Re: Saturable Reactors

Eric Davidson <edavidson-at-icva.gov> wrote:

>The construction of a 'home brew' saturable reactor is quite a bit 
>more complicated that it first appears.  I looked into the potential 
>of an SR to control my pig, about a year ago.  Most SRs are wound on 

So I gather you abandoned that approach?

>an E-I core with the controlling winding on the center leg.  The 
>outer two legs contain the reactance windings.  

Actually the first form of saturable reactor had just one control 
winding, and one power winding.  However they were only capable of 
"passing" half the phase, or 180 degrees (since during the other half of 
the polarity the power amp*turns would oppose the control amp*turns).  
They were originally used in pairs to achieve 360 degree control.

>They must be wound in opposite directions to cancel their MMF to 
>eliminate inducing AC in the control winding.  

Yes - the control winding is wound into two opposite coils, electrically 
in series, on the center leg.  Sometimes the center leg is gapped.  The 
two-coil control approach is also required for 360 degree phase control!

>The control winding is usually several HUNDRED 
>turns of 24 or so AWG wire, and is powered by a few hunderd volts.

Could be - you need to get enough amp*turns to saturate out the upper or 
lower half of the leg (depending on which half of the 360 degrees you 
are in)

>  I contacted several transformer/reactor manufacturers and they said 
>that they are very difficult to design properly. Sure, you could home 
>brew one, and it may work ok, but look for alot of nonlinearities 

I'm guessing w/o the proper lamination material (with the square B-H 
curves) performance would be greatly compomised.  Sadly, regular 
transformer laminations do not have this property.

>   A pig itself can be used as an SR.  To me, this is a better 
>possibility than building one from scratch.  The way you do this is 
>to open the tank and connect the secondary windings in antiparallel 
>X1 to X4 and X2 to X3) and place them in series with the load.  
>Connect the DC controlling current to the HV bushings.  Make sure you 
>have the proper connections on the secondary, or the HV will appear 
>on the primary!!  This is not good.  Good luck, and be careful!

Hmmm...  Have to think about that one a bit.  Did you try this?  Seems 
like it *might* work!  The problem is that the "control" winding in this 
case does not consist of two contrary-wound coils... :( ??  Maybe if you 
could get a pig with two HV windings - then you could do the 
antiparallel trick to them and make it work.

Maybe if I could find a 3-legged core (like out of a 3phase xfmr) some 
experiments could be performed...

Thanks for that post, Eric!


Get Your Private, Free Email at http://www.hotmail-dot-com