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

Re: Blown tranformer

Original poster: hvdave <hvdave@xxxxxxxxxxxxx>


I'm not advocating using an RLC filter, however if an RLC filter is over damped then there isn't any extra ringing. Even for a single pulse. The situation gets down to the actual component values chosen. For a given value chosen for R and C, there is a safe amount of L that can be added. Unfortunately for practical values of R and C that you'd like to use with an NST, the amount of L that can be added doesn't make a significant improvement. We can't have R too large since that sucks up too much power and too much C will load down the NST or worse cause undesirable resonance with the NST. So these limit how much L you can really add.

Since the filter C is usually placed across the NST output, then the thousands of Henries of the NST secondary winding is moot. The filter series RL just see the capacitance of C.

Really though, the NST winding has a significant amount of distributed capacitance since it is wound is consecutive layers. I've seen some strange resonances coming out of my 15k/60mA NST's with various loads. These resonances look like low order harmonics of 60Hz resulting in the output to not look like a sine wave at all. I have two NST's, both with the same 15k/60mA rating but one is a bit larger than the other. The resonances between the two are definitely different. Light load or adding some capacitance across the NST can make them worse. Under heavy load the output waveforms seem more normal and closer to what you would expect. The NST is definitely a weird device.

On my initial coil build, I used an RLC filter and had no problems. I did carefully select the component values to insure that the filter was over damped. I did change over to a modified Terry filter (RC) since I believe that has the advantage of being able to use a lower R value which doesn't throw as much power away. I wanted bigger sparks. The Terry RC filter does have less attenuation at the secondary resonance, but the coil puts out bigger sparks. The NST just doesn't seem to care one way or the other, especially with those MOV's. So my experience is that the the Terry filter works better than RLC.

Regards
Dave Lewis



 since I didn't really see any need for the L.

Tesla list wrote: 
Original poster: "Lau, Gary" <gary.lau@xxxxxx>
I don't agree with the posted advice on how to construct an RF filter.
The secondary winding inductance of an NST measures in the THOUSANDS of
Henries.  Adding a few microhenries in series with each winding will
accomplish nothing useful.
I was unfamiliar with ignition ballast resistors so I had to Google
that.  They would appear to range in value from 1.3 to 1.6 Ohms.  Given
that the resistance of an NST's secondary is several KOhms, adding such
a low series resistance will accomplish nothing useful.
Even if an L-C low pass filter were used, this would only be useful for
attenuating a steady-state signal.  When L-C filters are used with spark
gap coils, the series inductors will ring at hazardous voltages each
time the gap fires, and the NST will experience worse RF than if no L-C
filter were used.  Adding a resistor (typically several KOhms) in series
with the inductors will reduce their Q and reduce their ring-duration,
but the initial oscillation is just as "loud".
The current wisdom for constructing an RF filter is to use an R-C
configuration.  About 500pF from each NST HV bushing to RF ground, and
about 1K from each HV bushing to the main gap terminals.
I whole-heartedly agree that a PROPERLY SET safety gap should be used in
addition to any filter, as each protect against different hazards.  My
complete thoughts on this topic are on my web site:
http://www.laushaus.com/tesla/protection.htm
Regards, Gary Lau
MA, USA