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

Pulsed Tube Coil Work (with updates)

 All,
 
 Here's a brief overview of the pulsed tube coil work I've done:
 
 1) standard tube TCs:  These are pulsed in the sense that the
 vacuum tube rectifies the AC thereby causing the coil to run only 
 on the positive AC half-cycles or about 1/2 the time.  The oscillator 
 starts and stops 60 times per second.  Spark appearance ranges
 from stabbing points to tangled network of writhing sparks.  Longest 
 sparks obtained were 36" at 9000 watts, but I later obtained 35" at 
 5600 watts.    (update; obtained 38"+ sparks, then a tube arced
 and was destroyed)
 
 2) Filtered DC power supply tube coil, steady running, not pulsed: 
 Achieved up to 7" spark.   Spark is wide and short and hisses, 
 looks like a plasma flame, no individual sparks.  Draws a lot of 
 power.
 
 3) Pulsed AC tube TC:  Tried gating the grid leak network so the
 coil would run only near the peak of the AC cycle, but if the "on" 
 time was less than 2 milliseconds per 60 Hz half-cycle, the spark 
 was dim and short.  This method increased spark output efficiency 
 by about 13% but in general, the results were very disappointing 
 and caused a lot of destructive kickback.
    
 Next, I let the coil run for one full AC 60 Hz half cycle, followed by
 multiple AC cycles with the TC oscillator "off".  I used a -400 VDC
 voltage, controlled by a solid state timer circuit to disable the TC
 for any desired number of AC cycles.  The -400 volts was applied 
 to the connection point between the grid coil and the grid leak 
 network.  In general, this method worked very well, and gave slightly
 longer sparks than the standard tubecoil.  The pulsed operation 
 resulted in a much lower input current and cooler tube operation. 
 It was easy to increase spark production efficiency by 400%
 or more.   The spark took on a straight sword-like appearance in
 many cases, and produced a rapid-fire "staccato" type sound
 which varied as the timing of the control circuit was varied by
 potentiometer.  Efficiency is high, but a large power transformer 
 is needed to pass the heavy currents during TC "on" time.
    (update;  I now use a triac, SCR, or thyratron between the
 tube cathode and ground which is used to disconnect the cathode
 from ground and thereby disable coil operation during staccato
 "off" times.) 
 
 4) Pulsed, filtered, DC tube coil:  First I just applied a control 
 pulse to the grid-circuit point as explained above, but this did not
 work at first because unlike in the AC tube coil, the control pulse
 now had to TURN off the oscillator while it was running at full power,
 rather than simply keep it off, from a previous negative half cycle 
 (as in the AC system) during which time it was off anyway.  It 
 takes a lot more negative grid voltage to turn off the coil when it's
 running--maybe -1500 volts or so.  Thus, I used a mechanical 
 motorized rotary switch to control the TC.  This pulsed the
 system, but the sparks were short, weird,  and thick, at low 
 interupt rates, and long "on" times, and at faster interupt rates 
 and shorter "on" durations, the sparks were still short (but longer 
 than steady DC sparks).  Since I was using a zero-cutoff tube, I 
 simply used the mech. switch to un-ground the grid leak network  
 to disable the TC during "off" times.  This mechanical
 interupt system can, by the way, be used on a pulsed AC tube
 coil and works very well, by speeding and slowing the motor, the 
 system sounds like a gas engine revving up!  But the problem in 
 the DC system is not the mechanical interupt.
     
 To obtain long sparks from a pulsed DC tube Coil, it may be 
 essential that the TC oscillator starts up slowly and gradually*.   
 This occurs automatically in a pulsed AC tube coil (or regular tube
 coil), as the osc. power builds in time along the slope of the 60 Hz
 AC sine wave.  In contrast, in the DC pulsed system, the oscillator 
 is forced to start instantly at pulse turn on.  This fast start up 
 seems to cause short,thick sparks.  It is possible that the fast turn
 on drives the oscillator off frequency, or maybe the effect is related
 to some aspect of the ion-cloud creation and movement,
 or maybe the effect has some other cause that I have not thought of.
 
 Anyhow, to solve this problem, I used a tetrode tube for the main TC
 oscillator and fed a source of 600 volt, 60 Hz AC, through a dropping
 resistor, to the screen of the vacuum tube.  So the plate received
 pure DC, the screen received 60 Hz AC and the grid was connected
 in the traditional tube TC fashion.  Results were now excellent--the 
 oscillator starts gradually under the influence of the screen grid AC 
 and the sparks appeared just like the AC pulsed tube coil sparks.
  
 Since the system uses filtered DC, it is possible to use a small, 
 weak, transformer which slowly charges the filter capacitor over
 multiple AC cycles during TC "off" times.  Then, when the filter 
 capacitor is fully charged, the TC is turned "on" and the tremendous 
 stored energy in the filter capacitor is instantly available to the TC 
 which produces a giant spark pulse which partially depletes the filter
 capacitor's charge and the cycle begins anew.  I use about 3 
 pulses/sec which appears reasonably steady to the eye.

 The potential efficiency advantage of this design is enormous.
 My 35" spark, steady AC tube coil uses 5600 watts.  Using DC 
 pulsed operation, a 35" spark should require only about 1000 watts
 or less.  This approaches the efficiency of many spark-gap TCs.  I 
 have not yet tried this technique with the "35-inch-spark" tube coil,
 but I did try it with an "18-inch-spark" tube TC.  Power consumption 
 decreased from 1400 watts to 140 watts, and I was able to replace
 the original 800 watt transformer with a 135 watt transformer.
 AC pulsed or "staccato" operation will of course give the average
 input power reduction benefit, but will still require the heavy
 transformer.
   
   *(update; I think it might be the long full-powered "on" time that
 causes the short sparks with DC staccatto, rather than the fast 
 power ramp-up, but I haven't tested this) 

  Note:  In the above report, I am defining efficiency in a practical
 "coil builders" sense as input power versus spark length.

 There may be other radically different ways to build efficient
 pulsed tube TCs that I have not tried.  I feel that tube coils have
 been "under-explored" over the years, and offer intriquing 
 challenges as well as opportunities. 

 I also did some work with tube magnifiers, tuned secondary tube
 magnifiers, use of ferrite to increase secondary inductance, use
 of toroids, etc.

 John Freau