Re: Power factor correction capacitors for MOTs

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

Original poster: "Kurt Schraner" <k.schraner-at-datacomm.ch> 

Hi Ed,

while not supplying a satisfactory answer, I "think" beeing able to support 
your view of needing to consider the "backtransformation" of the  primary 
tank-cap to the low voltage side of the transformer(s), in order to 
estimate the power factor(PF). In fact I'm quite shure, this is a decisive 
factor. The system must be considered as a whole, in estimating  real- and 
apparent power draw. Extended PSPICE simulations would probably supply an 
answer. I'm not shure, if perfect compensation is possible at all, because 
of the nonsinusoidal waveforms we have, but fairly good compensation of 
around 0.9 should not be too hard to attain.

I can offer a number of measurements, taken during my 2002 presentation of 
the somehow biggish coil "B&W", in the Voltahalle place, which can be seen at:

http://home.tiscalinet.ch/m.schraner/Voltahalle1.htm etc.

The measurements, however, are partly inconsistent (with my view of 
matters!). Here are the facts:

26.06.2002	setup 3				
Transformers: 4 serial 100V connected PT's with parallel 16kV sec's.		
Primary Tank capacitor = 218nF; transformed to primary side = 348.8uF		
Ballast inductance by bridge type L-meter -at- 120 Hz = 11.6mH			
Resonance of transformed Ctank & primary-ballast-L: f = 79.12Hz

Line  Transformer Current VA draw  Wattmeter	Power
Volts  prim.Volts  draw            reading P	Factor
[V]     [V]         [A]    [kVA]     [kW]	[ - ]
300     310         40       12      10         0.8333
340     350         42     14.28     11.6       0.8123
360     380         45     16.2      13.5       0.8333
380     400         46     17.48     14.2       0.8124
380     400         46     17.48     14         0.8009
400     410         48     19.2      14.5       0.7552

(PSPICE simulation of this setup gives a PF of ~0.808)
The coil was always operated with an SRSG at 200BPS. The higher value of 
primary transformer volts vs. line volts fed, reflects some resonant rise, 
while the cap is yet smaller than resonant.


26.06.2002	setup 2				
Transformers: 4 serial 100V connected PT's with parallel 16kV sec's.		
Primary Tank capacitor = 218nF; transformed to primary side = 348.8uF		
Ballast inductance by bridge type L-meter -at- 120 Hz = 15.7mH			
Resonance of backtransformed Ctank & primary-ballast-L: f = 68.01Hz		

Line  Transformer Current VA draw  Wattmeter	Power
Volts  prim.Volts  draw            reading P	Factor
[V]     [V]         [A]    [kVA]     [kW]	[ - ]
300     260         32      9.6       6         0.625

This measurement doesn't fit in my view: f = 68Hz seems closer to the 
line's 50Hz value for resonance, but the PF is less, as well as the 
transformer's voltage shows no resonant rise. (I don't quite trust the 
15.7mH value of the ballast).


26.06.2002	setup 1				
Transformer: 400V / 16kV leak-flux-Xfrmer; unknown leak inductance.	
Primary Tank capacitor = 102nF; transformed to primary side = 163.2uF		
Ballast inductance is the unknown leak inductance of the transformer		
Resonance of backtransformed Ctank & primary-ballast-L: f = ?			

Line  Transformer Current VA draw  Wattmeter	Power
Volts prim.Volts   draw            reading P	Factor
[V]     [V]         [A]    [kVA]     [kW]	[ - ]
360	350         26      9.36      8         0.855
360	350         24      8.64      7.5       0.868
380	360        27.2    10.336     9         0.871

Here, good PF seems not to correspond with elevated primary voltage on the 
Xfrmer, when activating the coil, what I would expect. May be, I need the 
list member's help, to see something obvious... ???

BTW: Power factor was never an issue during our tests with B&W. The above 
results are just a sideproduct from my operations-notebook.

Best regards,   Kurt