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

Re: [TCML] still stuck at 36



Hi Bart,

Full energy transfer occurs at the first notch. So if you quench at the first notch, that's the ideal time to quench. If it takes 6 half cycles to do a full energy transfer, then the first notch will be at the 6 half cycle point. I'm
saying that the gap will never quench at the 3rd half cycle which would
be at a point well before the first notch. I'm saying the coil will never
quench before the 1st notch, unless a ground stike happens maybe,
but that's a special case.  Usually small coils quench at the 3rd notch
or so.  This 3rd notch is not before the 1st notch, it's well after that
time.  Now let's look at how many half cycles occur by the 4rd notch
for a poor quench situation.

If there are 6 half cycles to the first notch, then there are 18 half cycles to the 2nd notch, and 30 half cycles to the 3rd notch, and 42 half cycles to
the 4th notch.  Now if the coil runs at 100kHz, then each half cycle is
5uS.  Now assume a poor quench at 4th notch.  5uS x 42 half cycles
= 210uS.  This is a considerably longer duration than the 30uS duration
to the 1st notch.

In the example you gave of the 50hHz coil, it's true the full energy
transfer would occur in 63uS.  But why do you think the gap would
quench at that time or earlier if it's a poor quenching gap.  Only
the best gap might possibly quench at that 63uS which is also the
first notc
h.  But if that coil has poor quenching, then it wil quench
much later than the time of full energy transfer.  When this happens,
some energy comes back from the secondary to the primary and
keeps the gap lighted, then transfers back to the secondary and
keeps making these round trips until the gap finally quenches at
perhaps the 4th notch in a poor quench situation.  Now at a
4th notch quench, the spark duration is actually 512uS.  If the
cap charges in 20.75mS, then the ratio is 40.53.  This is about
2.5% which is considerably larger than the 1% charge time that
you mentioned.  And this is the point I was making.

Your analysis assumes 1st notch quench or sooner.  I'm saying
the quench never occurs before the 1st notch.  I'm also
saying that in a poor quench situation, the quench will occur
much later than the 1st notch (perhaps the 4th notch).

Now with a smaller coil that runs at 100kHz, and 300bps,
and assuming the same charge time you mentioned, then
the ratio is 81.06 which is still more than the 1% figure you
mentioned.  I agree that the cap charging issue is not a large
factor with the poor quench.  Since we want to get the
details right, I wrote this to give the correct math and show
the way to calculate the durations to these various quenching
notch-points (3rd notch, 4th notch, or whatever)

Cheers,
John
--


-----Original Message-----
From: bartb <bartb@xxxxxxxxxxxxxxxx>
To: Tesla Coil=2
0Mailing List <tesla@xxxxxxxxxx>
Sent: Wed, 28 Jan 2009 10:24 pm
Subject: Re: [TCML] still stuck at 36


Hi John, 
 
When I mentioned quenching "early", I should have "detailed" my meaning. I was describing early as "before the full transfer time expires", not that one would actually quench "early". Quenching early should always have meaning as quenching prior to the maximum number of cycles for 1 complete "full" energy transfer. If we have 6 half cycles for full energy transfer and we quench on the 2nd primary notch (3rd half cycle), we quenched early (meaning not all the energy in the cap was transferred before the gap stopped conducting). I believe in most cases (not all), the gap will usually quench when nearly all the energy is transferred (should average typically 3rd notch due to the norm of coil sizes and coupling). 
 
Regarding lower RF frequency coils. Say a 50kHz coil using 10kva pig and large 200nF cap (big coil, low Fr, powerful). Total energy transfer time is 63us. At 300 bps on a rotary gap (will need a rotary for this large coil), charge time would be 20.75ms. This is a ratio of 329 which is still very large (1% range for transfer time affect on charge time). 
 
But if we assume a very small C (say 20nF) used with this large low RF coil, then 69us transfer time and 2.075ms charge time under the same conditions. This ratio is 30. At this point, yes, t hen it becomes an issue and starts to eat away at charge time. But this is an extreme and not the normal coil. However, if any coil was extremely underpowered and/or really high bps, all bets are off (low RF or not). 
 
In any event, I agree with your summary. Just as Bill has found that a change in the gap has increased his spark length, I think all of us who have been down this road have found the same. The greatest improvement to spark length I personally can attribute to, has always been spark gap efficiency. Processing power across the gap with minimum losses is the goal for a performing coil. Bill's coil is yet again another proof of this. 
 
Take care, 
Bart 
 
futuret@xxxxxxx wrote: 
Hi Bart, 
 
You're correct that the details are important. 
 
At lower RF frequencies, and if the coil quenches on for example the 
5th notch, and if the break rate is lets say 300 bps, the 1000x
situation 
won't hold. The total time the gap is lit during one bang could be 
200uS. And if the cap has 3000uS to charge between each bang 
then it's a 15X ratio which is somewhat significant. But quench time 
is not a huge factor, that's why I said the quenching only needs to 
be adequate (not excellent). 
 
What often occurs if there's not enough airflow is that the gaps 
will run hot. T
his will cause the gaps to fire at a lower voltage and 
greatly reduce the bang size, and increase the break rate. This 
greatly reduces the spark length. This factor is what hurts the 
spark length the most; the reduced firing voltage when the gaps 
heat up. And if the gaps heat even more, then true 60Hz power 
arcing can occur which completely or partially prevents the cap 
from charging. 
 
Regarding the early quench scenario... it never happens. It's >
impossible. 
If the gap fires, the best it will quench at is first notch (never >
sooner), 
due to air flow. When there's too much air flow, the gap won't quench early,  instead it will simply fail to fire at times. This is a point that is often misunderstood. 
 
I agree that the gap resistances don't increase the transfer time
very > much, 
and are not a large 
factor in gap losses. Gap losses are more due to having 
multi-gaps. 
 
I go back to what I said that a single gap has less losses than a 
multi-gap, but a single gap is harder to quench and harder to keep 
cool, so it needs more airflow. 
 
To summarize again: 
 
1) Poor quenching can reduce the time available for cap charging, 
but is a minor factor. Also I believe it's mostly the first transfer energy that 
dete
rmines the spark length, with subsequent transfers contributing 
little or nothing to the spark length. 
 
2) Hot running gaps will fire at a lower voltage and give a smaller 
bang size which reduces the spark length. This is a very large 
factor. (need more air flow) 
 
3) Even hotter gaps will power arc and prevent cap charging and 
kill the spark length severely. (need more air flow) 
 
3) Too much air does not cause a too-early quenching. Too-early 
quenching is impossible. Instead too much air prevents gap firing 
at times. (need less air flow) 
 
4) Multi-gaps have higher losses and give shorter sparks, but 
quench more easily, and run cooler with less air needed. 
 
So there are a number of issues to consider. But it's a logical 
progression to first select a low-loss gap (single gap), then 
cool it with air because single gaps run hot and may have 
trouble quenching or may start firing at a lower voltage if 
too hot. Also make sure there isn't so much air that the 
firing 
is interupted at times. A steady firing is needed. 
There's no need to worry about quenching too soon, 
that can't happen. 
 
John 
 
 
 
-----Original Message----- 
From: bartb <bartb@xxxxxxxxxxxxxxxx> 
To: Tesla Coil Mailing List <tesla@xxxxxxxxxx>0
Sent: Tue, 27 Jan 2009 8:44 pm 
Subject: Re: [TCML] still stuck at 36 
 
 
Hi John, > > The single gap with low loss sounds reasonable. The idea
of quenching > sluggishly, well, some variation is possible, but the consequence is > not a cap charge issue. The charge time on a typical coil versus the > maximum energy transfer time is over 1000x. For my 4.5"D coil, a 12/60 > NST charging a .0188uF cap, charge time is 18.8ms for 120bps operation > versus a 17.16us for complete energy transfer (3.9 cycles at 230kHz). > If we quench slowly, the cap has all day to charge since the 17us > makes little difference (0.01% of charge time). > > A quench variation entails a quenching early scenario which would > leave a charge on the cap (and energy in the cap). The cap would then > charge to breakdown more quickly to the next due to the charge left on > the cap (the line between A and B just got shorter, bps increases, > spark length diminishes). But this also states that all the energy in > the cap did not make it through the transfer (this is why spark length > decreases and not so much the bps which was simply a result). I think
nearly all of the energy transfers, an 
d so much so, that the voltage left on the cap is insignificant in
the > same fashion as the ratio between cap charge time and energy transfer > time, and more so in a static voltage clamping gap as opposed to a > rotary. > > Energy transfer time does make a difference of course (energy and > time). But the gap cannot affect the energy transfer time > significantly by quenching early (but it can by preventing extensions > of the transfer time in an extremely bad situation). This would be due > to gap impedance. > > Energy transfer time is based on the mutual inductance and frequency. > For my 4.5"D coil, reactance as based off of the mutual inductance is > 123 ohms. If the gap was ridiculously lossy, it's resistance would add > to this reactance. In a completely theoretical case where the gap > resistance was equal to this reactance, the transfer time would be > double. It is still insignificant regarding cap charge time. However, > the em force enveloping the secondary would certainly be reduced (time > and energy), and this would affect energy stored in the top load. > > But now for the gotcha! Gap resistances are less than 1 ohm. So gap > losses are not affecting energy transfer times. I think it's all about > simple power losses at the gap that are the real cause of pain and > suffering with spark lengths which is why significant air cooling > drastically improves any static gap style (some styles are better than 0D 
others at accomplishing this). There is always the issue of too high
a > velocity at the breakdown point that blows out the spark. This is a > quenching early issue, which of course doesn't transfer "all" the > energy in the cap and thus we see spark lengths diminish and bps rise. > This is why some find using a variac on the air voltage is not always > best at 100%. > > Hope I'm making sense. Sorry for being long winded, but it's one of > those "devil in the details" subjects. > Best Regards, > Bart > > futuret@xxxxxxx wrote: >> I think what makes a hyperbaric gap perform well is that it's >> a single gap. This makes it a low loss gap. Once you have a >> low loss gap then all that's needed is to quench it adequately. >> Since a single gap is harder to quench than a multi-gap, this >> is why a good amount of air flow is needed. >> >> The air flow is >> also needed to cool the electrodes. If there's not enough >> air flow the gaps may run too hot and may begin to power arc, >> or quench very sluggishly. If a gap quenches too slowly, >> then less time is available to recharge the capacitor and >> output suffers. >> >> If there's too much air, then >> the gap may fire intermittently with a sort of stuttering >> action and this cuts the spark length too. It's not really >> over-quenching, it's simply erratic firing=2 
0in the first place. >> It's not that the gap quenches too fast,
rather it's that >> the gap does not fire consistently. >> >> John >> -- > >> >> - ----Original Message----- >> From: DC Cox <resonance@xxxxxxxxxxxx> >> To: Tesla Coil Mailing List <tesla@xxxxxxxxxx> >> Sent: Tue, 27 Jan 2009 4:58 pm >> Subject: Re: [TCML] still stuck at 36 >> >> >> >> It's not clear why to me either. My best guess is better cooling due >> to the >> high velocity air, and perhaps increased turbulence >> around the ends of the copper pipes. >> >> I know that most coils performance jumps in amazing quantity of spark >> output >> when the hyperbaric gap is used. I have never used it above 4 kVA as > a >> single gap, but did some experiments at 7 kVA with a dual series > > hyperbaric >> gap and it seemed to work as good as a small rotary. >> >> Perhaps looking at scope ringdowns with and without a hyperbaric gap >> might >> provide some clues. >> >> Dr. Resonance >> >> >> >> >> On Tue, Jan 27, 2009 at 1:51 PM, Lau, Gary <Gary.Lau@xxxxxx> wrote: >> >>> I don't disagree that hyperbaric gaps work well, but I'd like to be >> clear >>> on why. Too often on this List, the word "quenching" is used > > incorrectly. >>> My measurements showed conclusively that hype 
rbaric gap losses were 
lower >>> than in a sucker gap, but I made no attempt to measure
actual > > quenching. >>> Quenching is unrelated to gap losses and refers to how many pri-sec >> energy >>> trades occur before the gap extinguishes. In fact, I think that > > typically, >>> lossy gaps (at least multi-segment gaps) have superior quenching. >>> >>> So, have you seen evidence that actual quenching, as observed on a > > scope, >>> was superior with the hyperbaric gap? Was this with free-air > > streamers or >>> grounded (connected) streamers? Seems like a very difficult thing to >>> accurately compare between two systems. >>> >>> Regards, Gary Lau >>> MA, USA >>> >>> > -----Original Message----- >>> > From: tesla-bounces@xxxxxxxxxx [mailto:tesla-bounces@xxxxxxxxxx] > On >>> > Behalf Of DC Cox >>> > Sent: Tuesday, January 27, 2009 1:31 PM >>> > To: Tesla Coil Mailing List >>> > Subject: Re: [TCML] still stuck at 36 >>> > >>> > Bill: >>> > >>> > What was your maximum spark before you used the hyperbaric > sparkgap? >>> > >>> > Also, put a variac on the vac. cleaner motor, and fine adjustment >> of the >>> > speed will increase your output even more. Usually best output >>> > around 75% of the variac. >>> > >>> > I know they work great. They always increase your sec spark output >>> 20> dramatically which means we have been underquenching >>> > with most of our previous designs. >>> > >>> > Dr. Resonance >>> > >>> > >>> > >>> > >>> > On Tue, Jan 27, 2009 at 5:50 AM, Bill Bohn <b.jbohn@xxxxxxxxxxx> > > wrote: >>> > >>> > > I built the hyperbaric gap as described WOW!Now I get 44" I am > > using a >>> shop >>> > > vac for suction.I am using205 of the 6 strings of caps.I think I >> did >>> some >>> > > damage >>> > > >>> > > To my 5-1/4 coil.Now I have a 6.625" secondary wound 26.5" 1060 >> turns. >>> Top >>> > > load is 4.5x18 toroid and a 7" sphere.Seems best with the > primary >>> clamped >>> > > >>> > > At turn 5.The spark sounds somewhat eratic and choppy.IT will > > smooth >>> out >>> > > some by adjusting the spark gap closer at the loss of streamer > > length . >>> > > >>> > > What is the method of measuring BPS?IT doesn't seem to sound > > right.The >>> > > coupling between primary and secondary is 0.I am going to put > the > 6 >>> strings >>> > > of >>> > > >>> > > My mmc cap in but I think it will be best tuned at about turn 2 >> and >>> become >>> > > even more erratic. >>> > > >>> > > I am getting positive results and thank everyone for all the > > ideas20and >>> 2 >>> > > cents worth. >>> > > >>> > > _______________________________________________ >>> > > Tesla mailing list >>> > > Tesla@xxxxxxxxxxxxxx >>> > > http://www.pupman.com/mailman/listinfo/tesla >>> > > >>> > _______________________________________________ >>> > Tesla mailing list >>> > Tesla@xxxxxxxxxxxxxx >>> > http://www.pupman.com/mailman/listinfo/tesla >>> _______________________________________________ >>> Tesla mailing list >>> Tesla@xxxxxxxxxxxxxx >>> http://www.pupman.com/mailman/listinfo/tesla >>> >> _______________________________ ________________ >> Tesla mailing list >> Tesla@xxxxxxxxxxxxxx >> http://www.pupman.com/mailman/listinfo/tesla >> >> _______________________________________________ >> Tesla mailing list >> Tesla@xxxxxxxxxxxxxx >> http://www.pupman.com/mailman/listinfo/tesla >> > _______________________________________________ > Tesla mailing list > Tesla@xxxxxxxxxxxxxx > http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________ 
Tesla mailing list 
Tesla@xxxxxxxxxxxxxx 
http://www.pupman.com/mailman/listinfo/tesla ;
 
_______________________________________________ 
Tesla mailing list 
Tesla@xxxxxxxxxxxxxx 
http://www.pupman.com/mailman/listinfo/tesla ;

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