Parallel to Series: "Ultra Capacitor" Turbo?

[safe]

Okay, here's a wild idea...

You take a bunch of parallel connections from your battery and feed that voltage (let's just say it's 12 Volts) to a row of 10 ultra capacitors at a rate that is "conservative" for your batteries drain rate. So you make sure that if your battery chemistry can only handle 1C that you drain them into the capacitors at no more than this rate.

Okay, so now you have a bunch of "charged" capacitors.

Now what you do is press the "Turbo" button and all these capacitors get linked together not in parallel, but in series. How do you do this? I don't know, maybe with relays, but the idea is that you charge at one voltage and then discharge at the combined voltage.

All of a sudden your normally "tame" motor that is normally getting 12 Volts as input all of a sudden has 120 Volts that will for a short period of time produce a "rocket". Once the capacitors are spent then you go back to your normal voltage and in the "background" the capacitors are being recharged for the next time you use the "turbo".

You could even use the "regen" to fill the capacitors along with the battery. But in a lot of ways having a freewheel is better because you could get this huge burst of accelleration and then just coast for a long way. In this way you could have for very short periods of time an "extra taller gear" that would not ALWAYS need to be set up this way. The normal riding would be at the low voltage and only the "turbo" would be at the added voltage. A factor of ten might be too much, but you get the idea of how far you could take this if you wanted to.

The main "idea" is the "parallel to series" connections that boost the voltage to obscene levels for a short period of time.

So what's going to blow up first with this?

You might want to "go around" the controller on this one. Somehow bypass the controller and just get a straight shot from the capacitors directly into the motor until the energy runs out...

 

[xyster]

So what's going to blow up first with this?

Your shoulder sockets trying to hang on!

Another of Safe's ...err... creative ideas. I don't know enough about capacitors to offer any other response. But maybe I will after the electrical gurus jump in.

 

[Leeps]

I threw that idea around a little bit on the old forum. The problem ended up being that two one farad capacitors in series ends up being a one half farad capacitor with twice the voltage capacity. When you put a bunch of the ultra capacitors in series you lose the ultraness, or you add a bunch of them in parallel. This can get big and expensive quickly.
I still like the idea it just turned out to be not as effective as i thought it would. 12 volts to 120 volts is well insane, i would go for double voltage, even if you hit the button at full speed you would end up with stall torque coming from the motor. The parallel to series thing seems complicated, i would use a charge pump or a boost converter to slowly charge the capacitor bank at a controlled rate.
Joe

 

[xyster]

It does strike me you could use small, high power batteries, like a123's, in lieu of capacitors.

 

[safe]

The problem ended up being that two one farad capacitors in series ends up being a one half farad capacitor with twice the voltage capacity.

So you're saying you lose energy in "total", but you do get the doubled voltage right? I'm not expecting this to be without some "losses" (the efficiency might not be so great) but the idea is to give a huge "turbo" boost every once and awhile. You might even have a little LED light that tells you when the charging of the capacitors is complete.

And you do get the idea of:

"Charge in Parallel"

"Discharge in Series"

Right?

 

[Leeps]

I suppose you could, the nice thing with capacitors is they dont have a cycle life and they dont have a charge profile. The bad side is that the capacitance drops as you put them in series, and that they charge/discharge linearly. I wonder what the ESR on those things are.
Joe

 

[safe]

The bad side is that the capacitance drops as you put them in series.

In charging mode or discharging mode?

The idea is to charge in parallel and discharge in series through relays switching the connections.

 

[xyster]

I wonder what the ESR on those things are.

What's 'ESR'?

The a123's *should* be good for at least 2000 cycles from 100% DoD.
Each cell is 2.2 ah and 3.6 v nominal, 60C intermittent discharge, with 5-minute recharge time (if you can feed them that kind of power).

 

[safe]

The a123's *should* be good for at least 2000 cycles from 100% DoD.
Each cell is 2.2 ah and 3.6 v nominal, 60C discharge, with 5-minute recharge time if you feed them that kind of power.

Bingo!

That sounds like a good idea if capacitors have some "weirdness" that makes this "charge parallel" and "discharge series" concept hard to accomplish.

 

[safe]

2.2 ah
3.6 v nominal,
60C discharge

So let's say you have 20 of them:

20 * 3.6V = 72 Volts

Maximum amps:

2.2ah * 60C = 132 amps peak

That would be a real "burst" on a machine used to 36 Volts normally...

 

[xyster]

That sounds like a good idea if capacitors have some "weirdness" that makes this "charge parallel" and "discharge series" concept hard to accomplish.

I don't know, but it seems like the a123 "turbo" boost pack may have to be married to a low-impedance main pack that has a high discharge rate, like SLA, lithium manganese, lithium iron phosphate, or high rate lipoly in order to recharge the a123's fast enough for use every 5-10 minutes.

 

[safe]

I think the idea is to allow battery chemistries that can't handle high discharge rates to be able to on occasion "perform" at high discharge rates through the intermediate ultra capacitors. (or these A123 cells)

On a long uphill this doesn't help you any, but from a "fun" perspective of wanting to on occasion get a "thrill" it might make sense. And if the capacitor solution could work you could be charging while you are braking or even at times when you are at less than full power.

By definition:

"The capacitors CHARGE whenever you aren't using the full discharge of the battery." (it soaks up the unused potential of the battery)

It sort of "runs in the background" getting ready for it's use...

This would be an easy way to make it into the 50 mph Club!

 

[Leeps]

ESR stands for equivalent series resistance, or simply the resistance of the capacitor.

The capacitance will be divided by the number of capcitors (assuming there all the same capacitance if not theres a formula to find the capacitance) when they are mounted in series. This will manifest itself as the voltage dropping quicker than expected. For example you have two imaginary capacitors you charge each one with one amp for ten seconds to arrive at 10 volts, this gives you a one volt per amp per second change (or one farad). Now you hook them up in series pull one amp and the whole thing is discharged in ten seconds. This time you dropped twenty volts in the same time with the same current, this is a two volt per amp per second change(or one half of a fara).
Joe

 

[xyster]

Seems to me you wouldn't have to switch from parallel to series with the a123's themselves. Say you have a 36 volt main pack, and a 10-cell 36 volt 'turbo' pack of a123's. During times the main batteries have excess rate capability, flick a switch to connect the turbo pack (itself in series) in parallel with your main pack to recharge. At some point before the a123 turbo pack had fully recharged, it's low impedance would mean it'd began to drain back to supply power to the motor along with the main pack. But maybe they'd get mostly recharged, and when they do, flick off the recharge/standy switch (or have this automated) and when ready for 72 volts, hit a separate turbo switch that throws the 36v a123 pack in series with the main pack until total system voltage drop below 56 volts (2 volts per a123 cell), then flick the standby/recharge switch back to recharge, throwing the turbo pack into parallel for another cycle.

 

[Leeps]

I would want a more civilized manner of charging the boost pack.
Joe

 

[safe]

For example you have two imaginary capacitors you charge each one with one amp for ten seconds to arrive at 10 volts, this gives you a one volt per amp per second change (or one farad). Now you hook them up in series pull one amp and the whole thing is discharged in ten seconds. This time you dropped twenty volts in the same time with the same current, this is a two volt per amp per second change(or one half of a fara).

Translation into math language:

Charging in Parallel:

Capacitor One: 10 seconds @ 1.0 Amp @ 1.0 Volt = 1.0 Farad
Capacitor Two: 10 seconds @ 1.0 Amp @ 1.0 Volt = 1.0 Farad

Total initial energy stored while charging = 2.0 Farad

Discharge in Series:

10 seconds @ 1.0 Amp @ ?? Volts = 1.0 Farad

Total energy discharged = 1.0 Farad

Conclusion:

The efficiency of capacitors used in series is 50%?

This seems wrong to me.... can you express the relationships in terms that follow this kind of "mathematical proof" format? (so that I can understand you better?)

 

[xyster]

The math is explained here:
http://en.wikipedia.org/wiki/Capacitor

In parallel the effective area of the combined capacitor has increased, increasing the overall capacitance. While in series, the distance between the plates has effectively been increased, reducing the overall capacitance.

In practice capacitors will be placed in series as a means of economically obtaining very high voltage capacitors, for example for smoothing ripples in a high voltage power supply. Three "600 volt maximum" capacitors in series, will increase their overall working voltage to 1800 volts. This is of course offset by the capacitance obtained being only one third of the value of the capacitors used. This can be countered by connecting 3 of these series set-ups in parallel, resulting in a 3x3 matrix of capacitors with the same overall capacitance as an individual capacitor but operable under three times the voltage. In this application, a large resistor would be connected across each capacitor to ensure that the total voltage is divided equally across each capacitor and also to discharge the capacitors for safety when the equipment is not in use.

 

[Leeps]

Okay maxwell makes a 2.5 volt 120 farad capacitor, no idea how much it costs but it offers 5 milliohm ESR. Theyre sized like C-cells for some reason.
So 72 volts/2.5 =29 capacitors in series, lets make it 30 just to be safe.
120 farads/30caps=4 farads
theres a 36 volt useable range so 36volts*4farads=144 ampseconds before the voltage drops to normal pack voltage.
well it is what it is
Joe

 

[safe]

I'd still like the "mathematical proof" kind of logic like I demonstrated above. Rather than trying to explain what is going on in vague sentences with numbers just prinkled in I'd like to see the "proof" or the "equations" done properly.

It's normally not that hard... (being "smart" means being a "good teacher" most of the time)

I frankly think that we are "missing" the central idea in looking at "static" descriptions of series verses parallel with capacitors.

The ESSENTIAL ingredient is the "switching" done by the relays so that what is CHARGED in parallel is DISCHARGED is series...

A simple question:

What is the energy efficiency of a capacitor?

 

[fechter]

Don't confuse energy stored with capacitance. Energy is 1/2 CV^2.
Two capacitors in series have half the capacitance, but twice the voltage.
Since the energy is porportional to the capacitance time voltage squared, the energy is doubled.

The units of energy are Joules. One Joule = one amp-second.

Capacitors are usually very efficient. There is some resistance loss, but generally over 98% efficient.

 

[Leeps]

Safe a farad isnt a unit of energy, its a measure of capacitance. Energy is 1/2CV^2.
use that for your equation youll find it works out
Joe
Fechter got it in while i was testing the equation

 

[safe]

Two capacitors in series have half the capacitance, but twice the voltage.

Okay, now I want to make ABSOLUTELY sure you understand what I've proposed here. Here's what I want to happen:

The capacitors are charged in PARALLEL.

...then we take a deep breath and switch the wires around with relays and then:

The capacitors are discharged in SERIES.

Now my "very simple" questions are:

1. EXACTLY how much energy is lost as a percentage? (rough guess)

2. Do I get increased voltage as a result? (like equal to the number of capacitors times the initial charge voltage)

 

[Leeps]

absolutely, i assure you we understand.
for the first one i dont know for certain but there isnt much loss

for the second thing yes your voltage is doubled, tripled, or whatever space and wallet size allows.

Joe

 

[fechter]

Yes, you can switch capacitors to make a very efficient voltage multiplier.
It won't be efficient if there's too much voltage difference between the capacitors and the source or load. By using FETs to switch at high speed, you can minimize this kind of loss.

 

[safe]

I was reading in Wikipedia about capacitors in series and that looks not to be a good thing because it actually tends to work against the ability to store energy. From Wikipedia:

"In parallel the effective area of the combined capacitor has increased, increasing the overall capacitance. While in series, the distance between the plates has effectively been increased, reducing the overall capacitance."

But I want to keep the capacitors separate and in parallel while charging and then "somehow" switch thing so that for discharge they combine. If I used relays to reconfigure the connections what would happen? It's a "different question" to talk about charging capacitors "in series" because it appears to "screw things up" a bit.

See my confusion?

How do you achieve the "high efficiency" charging and then flip things and get the "high voltage" output?

Is there something else that could be added into the relay cicuitry so as to "extract" the energy and not screw things up... assuming that things actually do get screwed up.