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

Re: X-mas Experiments



Richard,

Thanks for the update!! A whole bunch of us were extremely interested in
the experiments and results. The thyratron gap certainly has GREAT
potential for precise comparative measurements. My comments are
interspersed below...


Tesla List wrote:
> 
> Subscriber: hullr-at-whitlock-dot-com Tue Dec 31 22:31:57 1996
> Date: Tue, 31 Dec 1996 11:10:12 -0800
> From: Richard Hull <hullr-at-whitlock-dot-com>
> To: tesla-at-pupman-dot-com
> Cc: caydsi-at-aol-dot-com
> Subject: X-mas Experiments
> 
> All,
> 
> I have just come into the office to stamp out a small fire.  I note 331 E
> mails!!!  I can't read them all today.  I will report breifly on the
> result of all my approx 41 hours of experiments during the last week and
> a half.
> 
> conditions:
> 1. My standard 500khz small Tesla system set up as on prior occassions.
> primary tank presented a 38 ohm surge impedance (.008ufd -at- 12uh)
> 2. Used a DC supply reported on before.  Impedance of supply 60k ohms
> (deliberately current limited). supply set for 5kv prior to gap breakdown
> in all experiments.
> 3. I had three gap systems shunted across supply terminals. (switchable)
>         a. static series tungsten system of 5 series gaps.
>         b. Rotary gap 6 point set for ~ 240 BPS
>         c. Hydrogen thyratron set for <2 usec on time.
> 4. Toroid on coil with tungsten needle point protruding 2" out beyond
> shading field of toroid.
> 
> Measurement devices:
> 
> 1. 12" ball at range of 5' from coil attached to Keithley 610C
> electrometer to measure the DC or electrostatic component of the coil's
> output.  (note*  bandwidth below 1 khz)  Used as a coulombmeter only in
> these experiments.
> 
> 2. 4' length of wire (whip antenna) connected to a Tektronix TDS 210
> digital storage scope at a range of 8' from system.  This was to look at
> the emitted RF/EM waveform packet from the system.
> 
> Results of tests:
> 
> 1. The static series gaps quenched too fast and allowed a measured break
> rate of ~1200 BPS!!  Remember, DC is always there!  The electrometer
> showed the accumulated  Positive (+) charge of 10^-8 coulombs was reached
> in 12 seconds.  The O'scope showed a curious bulging of the ring wave
> prior to the normal high Q ring down.  This showed up as a wave "packet"
> of 5 cycles of ring up and down with a peak of 20 V p-p.  This was
> instantly followed by the classic secondary ringup and long ring down
> over many tens of cycles of the secondary.  This second action only
> attained 15 v p-p.  The DC, time averaged current from the supply was 16
> ma with the voltmeter falling to a time averaged level of 3.5KV.  Time
> averaged DC input power ( not necessarily total dissapation) 56 WATTS!
> Spark out of system to grounded rod, 5"

Sounds like you had one complete cycle of pri->sec->pri energy transfer 
followed by a second pri->sec transfer and _then_ a quench. The 1200 BPS
may be more indicative of recharge time and low gap breakdown/reignition
voltage...

Implications: Full energy transfer to the secondary WAS achieved, but
sec energy was then given back to the primary. The highest secondary
voltage peak was hit, in passing, during the first energy transfer.
Output magnitude was low due to low Vgap, but spark length good because
of high rep rate...

> 
> 2. Rotary gap was used to reduce break rate from between 120 and 480
> breaks per second.  (forced dead time for the supply) Speed was adjusted
> to produce the maximum DC charge received per unit time.  This turned out
> to be 120-240 BPS.  Electrometer accumulated 10^-7 coulombs in only 3-4
> seconds!  O'scope indicated two crisp and distinct ring wave packets
> prior to the normal high Q ring wave.  Each were of 5-6 cycles with the
> first being over 40v p-p, immediately followed by the second of 35v p-p
> and immediately by the high Q wave of only 15 V p-p. (voltages are the
> peak readings, of course).  The power supply read 4kv and 5 ma time
> averaged. (20 Watts).  Spark out of system to grounded rod ~4".
> 

Higher initial energy per bang than above, but poorer quenching, with 2
complete pri->sec->pri cycles before final pri->sec transfer and quench. 

Implication: Full energy transfer from pri->sec occurred (in passing),
and peak energy was probably 4X that above (assuming proportional to
Vout^2).

> 3.  The hydrogen thyratron circuit was next used.  It was set up to allow
> a pulse repetition rate of 200 bps.  These pulses lasted only about 1.5
> usec.  Thus, the gap shorted the primary cap to the primary for only
> under 2 usec.  This is an incredible feat compared to our air gaps.  The
> coulombmeter drifted over many minutes of operation toward the positive
> direction but would also drift back, too.  The overall impression was
> that no real charging was taking place.  The desired goal of 10^-8
> coulomb was never even half reached after 3 minutes of operation!  The
> O'scope waveform was "TEXT BOOK"!!  An instantaneous ring up (1 cycle)
> and a long and smooth ring down over many cycles with a peak voltage
> recorded of 6V p-p.  The time averaged readings were 5Kv and 10ma.  (50
> Watts.)  the spark out to a grounded rod.  1.5"
> 

Incredible micro-quenching feat, Richard! It looks like quenching was
done at the first primary current zero-crossing, before all the primary
cap's energy could transfer to the secondary. Only a small fraction of
the primary cap's energy actually got transferred to the secondary.
Based upon the previous measurements, increasing the thyratron's ON time
to permit complete energy transfer should significantly improve output,
and more closely match conditions of your second experiment. 

> The precision Pearson current tansformer indicated the exact same tank
> current for the primary ring up of all gaps, showing each to be doing
> about as good a job of transferring the energy from the cap to the
> primary coil as any other.  (~80-100 peak amps).  Only the Thyratron
> indicated a smooth none variant action at 80 amps.  (controlled gas and
> conditions of break).  Air gaps did show occasional higher peak pulse
> energy supplied in the 100 amp range.  Overall there was little
> difference in tank current, though.
> Conclusions.
> 
> 1. The common old neon sign transformer the the preferred power source
> for any TC where the production of DC electrostatic energy is
> required!!!
> 
> The reason for this statement is that the result of the 1000+ BPS was
> that of holding back the DC charging rate.  The "agents of charge"
> appear to require a "flee time".  This bodes well for the ionic
> production scenario.  The AC supply will naturally allow the ideal
> 120-240 breaks per second required due to natural dead time of the AC
> sine.  If DC is used a forced break (rotary) is required to make DC.  The
> electrodes on the rotary must be small due to the need for not rapidly
> refiring during presentation.  Thus, a rotary with a DC system should not
> "present" longer than 500usec.
> 
> 2.  The system absolutely relies on multiple, forced, interactions of the
> primary and secondary fields.  The first "knee jerk" reaction is to state
> that sloppy quench is a must!!!  This may not be totally true and even if
> it is, long dwell times might hurt the reaction.  Much more investigation
> is needed here to pin down the ideal interaction period.

Certainly, the first full pri->sec energy transfer must be allowed to
complete, so that the peak secondary voltage can be obtained. While this
occured in the first two experiments, it did not in the third. Longer
"dwelltimes" (in increments of energy-transfer times) with the thyratron
gap should precisely nail down the answer to this one.

> 
> 3. Longest spark is not a direct function of input energy!!!  A key
> point!!!  I have known this qualitatively for years and screamed it out
> all too often!  It is shown here under controlled conditions.  The
> biggest power hog was the thyratron and it exhibited the least spark.
> Once again, (I have screamed this before, too), the ideallic LC
> oscillatory exchange, the perfect radioman's/RF/EM waveform, is a sure
> indication of not only lousey spark production but of poor DC
> electrostatic performance as well.  Text book stuff is a trip "downhill"
> where sparks and DC output are concerned.  The HI-Q ring part of the
> system might not even be invloved in what we coilers are after at all!!!
>  In all cases above, the HI-Q portion of the wave was grossly reduced in
> peak amplitude over the initial "wave packet".

Very true!  However, I don't believe you achieved an ideal LC
oscillatory exchange in the third experiment. Since your coupling
coefficient is relatively low, there is simply no way to transfer the
all of the primary cap's initial energy into the secondary more quickly
than 5 - 8 half-cycles. At Fr = 500 kHz, this implies a dwell time of at
least 5 - 8 uSec. In the  third experiment, lousy spark production AND
poor electrostatic performance  have the same root cause - low energy
transfer to the secondary. Increase the dwell time and I bet you'll see
a _dramatic_ performance improvement. You can be shorter than the ideal
time and see pretty good results, but 1.5 uSec was much too short!

> 
> 4.  It appears that there is no extant artifice which will reverse the
> sign of the accumulated charge.  It was positive (+) inspite of over 22
> iterative connections of the primary tank,  power supply polarity
> connection,  power suply/ground linkage,  resoantor base/tank linkage,
> DC/Ac supply usage, etc.  I just couldn't acquire a negative charge under
> any circumstance!!!  Please, someone show me it can be down!!  I have
> exhausted my possibilities here.

This tells us the effect is independent of output peak polarity, and is
only a function of the charge-carrier mechanism. This is still the
biggest Mystery!!

> 
> 5.  The continual reading of negative voltage on the electrometer while
> recording positive charge remains a mystery.

This is the 2nd biggest mystery! I think you are correct about the
leakage current being the difference, perhaps in combination with some
imbalanced induction effects?? 

> 
> All of the above is in visual format at the end of our report tape #55.
> I had a visit on Sunday from Richard Wayne Wall.  He is a very nice and
> intersting fellow who is commited to undestanding this too.  A lot of
> other research was down regarding dielectrics, shileding an such which
> may be reported later.
> 
> I'll be back on the 2nd.
> 
> Richard Hull, TCBOR

Welcome back, safe coilin', and Happy New Year to you!

-- Bert --