capacitors for heat

[fanman]

hello all

this is fanman back again, ive been off the net for a few months during my relocation, now im ready to hit the wind power again, anyway i built a axial flux machine a winter ago, with the idea of no bateries, only to heat with it, and the controller was my next step, now i see that were trying capacitors in line with the heaters to make the gen start up easier, thats a great idea, but how exactly does it hook up? i have a 120 volt  three phase axial flux capable of 5kw, 16ft blades, i run three wires down my tower to a three phase bridge rectifier, then to three 120 volt water heating elements, some of your sites say to hook caps up in series with the heater, could someone make me a schematic of this, i would very much appreciate it,

[tecker]

I think the caps need to work with the elements more IE rc time constants .I'm investigating this .

[electrondady1]

zubbly did two very good posts  and explanation(mostly over my head ) a few months ago.

http://www.fieldlines.com/story/2006/9/16/225354/917

thats the first one.

[veewee77]

If all you are doing is heating, why do you even need rectifiers.  The heating elemement will not care if it is DC or AC.

Enlighten me. . .

Doug

[RP]

One example would be where you have 1 or 2 heaters to connect across 3 phases in a balanced way.  In this case he has 3 phases and 3 heaters though so all 3 could indeed be connected with individual series capacitors and skip the rectifiers entirely.

[harrie]

Hi fanman, glad to see your back! I suppose you are using your battery bank to allow the mill to get up RPM before cutin, and also to use for other things beside heating??.Im trying to understand the use fo capacitors also, so Im not much help to you.

[smidy]

try a search for "putte" and i thinck you will find what your searching

[Gordy]

My understanding is that the caps. are needed to act like batteries to let the mill get started. With out them the heating coils act almost like a dead short on the mill leads.

Gordy

[CompDoc]

I believe this idea of caps was proven to be in-effective, can't remember the particulars...

I have posted a simple, yet powerful, circuit to 'actively' harvest turbine peek-power into heater elements on the DC side.  You'll find this is what you are looking for.  

If you are new to micro-controllers/electronics, I'll be happy to help you get going.

[Flux]

You can't use the capacitor method with dc.

If you load with ac, you need a load that is small at start up and increases with wind speed, ideally the load needs to increase such that the input power of the alternator increases with speed cubed. This comes close to output power rising with speed squared with typical alternator efficiency.

If you consider a circuit with a capacitor and resistor in series, at high speed ( high frequency) the capacitor  has a reactance smaller than the resistance and the alternator sees effectively a resistive load and delivers maximum power. As speed drops the capacitive reactance increases, this and the falling volts with speed means that there is less power lost in the resistor.

You can partly regard this as a variable impedance load, not quite the same as variable resistance but a bit similar. The capacitive current consumes no power so when the capacitive reactance is large the prop has to support little power even though the current flowing is still considerable.

This if chosen properly means a load that reasonably well matches the prop and allows start up and keeps sufficient load with speed to maintain a stable state.

The big snag is that with typical low speed alternators, the low voltage and low frequency means large capacitors and you need 3 sets for 3 phase.

The scheme becomes more attractive with high voltage and high frequency and anyone considering the method should wind for the highest practical voltage and frequency to keep capacitor cost within reason.

Flux

[sdscott]

Flux, can you elaborate on the size of non-polar caps we are likely to deal with in this real world of PM-alts?

[Flux]

To get some idea lets consider a machine with 120v line connected to a load consisting of 5 ohm resistors in star.

Phase volts will be 120/ root 3 = 70v approx.

Power per phase will be 70squared/5 = 980W  so we have a total load of nearly 3kW.

We would want Xc to be rather less than 5 ohms ( say 2 ohms) at the full load frequency.

If we assume a 12 pole alternator running at 500 rpm ( probably rather fast for 3 kW) then we need Xc to be 2 ohms at 50Hz.

10^6/( 2 x pi x 50 x C) = 2    Giving us about 1500uF per phase

You would need to fiddle to get the best compromise between starting and full load so it may work out lower perhaps 1000uF.

Now you can see the problem.

Flux

[sdscott]

That is a problem!  Thanks.

[alancorey]

I really like this explanation.  After 20+ years as an electronics technician I was still a little confused by capacitors in AC power circuits.  This I get:

It's a high-pass filter.  It's like hooking a tweeter up through a capacitor so it only gets power at high frequencies, because you don't want the bass driving it.  At low wind speeds (low alternator frequencies) the capacitor doesn't conduct much, but as the speed/frequency increases the capacitor conducts more until it approaches acting like a piece of wire.  Maybe that'll help someone else see it, I don't know.

  Alan