[TCML] SRSG "sputter"
phil at hvtesla.com
Mon Jul 12 01:41:44 MDT 2010
If the detection coil of wire is accurately placed where the stationary
electrodes are located, and the magnet positioned where the rotating
electrodes are, you will get a reasonably accurate representation that takes
in the SRSG's rotor position with respect to the motor itself, and this
tells you where the coil is really firing (see note 1) in relation to the
AC sine wave.
** It is not absolutely necessary to do it this way though (unless your a
nerd) - as most people just swing the motor around and that works just as
But it's nice to see and physically check that the chosen cap value in the
phase adjuster is actually working as intended. Using a photo-electric cell
positioned where the stationary electrodes are, and allowing the rotating
electrodes to trip it is far more accurate though.
Someone else, I think it is Finn, puts a magnet and a coil of wire around
the stationary electrode, and uses the passing electrode to disturb the
magnetic field. You would need to temporally substitute the Tu electrode for
a steel one though I think
** ( note 1)
Both methods still only give 'ball-park' settings as the spark will jump to
the advancing electrode and advance the timing anyway, but, importantly,
the phase shift movement is still relative.
From: "Joe Mastroianni" <joe at killerowls.com>
Sent: Sunday, July 11, 2010 11:29 PM
To: "Tesla Coil Mailing List" <tesla at pupman.com>
Subject: Re: [TCML] SRSG "sputter"
> HI Gary -
> Ok - some measurements available on YFrog
> <http://yfrog.com/13t2qj> This is the maximum phase shift I've measured.
> What you're seeing is the sine wave from utility power against a distorted
> triangular wave. This is the voltage between my household hot lead
> (black) and the household neutral lead (white).
> That triangle wave is measured from the point where the variac L output
> connects to the run cap and feeds into the motor relative to the neutral
> connection (white). So the motor is across the neutral and this point
> which I am measuring. Notice the maximum phase shift appears to be
> between 1.2 and 1.6ms (lets say 1.4). That would put the max phase shift,
> as measured from the power supplies, at 360*1.4/16.67 = ~30 degrees.
> Now, we could argue that the waveform from the power supply going into the
> motor may somehow not relate directly to the motor timing angle. Ok, I
> would kind of wonder why that might not be so, but for the sake of
> argument I have added these videos.
> I attached 2 rectangular neodymium magnets to 2 of the rotating tungsten
> rods on the SRSG wheel with wire ties. I then wrapped 20 turns of
> 26gauge magnet wire around one of the stationary tungsten electrodes.
> The first video shows 2 waveforms. <http://yfrog.com/0jrelmotdrivez>
> Channel A is the power feeding the motor, out of the variac/cap circuit.
> Notice how that waveform distorts as the variac L is progressed from "0"
> to about 1/2 way. Channel B is the voltage "blip" taken off the coil as
> the magnets rotate past. I think we can presume TDC is when the
> magnet/timing waveform crosses "0". I rotated the variac until the motor
> began to lose lock (you can hear it in the background, and see the
> waveform begin to oscillate). Notice the magnet pulses do not change
> phase with respect to the power supply to the motor, even though that
> voltage waveform begins to distort when the variac L is moved off "0".
> The second video shows 2 waveforms. <http://yfrog.com/bcrelutilz >
> Channel A once again is public service power, taken from the feed to the
> coil. Channel B is the voltage taken off the 20-turn magnet wire,
> measuring the passage of the magnets attached to the spinning tungsten
> electrodes. Here you can see the phase change quite dramatically, and
> it's clearly about 30 degrees before the motor loses lock.
> Apologize for the video quality - this is taken from my iPhone as I kneel
> beside the coil and rotate the variac.
> Hope this data is useful.
> On Jul 11, 2010, at 9:03 AM, Gary Lau wrote:
>> Hi Joe,
>> I didn't understand the before and after circuit configurations of your
>> phase Variac. A Variac is a 3-terminal device - Call the two ends of the
>> inductor A (hot) and C (neutral), and call the variable tap B. Mains
>> only goes to the case for safety (and should be used). In the phase
>> circuit, we want just a 2-terminal variable inductor, so use terminals A
>> & B
>> or B & C (plus the case ground).
>> If I understand your post, you previously used "the hot lead from the
>> to the variac and the hot lead from the output". Terminals A & B.
>> You then switched to "the hot and neutral from the output circuit".
>> Terminals B & C?
>> Both configurations should go from 0-100% of the total inductance, the
>> difference being that in one the full inductance is at the CW end and the
>> other is at the CCW end. I don't think there should have been any
>> difference in total phase shift achievable, if I understood the
>> configurations correctly.
>> You mentioned seeing phase shifts of 15 and 25 degrees. Since 360
>> is 16.67 msec, 15 degrees is only 0.69 msec, and 25 degrees is 1.16 msec.
>> this correct? It would be good to verify that the motor shaft varies by
>> same amount.
>> Regards, Gary Lau
>> MA, USA
>> On Sun, Jul 11, 2010 at 1:54 AM, Joe Mastroianni <joe at killerowls.com>
>>> HI Gary and John,
>>> I have made some mods to the phase control variac. Previously, what I
>>> been doing is using the hot lead from the input to the variac and the
>>> lead from the output (ignoring the neutral and ground) and using that as
>>> variable inductor.
>>> I put the phase shift circuit onto the oscilloscope channel B along with
>>> wall current channel A and looked at what happened when I rotated the
>>> variac. I got a phase shift along with a distortion of the waveform
>>> from a
>>> sinusoid to a distorted triangular shape, the peak of which maxed at
>>> 15 degrees away from the top of the public service voltage before
>>> in inductance caused more voltage drop than phase shift. After about
>>> degrees I was down to about 60V and the motor lost lock. This didn't
>>> too out of whack with what I had been told - that I'd only get about 15
>>> Then I tried just using the hot and neutral from the output circuit only
>>> in the picture on this page (http://www.hvtesla.com/phase.html) and I'm
>>> noticing about 25-30 degrees deflection before the motor loses lock.
>>> to Brandon for the help on that one.
>>> Now, as it turns out I have been out of the house all day on a family
>>> excursion, and only returned home this evening to do the experiments.
>>> a bit late to run the coil as it would indeed wake the neighbors, so I
>>> going to have to wait till tomorrow AM to run some experiments watching
>>> streamers. I did try to get a video using my iPhone yesterday in the
>>> semi-darkness, but the dark image response of that thing is pretty
>>> awful -
>>> even the streamers are hard to see in what it thinks is pitch black. I
>>> to get a real camera.
>>> More to come.
>>> With regards,
>>> On Jul 10, 2010, at 3:08 PM, Gary Lau wrote:
>>>> Hi Joe,
>>>> There's no way that having the gap BPS being a sub-multiple of the
>>>> frequency would affect anything. When the SRSG phase is significantly
>>>> of whack, the gap fails to fire at the end of a mains half-cycle
>>>> electrodes aren't aligned at the voltage peak. So the voltage now
>>>> charge in the opposite direction. But contrary to intuition, this
>>>> isn't simply "lost". In an L-C resonant circuit, the energy transfers
>>>> alternately between voltage in the cap and current in the inductor.
>>>> circuit at this point is the cap, and the NST secondary winding,
>>>> at or near 60Hz. Yes, you have a Larger Than Resonant sized cap, but
>>>> NST secondary currents flow through capacitive loads, we think that the
>>>> current limiting shunts saturate, and ferroresonance kicks in and
>>>> stuff happens, permitting at or near mains resonance. Bottom line, the
>>>> energy isn't lost, but when the voltage reverses, then charges to the
>>>> opposite polarity, it will reach a FAR higher voltage than on the
>>>> half-cycle, enough to break down a safety gap or insufficient RSG
>>>> As John said, it's not possible that the cap can carry a charge between
>>>> bangs. Once a spark gap fires, it won't stop firing until pretty much
>>>> of the energy has been dissipated. This happens pretty fast, and the
>>>> that the electrodes are spinning quickly doesn't matter.
>>>> Now that your gap timing is in the ballpark, can you describe what
>>>> as you sweep the phase adjustment Variac through its range?
>>>> Regards, Gary Lau
>>>> MA, USA
>>>> On Sat, Jul 10, 2010 at 12:09 AM, Joe Mastroianni <joe at killerowls.com
>>>>> Hi All,
>>>>> Well, indeed the problem was curable by advancing the timing on the
>>>>> about 30 degrees. Now I can bring it up to full power - BUT - I would
>>>>> that on full power the arcs aren't as large as they were when I had
>>>>> "sputter" problem. After doing a bit of reading I've come to the
>>>>> that the sputter was indeed some off-timing resonance issue where I
>>>>> entirely discharging my MMC bank. Perhaps it was some out-of-control
>>>>> ringup issue that the safeties were saving me from, with the pulses
>>>>> the cap bank at some multiple of the resonant frequency. I measured
>>>>> Thus the safeties were firing to offload the additional charge.
>>>>> Java TC suggests my resonant freq is around 130khz but it wouldn't
>>>>> me if it was closer to 120khz, which, perhaps, would cause some issues
>>>>> 120bps under certain conditions. It would take me going out and
>>>>> some HV probes for my scope to see it for myself...
>>>>> Meanwhile, it works now fine.
>>>>> Of course, this will not stop me from trying the other SRSG designs.
>>>>> too curious.
>>>>> I'm also going to try an ARSG, as I do like the sound.... ;-)
>>>> Tesla mailing list
>>>> Tesla at pupman.com
>>> Tesla mailing list
>>> Tesla at pupman.com
>> Tesla mailing list
>> Tesla at pupman.com
> Tesla mailing list
> Tesla at pupman.com
More information about the Tesla