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

Re: VARIAC's -> parallel??




>Original Poster: Bert Hickman <bert.hickman-at-aquila-dot-com>
<snip>
>>         My friend was talking to an engineer at a variac company the
>other day.
>> He warned against putting variacs in parallel.  Apparently, if the output
>> voltage of each variac is a little different, there will be heavy current
>> flow between the outputs due to the unbalanced voltage.  He had some
>> special choke or something for $100 that would fix this problem and allow
>> them to be put it parallel.  Anybody ever heard this before??
>>
>>         We decided it was ok because we put the ballast resistors on each
>leg of
>> the three phase variac and then combined them.   So the resistors would
>> balance the voltages out before they were joined.  The engineer's concern
>> did seem very valid to us.
>>
>> Something to ponder...
>>
>>         Terry
>
>Terry,
>
>These are called paralleling chokes, and the commercial ones used for
>variacs are actually small, high current, 1:1 transformers. These are
>wired with the windings in series such that any voltage mismatch between
>the pair of variacs is effectively "cancelled". If the voltage
>difference between outputs of a pair of variacs is defined as V1-V2 =
>Vx, the balanced 1:1 transformer circuit (below) will share this
>difference at the output without causing large circulating currents.
>
>
>                    + Vx/2  -
>       V1  ----------OOOOOOO-----------  Vout = V1 - (V1-V2)/2
>                     -------    |
>                     -------    |
>       V2  ----------0000000----
>                    - Vx/2  +
>
>A less "elegant" way to handle this would be to use a small high current
>single-winding reactor in the output leg of any variac to be paralleled
>to the first variac. While this method will reduce the magnitude of
>circulating currents, it does not eliminate them as in the above
>circuit.
>
>-- Bert --
>
>

I thought I had seen these chokes on Superior Electric's web page
(http://www.warnernet-dot-com/sev_ps01.html).  The site has been down for
a few days, and today I did not see them.  At $100 each, it would get
expensive since n-1 chokes are needed for n sections paralleled
together.  This number is needed since one choke will combine two
unequalized voltages into one.

Regarding paralleling being done by manufactures, here is a quote from
the web site mentioned above:

    "The 1156D-1256D & 1296D Series are offered in single through
    27-gang, single and 3-phase types with ratings from 6 to 365 kVA."

I did not notice anything about whether or not these sections were
balanced via chokes.  For anyone interested in ordering a 27-gang model,
be sure to put plenty of money in your checking account:  I was told that
the cost of a new 1256D was around $900 per section!  It would make a
pretty impressive addition to a control panel though.

Does anyone know why manufactures do not just make single units in
larger sizes?  Or do they?

**********************************************************************

> Something else to ponder,
>
>Is it possible, sensible even, to connect two variacs of difference rating
>together in some way? I have one 8 amp and one 15 amp at the moment which
>is fine for neons. I am getting a 8 kV 250 mA transformer and am worried
>that the brushes on the 15 amp variac might not like it too much. 8 amp
>variacs are more common than larger ones here in the UK so I could gang 3
>of them together if I have to, but for now can I do anything with what I
>have got?. Am I missing something really simple here?.
>
>many thanks
>
>bob golding

With this setup you have two potential problems.  You have the possibility
of unequal voltages.  In fact, for two completely different variacs
(rather than two variacs of the same type) unequal voltages, at least
at some settings, would be likely.  Then you have the problem of properly
balancing the power loads between the two units.  Depending on their
relative impedances, paralleling may not even allow as much current
as using only the 15A variac since the smaller one may try (via a lower
impedance) to pass more of the current than larger one.  This is
similar to only paralleling distribution transformers with matched
%impedances.

Jim Monte