[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Double Throw Spark Gap (fwd)
---------- Forwarded message ----------
Date: Sat, 13 Oct 2007 23:46:38 -0500
From: Crispy <crispy@xxxxxxxxxxx>
To: Tesla list <tesla@xxxxxxxxxx>
Subject: Re: Double Throw Spark Gap (fwd)
Thanks a lot, Bert. That pretty much sums up everything I've been
saying. Sometimes I'm not the best at getting across ideas, but that
was great.
I'd just like to add one note - you talk about the charging arc drawing
out due to the high charging current. Although this is completely true,
I don't think it's particularly relevant. The time it will take to
complete the charging cycle is, by my calculations, so short, that the
arc really shouldn't have to stretch at all.
Also, I have another question somewhat unrelated to this thread, but
goes along with the University limits. I obviously need an RF ground,
but I'm not allowed to pound anything into the ground. So, I need a
counterpoise. I have virtually no idea on what is acceptable here. Is
a 1ft x 1ft square covered in the middle with many layers of chicken
wire acceptable?
Thanks,
Chris B aka. Crispy
On Sat, 2007-10-13 at 21:58 -0600, Tesla list wrote:
> ---------- Forwarded message ----------
> Date: Sat, 13 Oct 2007 22:38:03 -0500
> From: Bert Hickman <bert.hickman@xxxxxxxxxx>
> To: Tesla list <tesla@xxxxxxxxxx>
> Subject: Re: Double Throw Spark Gap (fwd)
>
> Tesla list wrote:
> > ---------- Forwarded message ----------
> > Date: Fri, 12 Oct 2007 22:36:08 -0700
> > From: Barton B. Anderson <bartb@xxxxxxxxxxxxxxxx>
> > To: Tesla list <tesla@xxxxxxxxxx>
> > Subject: Re: Double Throw Spark Gap (fwd)
> >
> > Hi Bert,
> >
> > I don't think so. Are you saying your going to charge a 20nF cap in 69us
> > from a 12/30 NST source? (or am I misunderstanding your derivation?).
>
> Not exactly - I assumed that Chris first fully charged a 400 nF DC
> storage cap (Cs) to 17 kVDC and then used Cs to quickly resonantly
> charge the 20 nF cap through the dequeing diode and charging inductor to
> 2*Vc. Cs can quickly transfer a portion of its energy to the tank cap
> (Cp) - this is the source for the high charging current, not the NST.
>
> > I realize the charging inductors peak current is V/sqrt(L/C), but that
> > does not account for the impedance of the transformer. You've got 360VA
> > to play with and nothing you do in the charging circuit is going to
> > change that.
>
> I agree - the average power (over an extended period) can only be 360
> VA. Chris's HVDC supply uses a 12/30 NST, a full wave bridge, and a 0.4
> uF energy storage cap (Cs) (made up a a batch of electrolytic caps in
> connected in series). If Chris can temporarily store up energy (in Cs),
> then operate the coil in short high power bursts, he should be able to
> obtain the higher power bursts (and longer sparks) he wants without
> exceeding the university's current limit. The shorter the duty cycle,
> the higher the peak power available during bursts.
>
> > The charging inductor and cap will require time, and with a
> > 12/30 is going to require a lot of time. In a typical 12/30 system, a
> > 20nF cap will require 40ms to fully charge. The charging inductor is not
> > going to change that "unless" you have a non-limited source of current
> > (which you don't).
>
> Although Cs can't supply unlimited current, it can briefly deliver
> considerably higher current than the NST. Cs is not an ideal storage
> cap, since will it a has high ESR, but it should work as long as the
> caps can survive the high "ripple" current involved with recharging Cp.
>
> During the brief tank cap charging cycle, Cs supplies the 14+ ampere
> high current pulses necessary to charge (and voltage double) Cp via the
> charging inductor (Lc). Cs gets (at least partially) replenished from
> the NST/FWB during the dead time between tank cap charging pulses. The
> NST and FWB rectifier supply the DC into the following circuit (where Lp
> = 25 mH, Cs = 0.4 uF, and Cp is about 0.016 uF:
>
>
> Static |\ |
> + ----o----o o----| >|-----o o---------
> | Gap |/ | \ |
> | DQ A \ B O
> HV + | o O Lp
> DC In ----- | O
> ----- Cs | | | |
> | 0.4uF -----| |----o
> | | | |
> | Lc Cp |
> | ======= |
> - ----o-----------OOOOOOO--------------------
>
> Chris proposed using a static spark gap to (theoretically) allow the
> coil to operate in short bursts separated by recharging of Cs. In
> operation, Cs charges to a sufficiently high voltage to cause the
> initial ("cold") breakdown of the static gap and ARSG charging gap. The
> combination of Cp and Lc "ring", eventually leaving Cp with a voltage
> that's twice the voltage across Cs. During burst mode, the static gap
> will (hopefully) retrigger over a number of tank capacitor recharge/fire
> cycles. Each successive charging cycle transfers a portion of the stored
> energy from Cs into the tank cap. If the break rate is high enough, the
> NST can't fully recharge Cs between successive bangs, and the voltage
> across Cs declines. After some unknown number of TC bangs, the voltage
> declines to the point where the static gap can no longer reignite, and
> Cs recharges back to the previous "cold" breakdown voltage of the static
> gap and ARSG - sort of like a relaxation oscillator. If this approach is
> successful, Chris should get a burst of higher power sparks that are
> longer than he could obtain by simply driving the system from the NST alone.
>
> >
> > If I understand this scheme correctly, the charging connection is
> > limited by the dwell time of the electrodes.
>
> Sort of. Except that, once the charging gap in the ARSG fires, it
> becomes quite difficult to break the current flowing through the
> charging inductor. The relatively high charging current permits the
> formation of a heavily conductive arc (~13A peak in his system) which
> simply continues to conduct, stretching out as the rotary gap separates.
>
> The charging arc continues until the half sine wave charging current
> drops to almost zero (probably at less than an amp). At lower arc
> currents, the voltage drop across the arc rises to the point where it is
> comparable to maximum source voltage from Cs (and Lc*di/dt) and it
> finally breaks. The point is that effective "dwell time" of the ARSG
> lengthens so that it is always adequate to complete the resonant
> charging cycle for Cp and Lc. The faster you spin the ARSG, the longer
> the trailing arcs become on the charging contacts so that the effective
> dwell stays relatively constant...
>
> > That adds considerable time
> > to the situation (it's worse than a typical AC system). There's
> > something else which complicates the charging mechanism. If the rpm is
> > slow, the dwell time per electrode is extended (charge will reach
> > voltage with less passes). Increase rpm, and it will take more passes to
> > reach breakdown desired). But, here's the gotcha, typically RSG BPS is
> > based on timing. In this case, the gap distance becomes critical with
> > rpm. It's an odd thing to say the least.
>
> The arc stretching in Chris's system should permit virtually full
> charging (to 2 times the voltage across Cs) during a SINGLE charging gap
> presentation whenever the series static gap successfully fires. Since Cs
> is over 20 times larger then Cp, and Chris is using a comparatively low
> HV power source, once the static gap fires, he should obtain some
> (unknown) number of sequential bangs, although each bang may be smaller
> than the previous one at higher break rates.
>
> >
> > In Steve's case, he has a hefty transformer (14/250) at 3500VA. He's
> > using 12 electrodes at 1850 rpm with a 50nF cap. His dwell time is 470us
> > per electrode and 5.63ms per revolution. When you factor in the
> > impedance of the transformer, full charge time is 64.3ms. This will
> > require 11.4 revolutions to fully charge the cap. He's running at 370
> > ppr, but if he was to actually fire at 30KVDC, the actual firing bps
> > would only be 32 (by my calculation). He is likely firing at about
> > 17KVDC which is in the low 100 bps range and likely brought about by gap
> > distance, electrode edge surface, etc.). Gap distance is going to play a
> > significant role in these SPDT gaps.
> >
> > In Chris's case, there's a huge difference (3,140 VA).
>
> I agree. That's why the system Chris is proposing only runs for short
> bursts (working mainly off stored energy in Cs), then stops to recharge
> Cs, etc. Using a high voltage relay (instead of a static spark gap)
> would make overall operation straightforward and (perhaps) more
> controllable...
>
> >
> > Take care,
> > Bart
> >
>
> Hope this helped, and best wishes,
>
> Bert
> >
> >
> > Tesla list wrote:
> >> ---------- Forwarded message ----------
> >> Date: Tue, 09 Oct 2007 23:10:39 -0500
> >> From: Bert Hickman <bert.hickman@xxxxxxxxxx>
> >> To: Tesla list <tesla@xxxxxxxxxx>
> >> Subject: Re: Double Throw Spark Gap (fwd)
> >>
> >> Hi Adam,
> >>
> >> I suspect the tank cap will charge almost completely. Let's plug in some
> >> reasonable "guesstimate" numbers and see...
> >>
> >> Chris didn't indicate the size of his tank cap, so let's take an
> >> educated guess that it's 0.02 uF. We can always change this to the
> >> actual cap size Chris is using later if necessary. The starting voltage
> >> on the DC storage cap (when driven by a 12/30 NST and FWB rectifier)
> >> will be about 17,000 volts. With a 25 mH charging choke, the resonant
> >> charging frequency of the combination of the charging choke (Lc), tank
> >> cap (Cp), and DC storage cap (Cs) will be about 7300 Hz. If allowed to
> >> fully complete, the charging current will be a half sine wave lasting
> >> for 1/2 cycle, or about 69 usec, peaking at about 14.5 amperes. For
> >> complete resonant charging cycle, we must see a minimum effective dwell
> >> time of about 69 usec in th charging gap.
> >>
> >> Now, when the charging gap initially fires, the rotating contacts are
> >> approaching each other. Let's assume that the gap fires when the
> >> approaching electrodes reach a minimum separation of 0.25", the gap
> >> electrodes are 0.125" in diameter, and the DC charging current arc can
> >> be drawn out at least another 0.375" before being extinguished. YMMV.
> >> With these parameters, the estimated total distance over which the
> >> charging current would be connected (via arcing) would be about 0.75".
> >>
> >> Let's further assume that the ARSG rotor speed is 3000 RPM, and the
> >> electrode radius of rotation is 4" (8" diameter for electrode rotation).
> >> The effective "dwell" angle is about 10.7 degrees, and at 3000 RPM this
> >> translates to about 42 usec of total effective dwell time versus 69 usec
> >> to fully charge the tank cap. Does this mean that the gap will
> >> prematurely interrupt the charging cycle?
> >>
> >> I suspect the answer is - NO.
> >>
> >> Since the peak charging current is a robust 14.5 amps, I suspect that
> >> (at least for the relatively short arc distances we're discussing) the
> >> gap will be unable to interrupt the charging cycle until the charging
> >> current has dropped significantly - perhaps down to an ampere or less.
> >> Instead of prematurely terminating the charging cycle, I suspect that
> >> we'll see trailing arcs in the gap that effectively "stretch out" the
> >> dwell time so that the tank cap can achieve nearly 2 x Vsupply.
> >>
> >> It will be interesting to see the actual results that Chris sees... :^)
> >>
> >> Bert
> >>
> >> Tesla list wrote:
> >>
> >>> ---------- Forwarded message ----------
> >>> Date: Tue, 9 Oct 2007 03:46:42 -0700 (PDT)
> >>> From: Yurtle Turtle <yurtle_t@xxxxxxxxx>
> >>> To: tesla@xxxxxxxxxx
> >>> Subject: Double Throw Spark Gap
> >>>
> >>> I haven't really been following this thread, so
> >>> forgive me if this comment has already been addressed.
> >>>
> >>>
> >>> Regarding the following circuit:
> >>>
> >>> http://tangent.cluenet.org/~chules/hv/tesla/dtsg.html
> >>>
> >>> Does everyone think the cap can charge adequately in
> >>> the same amount of time it can discharge? Without
> >>> doing the math, I gotta believe that if one set of
> >>> flying electrodes zooms around at several hundred rpm,
> >>> the presentation time for the charging portion will be
> >>> too short to fully charge the cap. After all, most rsg
> >>> coils are charging the whole time they aren't
> >>> presenting (well not exactly). I guess you gotta know
> >>> the dwell time and the size of the pig feeding this.
> >>>
> >>> Adam
> >>>
> >>>
> >>>
> >>> ____________________________________________________________________________________
> >>> Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games.
> >>> http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow
> >>>
> >>>
> >>>
> >>>
> >>>
> >>
> >>
> >
> >
> >
>
>