[Prev][Next][Index][Thread]

The saturable transformer, or "Taming the MOT"



Gang!

I was searching the IBM patent database for saturable reactors, but
patents for the real thing are so old, that they ae not iincluded in the
database, it only reaches back to the start of the seventees.

One patent stood out, this:
4122385
Inventor: George A. Oberbeck

http://patent.womplex.ibm-dot-com/cgi-bin/viewpat.cmd/US04122385__

http://patent.womplex.ibm-dot-com/

This excelent patent describes the saturable transformer in some detail,
which is not all that hard to grasp, let me have a shot:

You grab 2 MOT`s and bring them to a machineshop that has a metal
bandsaw, and instruct them to cut them like this:

       Please view in a fixed-width font such as Courier.


                  z
                  z
                  z
   +++++++++++++  z  ++++++++++++     ++++++++++++++++++++++++++
   +           +  z  +          +     +                        +
   +++++++++++++  z  ++++++++++++     ++++++++++++++++++++++++++
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +   +   +  z  +  +   +   +     +   +                +   +
   +   +++++   +  z  +  +++++   +     +   ++++++++++++++++++   +
   +           +  z  +          +     +                        +
   +++++++++++++  z  ++++++++++++     ++++++++++++++++++++++++++
                  z
                  z
                  z
                  z

The one on the right may need hand sawing off the midle leg with a
hacksaw, due to the small dimention on the windows, but a good tool shop
will have a bandsaw, the perfect tool for the task.


Rearranging the cores like this:  

Please view in a fixed-width
     font such as Courier.


  ++++++++++++++  ++++++++++++++
  +            +  +            +
  ++++++++++++++  ++++++++++++++
  +   +ccc+    +  +   +ccc+    +
  +   +ccc+    +  +   +ccc<-----choke coil
  +   +ccc+    +  +   +ccc+    +
  +   +ppp+    +  +   +ppp+    + <--- choke cores
  +   +ppp+    +  +   +ppp+    +
  +   +ppp+    +  +   +ppp<------primary coil
  +   +++++    +  +   +++++    +
  + --flux-->  +  +  -flux->   +
  ++++++++++++++  ++++++++++++++
  ++++++++++++++++++++++++++++++
  +     ----flux-------->      +
  +   ++++++++++++++++++++++   +
  +   +ppp             ppp +   +
  +   +ppp    ssss     ppp<------primary coil
  +   +ppp   sssssss   ppp +   +
  +   +     sssssssss      +   + <--- Power core.
  +   + sssssssssssssssss  +   +
  +   +ssssssssssssssssssss+   +
  ++++++++++++++++++++++++++++++
  +                            +
  ++++++++++++++++++++++++++++++
       ssssssssssssssssssss
        ssssssssssssssssss
           sssssssssss      
             sssssss
               ssss

You have all you need for a transformer, whose normal state is that of
almost total current limiting.

The primary windings are put around all the cores, and it should be
noted, that any of the 3 cores are large enough to provide enough
impedance in the parallelled primary windings.
Therefore, the flux in the cores take the path of least resistance
(reluctance, really, but don`t worry about it), and the secondary on the
powercore doesn`t have a chance taking out any power, because the
primary is quite content on living off the 2 choke cores, and if you try
to take power off the secondary, the flux will shift over into the choke
coils.

That is, unless you decide to wind one more coil, the choke coil, which
will saturate the iron in the choke coils, so that more and more flux is
directed trough the power core.

The aforementioned patent takes special aim at patenting the concept of
2 more coils, shorting coils around the choke coils, to shunt
differences in the shunt coil output, due to production variation.

Otherwise, the transformet is old hat, and since it`s failuremode ->
faulty choke coil power source, it just falls back on good oldfasioned
choke current limiting.

This is a real sane way to deliver high voltage energy, and I will build
a prototype of this, but will also give others a chance to catch up, and
do some practical testing.

Cheers, Finn Hammer