[TCML] Re:Pros and cons

[uhvsystems@xxxxxxxx]

Jim,
 
Breakdown voltages and sparks in nonuniform gaps obey normal (gaussian) distribution law.It is established standard deviations tend to increase as voltage pulse amplitude  and gap lenght increases.Furtherly,experiments indicate sd. is higher for nonuniform gaps contaminated with floating charge than for gaps with relatively clean air.1.2 MV positive SI with near optimum time to crest at 50% probability level breaks aproximately 4.5 m long nonuniform air gap.If the gap gets contaminated by charge sd~12-16% (for this to occur it is enough to have only few impulses per second,not necessarily 100 or more) in the sample of 10000 discharges ,purely on statistical basis, one discharge more than  7.5 m long can be expected.I don't know  sd. for high frequency tesla transient sparks,but if the PPS rate is high enough and leader effect brings the thermodynamical  *inertia* of channel in the mix, sd must be substantial .Two good official references about SI gap breakdowns and related insulation co-ordination are  IEC 60071-1:2006 i IEC 60071-2:1996.
 
 
 
Dex,
My guestimate for voltage range (100-200 kV) where spark growth effect is seriously degraded is a very rough one.Generally we differ glow and streamer corona,and leader corona.The estimated range  is not a wild guess becouse for monopolar impulse leader corona to be excited requires aproximately twice as much voltage.SGTC up to 100 kV are able to generate only streamer corona.That doesn't mean the spark growth is impossible at 100 kV,but rather to get say 2x longer spark in comparison with single shot mode spark lenght , SGTC ought to be operated at extremely high PPS rate (for example, 2000 PPS instead of usual 100-200 PPS).
 
 
Regards,
Fez Zaev      
 
 
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
 
           
jimlux <jimlux@xxxxxxxxxxxxx> wrote:

uhvsystems@xxxxxxxx wrote: 
> John, Good points.I'm not expert on this but I have impression that 
> the concept of one and the same PPS rate which would be ideal for all 
> TCs is an illusion.By "ideal" I mean the best one for power 
> input-spark lenght relationship.Too many things affect spark 
> growth.From the geometry (top load) and size of the coil, output 
> voltage peak,voltage wave form and frequency to enviromental and 
> atmospheric cicumstances. 
 
I sort of agree. As a practical matter, tesla coils tend to have certain popular design sizes, based on the availability of HV transformers (may be less so for solid state), so for a "neon sign transformer" powered coil, of any configuration, there's probably an optimum pps, just because there's not a huge design space in terms of physical configuration that works at all. Likewise for pole transformer based designs, which tend to run at 14-15kV rms and a few kilowatts (here the practical limit might be the mains power supply.. it's inconvenient to wire up something for more than about 50Amps at 240V) 
 
> > Dex, 
> > I'm 99.99% certain there's no SGTC ( at sea level altitude) which can 
> generate 8 m long sparks at 100 impulses per second peaking 750 
> kV.Perhaps,in a long run this is possible with 1000 PPS or more but 
> I'm not sure about that.OTOH,I'm positive that TC peaking 1200 kV at 
> 100 PPS has a high probability ,in a long run,to break 8 m long gap! 
 
The latter is not much different from a transient on a EHV line triggering a long spark, and those, while not common, have certainly been observed. To a certain extent, the voltage at the electrode from which the spark starts isn't much different (in terms of spark growth) whether it's 100 Hz or 100Hz modulated 100kHz. 
 
> > Jim, 
> > 200-300 kV is a very interesting remark.This is a voltage range where 
> first appearances of so called *leader corona* and *leader flashes* 
> are experimentally observed.That should have consequences to the 
> behaviour of the sparks in HV impulses machines like tesla coils.To 
> illustrate,if 250 kV SGTC is in position to prolong sparks 3x their 
> single shot lenght at typical PPS rates,with 100 kV and 
> corespondingly smaller impulse energy it wouldn't do much better ,at 
> typical break rates,than about 1.3x spark lenght in 100 kV single 
> shot discharge. 
> >  
In fact, lower voltage coils,like the very small ones using a electromechanical interrupter (buzzer), tend not to have long sparks, but more of the brushy discharge or lots of shorter sparks, say, 10cm long. So your observation is interesting about leader corona. 
 
The maximum voltage at the top of the tesla coil is almost certainly limited by the radius of curvature of the electrode, and empirical evidence suggests that toroids are better than spheres, even though the max voltage is less for the same overall size. 
 
 
 
Dex Dexter <dexterlabs@xxxxxxxxxxx> wrote:

I guess people working with drsstc ,like Steve W.,may want
to check this theory about small voltage (200kV<) and limited
spark growth.For example they can set primary energy such
that in loseless transfer secondary is brought to  potential
of 200 kV.The voltage in a real world will be lower due to
loss.Sharp breakout point should be attached to toroid to let 
sparks form.Comparison of spark lenght between s.s. bursts and 
100,200,300 BPS runs will give us idea about general behaviour
of "small" output voltage coils.I'd like to check this myself,
but I don't have drsstc.My 5kv/20mA NST powered table-top coil
can't run  s.s,and I don't know exactly primary voltage and
energy (funny , I don't know even power input since ammeter
goes reading-crazy when the coil fires).

Dex       
 

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla