Deleting the e-starter

[Gorminrider]

...one thing that occurs as a theoretical...if the capacitor is really really big, and discharged, it acts like a short circuit (unless there's a ballast resistor in series.) So, whether series or shunt, the stator will have full output for some period of time...somewhat similar to recharging a very depleted battery... maybe a Fuse issue, eh..? This isn't going to be a problem kickstarting it, but ....

[Chuck78]

From the advrider post link above:

http://en.wikipedia.org/wiki/CapacitorCapacitor

A capacitor is two metal plates separated by an insulator. If you apply a voltage to the plates, they will store an electric charge. If you disconnect the voltage source the charge will remain on the plates and the measured voltage on the plates will also remain until it is drained off by some external circuit. The amount of charge stored on the plates is related to the physical dimensions of the plate and the voltage applied to the plates.

Charge stored = Voltage across the plates * Capacitance of the particular device.



Charge is measured in Coulombs, Voltage in Volts, and Capacitance in Farads.
An Amp is the rate at which charge moves, in Coulombs per second.
A coulomb is 6,241,506,000,000,000,000 electrons.

Capacitors also have a voltage rating, usually printed on the case next to the capacitance value. It refers to the highest voltage that can be safely applied to that capacitor.

In this case, guessing by the picture, the capacitor is probably 5000 microFarads. So when charged to 15V, it stores 15*.005 = .075 coulombs of charge. Compare that to a 7Amp-hour battery which stores 7*3600 = 25200 Coulombs.

The point of the capacitor is not to provide power for any accesories -it can only run a taillight for a fraction of a second. It is there to help the conversion of AC power from your stator to smooth DC power. The rectifier converts the AC into pulsed DC, the capacitor has to provide power for the dead time between the pulses. In the case of devices running off wall current, the cap has to power them for 1/(60Hz*2) = 8.3mS. For motorcycles, the frequency is higher, 1/(240Hz*2) = 2.1mS in the case of my XR. The *2 is because there are two pulses per cycle.

So using the XR as an example, The load is 6A, so the voltage on a 5000uF cap will drop 6 * .0021 / .005 = 2.5V. So if the regulator limits the voltage to 14.5V during the pulse, the voltage will drop to 12V just before the next pulse.



A battery uses an electro chemical reaction, and will generate a fixed voltage, that will slowly decrease as the chemistry is expended.

A rechargable battery is the same, but uses chemistry that can be "reversed". You feed power into it, it undoes the previous chemical reaction, you pull power from it, it re-does the chemical reaction.

A capacitor simply stores a charge... and will seek the voltage applied to it. So if you drop 13.2 volts across a capacitor, it will draw a LOT of current quickly until it stores enough energy to go up to the applied voltage (13.2 in this example). A capacitor will quickly go to whatever voltage you apply to it (or explode trying) based on the resistance of the current path. The more resistance, the slower the voltage goes up (by a deterministic RC time constant).

Thats a lot smaller cap then I would have expected for a battery replacement. And if I were building one, I would want some sort of current limiting, a simple resistor with a diode in parallel (arrow pointing away from capacitor) would do the trick nicely, it would make the cap charge slowly, but discharge quickly if necessary... though at the cost of a .6 volt drop.

All in all, a battery is probably a better solution, though something like this could work and weigh next to nothing. Mendelsons electronic surplus in Dayton Ohio has about 4 acres of electronic surplus, and you can probably buy a mil spec capacitor the size of a beer can for $5.

[Chuck78]

Basically most "battery eliminators that mostly bobber/chopper/"cafe" builders buy off the shelf are around 4700uF, and will give you this erratic voltage pulsing as mentioned by one of the two people I quoted, voltage fluctuating from regulator output (say 13.2v) down to 12v in 2 millesecond cycles... A larger cap of at least 14,000uf will have much more voltage storing capability and not drop like this, as some bikes will definitely have an erratic idle and flickering lights at idle speed with the smaller cap. And you don't have to have one the size of a beer can, look at the $15 22000uF cap I posted, size listed as 120mm x 50mm.

[Gorminrider]

mmm.well, as to ripple, and all that,
I got curious and so I idled my gsx400 yesterday without a battery or capacitor just as a curiosity but didn't actually notice any ripple in the lights at 1600-1800rpm...the bike will just quit if the revs are too low.My cheapo led digital voltmeter freaked out and went into a swoon so there is ripple, I bet.
Given stator coils are spinning at about 1800 rpm : divided by 60 seconds is 30 cycles per second even were it the simplest of 1 phase ac generators,let alone 3 phase-so say 90 cycle? compare this to housepower at 60 cycles.
but perhaps the capacitor is really needed because there's something not quite right ..lights not as bright, spark not as snappy..something.Perhaps where a battery or capacitor can "store" the peaks the regulator is passing, running the system without results in a lower average voltage. I will have to find an old analogue meter to see, I guess.

It occurred to me that if I were doing this,with a kickstart,and the Miraculous Capacitor I might keep the stop switch off for a few kicks?....

[850 Combat]

More than likely, below 1800 rpm, you don't have enough voltage to operate the electronic ignition. Points would probably keep it idling.

[Gorminrider]

aha-likely so.well spotted,850'-that's a better diagnosis.... especially cogent as my stator in this bike is amateur-wound and a bit low on the windings-count as well...

(but 'tis good enough in normal usage, I need say.)