Re: Resonant charging design

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "R.E.Burnett by way of Terry Fritz <twftesla-at-qwest-dot-net>" <R.E.Burnett-at-newcastle.ac.uk>


Hi John & all,

"John Freau" wrote:

> What was the reason for choosing an 85% power factor? Is it 
> the best that can be obtained, or was 85% chosen for some 
> other reason?  Is it possible that better than 85% can be
> obtained with certain ballast arrangements and breakrates?

There is nothing magic about the choice of 0.85 power factor.  I
picked this value because I thought it was a relatively good power
factor and represented a significant improvement over figures like
0.5 that have been mentioned here before.

Why not 0.9 or 1.0 ???

Like most things,  there is a trade off.   You could design for
a PF of 0.9 instead of the 0.85 that I picked.  Working through
the design for PF=0.9 would result in _smaller_ values for the
ballast and the tank capacitor.  (ie. a higher resonant freq.)

The trade off is two-fold.  For a higher power factor,  the tank
capacitor voltage must peak higher to process the same power due
to the reduced capacitance.  ie. Higher voltage stresses.
Secondly, the reduced ballast inductance gives a higher fault
current if the HV side became shorted.  I considered a PF of 0.85
to be the best compromise,  but I am open to suggestions.

The best PF that I have ever managed to get on the simulator with
any break rate and ballast combination was 0.93   I found that it
was not possible to get a power factor better than 0.93 because
the current drawn by the transformer is not a pure sinusoid.  Even
if the current and voltage are in phase,  the PF is ultimately
limited to around 0.93 due to harmonic distortion in the current
drawn by the TC.  We often forget about the horrible waveshapes !

> I do realize that it's often hard to obtain better than 85%
> power factor especially at higher breakrates.

Simulations suggest that it should be possible to get to PF=0.9 at
say 800BPS,  but the required ballast inductance would be very
low !  Maybe there would be problems with power arcing in practice ???

> I noticed that you did not extend the graph down to 100 bps,
> were there problems with the simulations in that area?

100 BPS,  (or 120BPS in the states) obeys different rules ;-))
Seriously,  below 200BPS another parameter enters the equation.
_ROTARY FIRING PHASE_
At 200BPS and above,  the firing angle has no influence on the
PF or power throughput,  but below 200 the phase has a large
influence and becomes important.

If you run at 100BPS or 120BPS my method is not so useful
because you can control the power factor and power throughput
by adjusting the rotary firing phase instead.  I guess I should
have stated that this stuff is really most useful for 200BPS and
above,  but it does work for both SYNC and ASYNC rotary designs.

> I guess the power factor could be improved even more by using
> unequal electrode spacings...

Yes possibly, but this gets quite complicated because there are
so many variables !

> Can you explain why the graphs have the shapes and parameters
> they do?  For instance why do the charging frequencies have to
> be as shown, for best power factor?  I realize that the
> simulations show them to be best for best power factor, but I
> was just looking for more insight into why...

I can offer some expanation for the shape of the graphs,  but
I will do this in a separate post,  as this is getting quite long
already.

Thanks for your feedback,

						Cheers,

						-Richie Burnett.