This is a radio frequency (RF) and high voltage spike and
kickback protection circuit. It protects your power control
cabinet, and the low voltage 60 cycle house wiring. This circuit
is recommended for the low voltage side of all types of step up
transformers, regardless of taps, grounds, etc. on the windings.

X1 is the step up transformer. The center core is grounded to the
dedicated RF ground that also grounds the safety gap and the
base wire of the Tesla secondary.

RFC 1A and RFC 1B are about 5-10 turns of heavy insulated wire
(sufficient to carry the current requirments of X1) on a large
iron powder or wire ring toroid core. I use 4-6 diameter iron
powder toroids and heavy cable to wind these chokes.

"Protective Capacitors PC1 and PC2 are not critical and can be
rated in the vicinity of .5 to 2 microfarads. Use a voltage
rating as high as possible. The usual 400-600 volt capacitors
will not withstand kickbacks for very long. 1 KV rating or more 
should work fine."

In place of this circuit, a large heavy duty commercial EMI/RFI 
line filter may be placed here. These filters offer the
convienience of good efficiency in a compact unit. Where current
throughputs are high I use several in parallel. Quality
commercial line filters employ iron powder chokes, as well as the
"PC" capacitors of the circuit at the top of this post. The Line
Filters I use also have RF choke coils in the ground path; the
ground wire can be run reversed (it is neutral) and can be used
to trap stray RF, preventing ground path contamination to the 60
cycle breaker box. I use a minimum of two independent grounds.


  Date: 08-08-94  15:25
  From: Terry Smith                                
    To: Richard Quick                              
  Subj: Tesla Coils

       PC1     X1       RFC 1A       RFC 2A
Ŀ> TO TESLA TANK
            )(              
            )(              
              )(              O
grnĴ      )(    grndo SAFETY GAP
              )(              O
            )(              
            )(              
ٺUUUUUUUUUU> TO TESLA TANK
       PC2              RFC 1B        RFC 2B

 RQ> "Protective Capacitors PC1 and PC2 are not critical and can  
 RQ> be rated in the vicinity of .5 to 2 microfarads. 


One little point of safety, which should be pointed out to the
folks with limited electronics safety knowledge, is that PC1 and
PC2 would have to be NO HIGHER than 0.1 uFd, in order to meet a
5mA ground current leakage standard. The circuit and values of
PC1 & PC2 shown above could be lethal if the building ground
connection were marginal or missing.  

Obviously that would decrease filtering effectiveness.  A 120 V
isolation xfmr, or additional stages to add both filtering and
safety related line isolation, as you later described, would be
possible solutions ot that problem.  BTW, I have seen foolish and
inattentive engineering of power supplies for broadcast equipment
from one reputable and fairly quality oriented manufacter, which
caused 6 and 12 mA of ground leakage with power on and off, due
to similar low cost but irresponsible RF filter cap size choices. 

 RQ> rating as high as possible. The usual 400-600 volt           
 RQ> capacitors will not withstand kickbacks for very long. I     
 RQ> prefer capacitors with ratings of from 2500 to 5000 (or      
 RQ> higher) volts"

What are "the usual"?  Orange drop or block type small leaded
film or mica caps?  I would think that current ratings (which
would be related to ESR, and filter effectiveness) might be more
important than voltage rating, so long as no less than 400-600
volt rated caps were used.  At 450 kHz, the Xc of the range of
caps we're discussing would be 0.1 to 3 ohms, which makes me
wonder if "voltage rating" isn't being substituted for physically
larger cap with higher thermal and current limits?  Such overall
higher rated caps should be more effective as RF filters due to
the same lowered ESR which results from other design parameters
being increased.  

Do you know if smaller cap failures are from voltage punch
through, or if it's really from cooking due to a 3-5 amp
repetitive surge current demand, which small inexpensive caps
can't handle?  

BTW, I'll be Freq'ing your archive tonight.  I'm curious about
what resonant circuits you might have brewed.  Thanks for making
it available! Terry  (203)732-0575 BBS (1:141/1275)


  Date: 11 Aug 94  17:01:49
  From: Richard Quick                                
    To: Terry Smith                                  
  Subj: Tesla Coils

       PC1     X1       RFC 1A       RFC 2A
Ŀ> TO TESLA TANK
            )(              
            )(              
              )(              O
grnĴ      )(    grndo SAFETY GAP
              )(              O
            )(              
            )(              
ٺUUUUUUUUUU> TO TESLA TANK
       PC2              RFC 1B        RFC 2B

 
RQ> "Protective Capacitors PC1 and PC2 are not critical and can
RQ> be rated in the vicinity of .5 to 2 microfarads." 

 TS> ...PC1 and PC2 would have to be NO HIGHER than 0.1 uFd, in   
 TS> order to meet a 5mA ground current leakage standard.  The    
 TS> circuit and values of PC1 & PC2 shown above could be lethal  
 TS> if the building ground connection were marginal or missing. 

Good point, though yours is a worse case scenario. There were
already several reasons why I did not like the circuit above, 
you just added another reason to the list.
  
 TS> Obviously that would decrease filtering effectiveness. A     
 TS> 120 V isolation xfmr, or additional stages to add both       
 TS> filtering and safety related line isolation, as you later    
 TS> described, would be possible solutions ot that problem. 

My improvements on Harry's circuit above work pretty well, and I
have always loved isolation xfrmrs as an additional safety/RF/
spike/surge measure. I am considering the addition of a pair of
back to back pole pigs in my low voltage feed lines for just this
purpose.

 RQ> "Use a voltage rating as high as possible. The usual        
 RQ> 400-600 volt capacitors will not withstand kickbacks for     
 RQ> very long. I prefer capacitors with ratings of from 2500 to  
 RQ> 5000 (or higher) volts"

 TS> What are "the usual"?  Orange drop or block type small       
 TS> leaded film or mica caps?  

Again, I am quoting Harry Goldman above, but generally speaking
the "PC" caps used in Harry's circuit are the typical "can" type
caps used in capacitive start motors. I have seen a lot of these
used (and NO they are not filter caps) simply because they are
cheap; surplus caps with a rating of .1 uf @ 400 vac are locally
available for around $0.25...

 TS> I would think that current ratings (which would be related
 TS> to ESR, and filter effectiveness) might be more important    
 TS> than voltage rating, so long as no less than 400-600 volt    
 TS> rated caps were used.  At 450 kHz, the Xc of the range of    
 TS> caps we're discussing would be 0.1 to 3 ohms, which makes me 
 TS> wonder if "voltage rating" isn't being substituted for       
 TS> physically larger cap with higher thermal and current        
 TS> limits?  

Very possible. 

 TS> Such overall higher rated caps should be more effective as   
 TS> RF filters due to the same lowered ESR which results from    
 TS> other design parameters being increased.

Good point.
   
 TS> Do you know if smaller cap failures are from voltage punch   
 TS> through, or if it's really from cooking due to a 3-5 amp     
 TS> repetitive surge current demand, which small inexpensive     
 TS> caps can't handle?  

I don't know. You could ask Harry Goldman (I have posted his
SNAIL address should you choose to correspond), as he recommended
the above circuit (recently too); I have never used it. But in my
experience voltage punch through has only occurred as a result of
a typical coiling type "incident" where there was no doubt as to
the cause (run away oscillator, direct strike to low voltage feed
lines, xfrmr breakdown, etc..). Perhaps this is what he was
referring to when he said the lower voltage caps don't last very
long. On the other hand a repeated 3-5 amp surge current demand
on the filter implies something else is wrong... perhaps he is
not properly grounded or choke/filtered at the RF side of the
xfmr???  Like I said... I don't know, but I would not recommend
the above circuit.

