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

Re: Microsim and filter design



Original poster: "Terry Fritz" <twftesla-at-uswest-dot-net>

Hi Dunckx, David, All,

I changed my big coil model to this system.  I changed it so it charges
across the cap so the filter signal is super nasty for testing.  I used
250mH chokes, 10K resistors, and 2nF filter caps on each leg.  the values
of the inductors and caps is larger than I would like, but the idea seems
to work very well!

I posted the MicroSim model at:
http://hot-streamer-dot-com/temp/Dunckx.sch

I also posted screen caps at:
http://hot-streamer-dot-com/temp/Dunckx.gif

The top picture shows the model with the modified filter.  The second shows
the filter with the cap signal in green and the NST signal in red.  Note
how the signal is significantly smoothed out by the filter :-)

The third shows the RMS current through the filter resistor and inductor.
the resistor is seeing about 40mA so 0.040^2 x 10000 = 16 watts per leg or
32 watts total is being wasted.  This compares to 74 watts total being
wasted in the coil's original design.  Of course, that can probably be
greatly improved with more work.

It does require large value (250mH) inductors that can withstand the NST
voltage across them (I was getting 14.8kV spikes).  That should not be too
hard to do since the coils are pretty conventional.

Cheers,

	Terry

 

At 10:27 PM 1/22/2001 +0000, you wrote:
>Hi David, Terry, All!
>
>Date: 22 January 2001 01:36
>Subject: RE: Microsim and filter design
>
>
>>Original poster: "Terry Fritz" <twftesla-at-uswest-dot-net>
>>
>>Hi David,
>>
>> You want the filter to reduce the high frequency ringing
>that reaches the
>>transformer.
>
>
><snip>
>
>
>>>I then
>>>substituted the filter you suggested and repeated the
>simulations both with
>>>and without the damping resistor. Results can be found at
>>>www.corridor-dot-net/deano/NewFiltr.zip . (336KB)
>
>
><snip>
>
>Thanks for your input on this.  I have "had a go" myself,
>though being a complete neophyte on the microsim thing, have
>no great confidence yet in my ability to drive it. The
>learning curve is vaguely Himalayan and my simulations are
>presently perched precariously half way up the South Col ;-)
>
>The following comments should therefore be treated
>_with_*great*_caution_ !
>
>
>My initial observations are:
>
>1) The "little L", L,C components need selecting on the
>basis of the radio frequency being generated by the coil.
>For 100kc/s, "little L" = 50mH, L = 400mH, C = 10nF+ looks
>interesting. Obviously the inductance and capacitance needed
>for higher frequencies is less. The previous values I
>suggested are better for e.g. 400kc/s.
>
>2) There may be quite a high current circulating in the
>damping resistor, which needs selecting on the basis of the
>"little L" and C values.  I remain uncertain about the
>appropriate power rating for this component.
>
>3) Care needs to be taken to ensure there is not a
>possibility of resonance between {"little L" + L} and C with
>the break rate 8<O
>
>
>In other words, the filter needs designing for each
>individual coil (though it isn't exactly hard!)  However,
>from my limited twiddling, it does seem that it _may_
>(emphasise uncertainty!) be possible to design a good filter
>with 1-2% input power loss.
>
>I _may_ (emphasise high probability) be quite wrong here
>(CAUTION: learner driver!) and it would be great if more
>simulations on this thing can be done to test these findings
>prior to cooking a transformer with this filter design!  I
>wouldn't want anyone to lose a tranny (or worse) thanks to
>my inexperience with microsim.  All input/comments
>gratefully received.
>
>Dunckx
>