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

Re: Why Secondary Q Matters



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

Hi Malcolm,

At 08:44 AM 9/6/2002 +1200, you wrote:
>Hi all,
>        Here are some thoughts on this topic for you to ponder. 

"I" ponder this topic all the time now :-)))

>
>    There is one situation when secondary Q doesn't matter - i.e. 
>when the need for it is over and done with: when streamers have 
>attached to something forming an arc.

Like a ground.  If they are air streamers, the coil spends a long time
driving it and gives Q a chance to drain even more energy.  Air streamers
actually are a surprisingly "light" load.  Grounds are a surprisingly "dead
short" load ;-))

>
>    There are two ways of generating attached streamers: the first 
>when an discharge rod etc. is placed close enough to the terminal for 
>an arc to form, and secondly, when an airstreamer has propagated 
>through repetition to the point where it connects with an object 
>considerably increasing the discharge currents and dragging the 
>loaded Q down - power losses in output discharges are good :)

Air streamers mostly draw current due to the extra capacitance they create.
 As this capacitance is feed with current through the highly resistive air,
it makes a hot path.  That's are beloved streamer :-))  Arcing to a ground
actually has very poor impedance matching but it tends to occur when the
terminal is at high voltage and can deliver a nice single blast of current.
 Times like this are when you want very good grounding.  

Note that Paul's Qvar tests suggest that the surrounding environment can
have a traumatic affect on secondary Q far aside form the mechanics of the
secondary coil!!

http://www.abelian.demon.co.uk/tssp/qvar070402/

Avoid wet rainy areas like the plague!!!

>
>    Scope traces of operating secondaries show that for air streamers 
>only, the loaded Q of the secondary remains relatively high. Multiple 
>energy trades with the primary occur (if the gap system allows them 
>to). Sparks do not propagate well if something inhibits the 
>secondary's ability to ring, either continuously if the gap quenches 
>at first notch (decrementing by streamer power dissipation of course) 
>or through re-ringing as multiple trades with the primary proceed. In 
>this second case, a high unloaded primary Q is also required. This 
>can be facilitated by operating with a high primary impedance and low 
>gap currents. High primary reactance at Fr is important for this to 
>occur.

This is a fun graph (in this case, it is a horrible graph!):

http://hot-streamer-dot-com/temp/OLTC08-31-01.gif

It shows in red the instantaneous power dissipated in the primary coil as
heat.  The puny green line is the streamer power.  As the secondary rings
down, the low secondary Q eats away (like a school of piranha) at the
streamer's power. 

>
>     Secondary Q is important for two reasons - to generate initially 
>the highest possible voltage for a lump of Ep (more bang for buck) 
>and to retain energy that is not consumed in discharges *over each 
>half cycle of ring*.

YES!!!

>
>     Typically, we want streamers to stretch out over great distances 
>and connect with something. Much more exciting than placing a 
>discharge rod close to the terminal IMO. In my experience, a system 
>suffering from low unloaded Q does not promote streamer growth 
>through repetitive primary bangs at all well. A symptom I've observed 
>with such a system is that arcs cannot be formed that are 
>substantially longer than the length air streamers reach. It appears 
>to be a requirement for streamer propagation that prolonged ringing 
>per bang is needed to heat an ever-increasing length of air.
>
>     I'm not a great fan of matching-secondary-impedance-to-spark 
>theory because the impedance of the spark is in fact going to be 
>determined by both by available energy to feed it and the type of 
>spark it is (air or attached). (I should qualify that by saying that 
>the pressure, temperature, mositure content and molecular content of 
>the gas must also be a determining factor)

When running models, you need a "good guess" to plug in for what the
streamer will do to the system.  220K + 1pF/foot of arc length does this
job. It is an "average" of a far far more complex event.  A streamers has a
very wide range if "impedances" over it's short life.  Over the years, all
our coils have really become quite the same.  We don't vary much from the
"recipes of success".  I strayed from it, and see what happens! :o))  Our
present coils are all about in the same order of magnitude for output
impedance that streamers like.  

The present OLTC project presses the models to the breaking point!!  With
it's very precise gap control, it can be tuned to different stages of
streamer impedance.  It "could" detect if a streamer were going to air or
connected to ground and adjust the gap to optimize the situation (stop
quenching and go for the big power arc >;o)).  But the simple 220k=1PF/"
falls far short...  Here is a modeled result that suggests my coil is not
that bad after all:

http://hot-streamer-dot-com/temp/OLTC09-05-06.gif

The low Q is bad, but not too terrible.  The difference is that this model
assumes different dynamic secondary impedances (this model is actually
probably right!).  I guess we can start abbreviating Dynamic Streamer
Impedance as "DNI"...

I ran secondary Q arrays today for some conventional coils and it seems
that, in most normal cases, Q is low enough not to make a giant difference.
 However, we are close enough to the "edge" that a screw up can really bite
you.  If Rsec gets close to 1000 ohms, big problems start!  I think I am
going to steer away from SonoTube from now on.  It is nice and cheap, but
it's low Q effects are gnarly!!

>    In the first case, the spark impedance is relatively high 
>(witness the relatively high loaded Q of a system producing 
>airstreamers) and in the second, rather low (the loaded Q of the 
>system has dropped into the dirt). This is why a large terminal 
>capacitance makes loud bright arcs that we all love. One might infer 
>from that that the secondary is doing most of the delivering to the 
>air streamers and the terminal to the arc, and we want the streamers 
>to generate the long stretch prior to a connection with on object 
>being made.  The scope waveforms strongly suggest this is what occurs.

Yes indeed!  Very high voltage and high impedance to start, and then high
current later on when the streamer is large.  Although us "modelers" just
lump it all together right now, the dynamic effects are starting to become
very important!  Expect to see for more dynamic streamer impedance data
real soon!!

BTW - Paul, I am going to order some brass balls of various sizes tonight
for the streamer impedance tests.  No one in town had anything well enough
made (goofy drawer knobs) for the proper test.

>
>I'm sure this will be regarded as a lot of unqualified handwaving in 
>some quarters. It seems to be borne out in the field though. I'll be 
>watching further development of more scientific modelling with 
>interest.

You seem to be waving you arms in the right direction.  You will note a
number of folks are following your lead ;-))

Cheers,

	Terry


>
>Regards,
>Malcolm
>
>