choke / reactor

[Vladpl]

Hello guys

I have designed many dc chokes (inductors). Now I have a job for a AC choke (Reactor). I have designed a few 3 phase reactors using formulas that I had form existing designs. So what I am asking is ... is there any differnce between a choke and a reactor? Would the design be the same for both since inductance is inductance. This reactor will be obviously connected to the mains and will operate at 50Hz. My understanding is that choke, reactor and a inductor are a same thing?

 

[Skogsgurra]

Yes, they are all the same thing. But, there may be a slight difference if you are designing commutation reactors for line commutated thyristor rectifiers, they have to handle the line-line voltage for the commutation period and that means some rather high du/dt and di/dt that you do not normally see in a pure DC or AC application. You may need to design for that or you will have a very acoustically noisy component.

If it is a normal reactor (just a piece of L) then I don't think that you should have to think about anything else then having laminated core - and you usually don't need the air gap found in DC chokes.

Gunnar Englund

http://www.gke.org

 

[zappedagain]

I like to think of a choke as a 'lossy inductor' that I use to absorb RF (AC) energy. Wikipedia gives a pretty good overview on inductors, chokes, and reactors:

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

 

[CarlPugh]

Reactors have ac flux in the gap. Any conductor near the gap, including windings will have loss.
Keep all copper or aluminum including conductors (Rule of thumb) 1.5 times the gap distance from the gap.
Multiple small gaps are better than a single or double gap.
There will be IR loss, core loss and gap loss.
Stainless steel is frequently used to hold the core together. Stainless steel has fairly high resistance and the loss will not be as great as if a good conductor is used.
If there is high current, it may be necessary to have multiple conductors and transpose the conductors in the winding.

 

[Warpspeed]

The design approach is quite different depending upon the relative proportions of dc current and ac voltage.

The extremes would be a dc solenoid, and an ac reactor.

A dc solenoid has no ac voltage present, so a solid iron core would work fine, there are no eddy currents. All you need to know is the flux density to ensure you are safely below core saturation. The only losses will be resistive copper losses.

A reactor may not have any dc component at all, but there could be a considerable ac voltage. So eddy current losses in the core, and flux swing are the main points of interest. Faraday's law, and normal transformer design parameters would be the way to tackle it. A reactor is basically an ac transformer without a secondary. There would not normally be an air gap in a reactor.

It becomes far more interesting when both dc and ac are present as in a choke, and the design will have to accommodate both.

My approach would be to design the number of turns and core cross section to support the ac component, just as with transformer design. Then the inductance and dc component is taken care of by introducing a suitable air-gap.

Quite often it takes several iterative attempts to come up with a suitable overall design that suits both the ac and dc requirements.