in the middle. Or has just not been properly tested.

Oh and punxor, a RC Lipo powered sound system sound be badass. Imagine how much power just a few 4S / 5 amp hour nanotech packs in parallel could put out. I've thought about building some portable noise weapons using some, for various nefarious reasons . Some of the higher end nanotech could put out a 2-3kW constant per 5amp pack.

All the threads about rc lipo hooking it up all kinds of ways and charging them all kinds of ways. We are not afraid of that, right ?

Lets not be afraid of a few diodes or fuses in our electric bike power supplies !

Or new and different ideas.

He is talking about using lifepo4 and diodes. That right there says he is not a newbee !

Newbees ask rc lipo questions !

 

[ProDigit]

heh, thanks!

I've bought 4 of these diodes: http://www.amazon.com/gp/product/B005OFLMQW
I plan on using 2 of them in parallel to each positive terminal of each battery (negative terminal connected to chassis).
mounted on an aluminum bar/ruler to draw out any heat, just in case.

That'll be 40A; I hope it'll be enough. Most likely it will, because the second battery will aid the primary battery. I'm quite sure I won't surpass the 35A peak on the battery (or 17,5A per diode).

I know there's a guy making the ideal diode; a bit expensive, but great for ebikes!

I told him if he wanted no voltage drop over the diode, that he should use a triode instead, and solder the gate on the input. I don't know if that will work, but if not, just solder a resistor to the gate, and then connect input and gate together. That way there's virtually no voltage drop.

I wonder why 2 BMS packs won't be able to operate together?

 

[Ricky_nz]

I know there's a guy making the ideal diode; a bit expensive, but great for ebikes!

I told him if he wanted no voltage drop over the diode, that he should use a triode

I highly doubt you would want to use one of these http://en.wikipedia.org/wiki/Triode.

I'm guessing you mean A FET and no tieing the gate to the "input" will not work at ebike pack voltages.

 

[texaspyro]

And you won't find a single 4F car audio cap that is anywhere close to 4F... I've tested dozens of brands. Virtually all are fraudulently rated crapassators (like less than 1/10 their advertised rating).


But caps can be used constructively. The latest "hybrid" race cars use them for regenerative braking... but why is beyond me.