Author Topic: Hydron's 160mm DRSSTC and topload current measurements  (Read 4694 times)

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Hydron's 160mm DRSSTC and topload current measurements
« on: August 28, 2017, 05:17:15 PM »
Rather a long time ago (2014) I took some interesting measurements of the topload and breakout point current in a medium sized (160mm/6" secondary) DRSSTC at a number of different power levels.

I was always intending to do some post processing on the data to learn more about streamer/arc impedance and topload conditions during a burst, but never ended up having the motivation/time to do so (the idea was to use it as an excuse to learn some python coding skills!). I think it's about time I put this data out in the public for all to have a look at, with the hope that some find it interesting and that it helps us learn more about tesla coil physics.

The other reason to get this out there now is that I'm forgotting some important details about the coil which was measured, as I've been living on the other side of the world from it since these measurements were taken and my notes at the time were limited. I unfortunately can't verify coil parameters or repeat the measurements with more documentation until at least next April. That said, the coil specs I have are as follows:


DRSSTC specs, info:

Secondary: 160mm dia, ~730mm winding length (see attached JavaTC file for more accurate numbers), 1920 turns of 0.315mm dia. wire
Toroid: ~170mm minor diameter, ~760mm major diameter (as above, see attached JavaTC file). Constructed of aluminium ducting.
Primary: 11 turns of 6.35mm dia. copper tube, spaced approx 15mm centre-centre, starting at approx 200mm diameter
Tank capacitor: 48x 2uF 1000V Aerovox snubbers, in 2 parallel strings of 24. Tappable for a range of ~400-166nF at 10-24kV (I believe these tests were at the lower end of this range, likely at 166-200nF)
Coupling: approx 0.15 (adjustable)
IGBTs: 2x CM300DY-24H, ~750A OCD setting
Bus capacitance: 2x4700uF in series
Rectifier: Voltage doubler
Variac: 15A, ~0-260V output
Driver: DIY design based on UD2.5. Can be assumed to be equivalent to any UD2.x design.
Interrrupter: DIY design based on oneTesla midi interrupter code.

The coil is not too sensitive to tuning, and runs happily on both upper and lower pole tuning. I find that with the >600V available with the voltage doubler rectifier that upper pole tuning with higher-Z primary gives nicer, more controllable streamers, especially when playing MIDI music with the coil. I also tried lower pole tuning with a lower impedance primary setup and non-doubler rectifier - this also gave good streamer length but they were more "chaotic", MIDI did not work as well and streamer length was much more sensitive to coil input power.

I've attached a few pictures of the coil in action, with some ground strikes in the ~1.5-2m range. Max strike length achieved is almost 2.5m, or a little over 3x secondary wound length. I have not pushed it properly - the only part I've blown up on this coil was a plugpack supply that was hit by a streamer during the prototype phase!






Video taken while controlling coil - dont drive and film if you want quality!. I believe this was lower pole tuning.

Measurement info:

Scope:
- The topload current measurements were done using a Cleverscope CS328A PC based oscilloscope (see https://cleverscope.com/products/), with the ethernet interface option installed. This allowed me to power the scope and a wifi router off batteries, and locate them on the topload while maintaining control and data download capability remotely from my laptop.
- I've attached a picture of the first rough test setup - the wires were tidied up and the wirewound resistors were replaced for the actual measurements, but the idea remains the same.
- With 5.8GHz wifi, a faraday cage shield could be located over the test equipment, with only a small hole needed to get the wifi signal out, allowing it to be completely safe from the output voltage.
- A small DC offset may be present in the measurements due to non-perfect calibration of the scope - this probably should be removed before doing any calculations using the data.



Probing/measurement setup:
- A 1R resistor was placed in series with the connection from the top of the secondary to the toroid, and the 2R resistor placed in series with the toroid and the breakout point
- All measurements are of the voltages across these resistors, referenced to the toroid. As such, the voltage measured for the breakout current (channel B) is 2V/A and 180 degrees out of phase with the toroid current (channel A), which is 1V/A. Inverting and scaling by x0.5 will correct the channel B voltages and give the breakout point current at 1V/A - channel A data needs no modification and directly shows toroid input current at 1V/A.
- Scope was setup for 30ns between samples (i.e. 33MS/s), with a 20MHz bandwidth filter on both channels. As the input frequency is not expected to be high aliasing shouldn't be present, despite the nyquist sampling criteria not being met. The lowest power measurement was at a slightly slower sampling rate, 50ns between samples (20MSa/s), cant remember why sorry!

Coil operation during measurements:
- The topload currents were measured at 5 different power levels (file naming is related to the descriptions below, should be self-explanatory):
   - Very low power (no breakout)
   - Medium power (breakout into streamer, no ground strike)
   - Higher power (breakout into streamer, no ground strike)
   - Even higher power (small ground strike)
   - Highest power (big ground strike)
- In each case BPS was set to ~100, with power adjusted by changing variac output (and possibly the on-time, unfortunately I can't remember)
- For each power level, the first 100 on-periods from interrupter signal being applied were captured by using the segmented memory of the scope. This allowed for a reasonable sample rate (33MS/s) to be used by only capturing actual data, not the time when the interruptor was off. The number of captures (100 = 1 second at 100BPS) is enough to show the evolution of a streamer or ground strike in a way that a single capture cannot.
- Videos were captured for 3 of the 5 power levels, and can be found here:
I have the original video files if they are useful to anyone, but youtube is more convenient.
- A picture of the coil in it's measurement setup is attached below. The item seen on top of the topload is a plastic block I used to keep the faraday shielding in place.



Notes on captured data:
- All 500 waveform captures are found here, 7zipped up (7zip compressed them much better than plain zipping): https://highvoltageforum.net/files/hydron_topload_current_waveforms.7z
- Each is in a CSV type format (I think the delimiting character is a tab rather than comma), with a few lines of header at the start.
- The "TriggerTime" number is in days from a start-of-1900 epoch (unsure what time zone etc); the fractional part can be converted into seconds by multiplying by 24*60*60. Difference between sequential captures should be very close to 0.010 seconds (100BPS).
- Trigger point will unfortunately not be in exactly the same point every time, as it is triggering off input current which varies depending on power level. Enough pre-trigger data was captured however to show the start of every on-period.


Hopefully this data is of use/interest to some, and can be processed to learn more about the complex impedance of streamers during growth, which may allow for better design decisions to be made when building coils. Some things to remember while doing so:
- This only captures current flowing into the topload capacitance, not the secondary coil capacitance. JavaTC splits these out into separate numbers.
- The breakout point also has it's own additional capacitance to ground. This can be estimated by looking at the low power (i.e. before capacitive streamer forms) measurements of current flowing into it in comparison to the toroid current (which will be the sum of current due to both toroid and breakout capacitance).

Edit: see a couple of waveform captures below, showing raw data before scaling and inversion of channel B
« Last Edit: August 28, 2017, 05:40:44 PM by Hydron »

Offline futurist

  • Global Moderator
  • High Voltage Technician
  • *****
  • Posts: 116
  • Karma: +0/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #1 on: August 28, 2017, 10:09:49 PM »
We finally see your coil in action, well done and thanks for the writeup :)

One thing I also noticed with my lab notes, it's fascinating how fast people forget some important details (or how fast the time goes by). After a year it's getting hard to repeat some of the experiments

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1048
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #2 on: September 04, 2017, 01:54:31 PM »
I love the amount of details in your post, it only lacks technical porn with closeups of the inverter, MMC and any small clever details you made up :)

Putting a 1500$ oscilloscope on top of a Tesla coil is also pretty bold, but demonstrated quite well that it did not matter with proper shielding, did you observe any corona from the antenna peeking out?

I hope Uspring got something clever to say about the waveforms, personally I will play a little around with all the data and see where it gets me!
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1048
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #3 on: September 05, 2017, 11:35:44 AM »
I started crunching numbers on the series of big_gnd_strike_0 to big_gnd_strike_99 and the result is likely a video animation of the run, it will take a little while and maybe ill first have something middle next week
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #4 on: September 06, 2017, 12:24:13 PM »
Was not concerned about corona, damage etc to the scope at all - only thing poking out of the shield was the breakout point; the wifi antenna was inside radiating through a small (but big enough for 5GHz wifi) slot.

I have attached a diagram which will hopefully make the measurement setup clearer, as well as a couple of photos of the inverter/MMC (in an incomplete state).

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #5 on: September 06, 2017, 04:13:48 PM »
Hydron, that's a beautiful data set you've got there. AFAIK the best one ever made. I haven't got around to analyzing it in more detail, since I've been away and need to write some code in order to do this. The many bursts in each set allow the study of the effect of a preheated arc channel and also the kind of variation of arcs in a run. Very neat, data sets like this maybe allow to estimate the best primary and secondary parameters when designing new coils.

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1048
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #6 on: September 06, 2017, 09:46:37 PM »
I know this was a tedious excel job, but I could just not wait or spend time on writing code when it was just a mere 100 data sets and a screenshot :o

I have actually been blown away by the data, how the charge/discharge current behaves from the start of a ground strike till the end.

From talking with Hydron on IRC we have so far come up with the shorter burst in the ground strike must be the OCD kicking in, but we are not sure as OCD LED is not shown in video. But it is really interesting to see how it builds up a additional cycle while recovering from the initial hard hitting ground strike that completely skews the current oscillation frequency.

/>
Edit: There is a conversion error in the excel data, that makes it look like there are dead-time in the breakout current, but that should be just as sinusoidal as the topload current, the error is that excel somehow sees numbers below 100.000.00 as 1.
 
« Last Edit: September 07, 2017, 08:46:02 AM by Mads Barnkob »
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #7 on: September 07, 2017, 01:10:28 PM »
That's a very nice visualisation of the data. The OCD kicking in makes much sense to me. The power draw on the primary is proportional to Iprimary * Isecondary. The ground arc removes most of the secondary tank energy and the top load current, which is the same as the secondary current, drops to almost 0. This implies, that power transfer from the primary to the secondary also drops to a small value. Without the secondary load, the bridge will boost primary current up from the already large value, triggering the OCD.

Another explanation, though, might be a high speed transient from the ground arc finding its way into the electronics.


Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #8 on: September 07, 2017, 02:59:10 PM »
It doesn't appear to me that the OCD is tripping immediately upon the ground strike hitting - while afterwards the current drops nearly to zero it does rise again before eventually trailing off, which I don't think is consistent with an immediate OCD trip.

As a guess I'd go with the first theory - maybe the ground strike pulls the system out of tune enough that less energy is transferred from the primary, allowing it to ring up to a level high enough to trip the OCD as suggested. It definitely requires more analysis and to look at the other data sets rather than just the ground strike one.

It's becoming obvious that simultaneous measurement of the primary current & drive signal would also be very useful - while I unfortunately won't have access to the original coil before at least next April, I'll try and get some of these measurements for a QCW coil (run in both QCW and non-QCW mode if possible), hopefully by the end of this year.

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #9 on: September 08, 2017, 01:03:46 PM »
Getting new data would be wonderful, especially since it probably is at quite different coil frequencies and burst lengths. Not that I've gotten really started analyzing the present one  ;)
Looking at arc development from burst to burst, do you think, that initial bus voltage changes between bursts? I.e. is the 10ms pause between bursts sufficient to power up the bus caps? I guess some info about this can be indirectly obtained from the top load current rampup, since that depends primary rampup and therefore bus voltage.

« Last Edit: September 08, 2017, 01:12:36 PM by Uspring »

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #10 on: September 08, 2017, 01:30:39 PM »
At 100BPS the bus caps will be charged (by 50Hz mains) at approx the same rate as the burst rate, so I'd hope (but cannot guarantee) that it's a reasonably constant starting bus voltage within each data set.

Any new data would likely be at 300kHz+ and (if in QCW mode) much longer burst lengths. While it's already making small sparks, the coil in question is still a long way off being done though (biggest job is a bunch of coding and PCB design that's not far progressed).

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1048
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #11 on: September 18, 2017, 05:03:39 PM »
In regard to the ground strike data, I thought I would bring forth this old discussion/information due to the high resemblance between the current waveforms of Hydrons measurement and Pauls simulations.

From a old TCML thread about heavy ground strikes, whiplash and wavefronts: https://www.pupman.com/listarchives/2012/Sep/msg00138.html

Steve Ward wrote:

> During a ground discharge, the voltage at the topload
> has been measured to collapse within 100nS (worst case)
> and 250nS (more typical).

> ... during this event is a wave front propagates down from
> the top of the coil.

> ... the capacitance from end to end of the secondary
> (including the capacitance from the topload to the coil)
> should essentially provide dispersion, making this
> wave-front not so steep,

Paul Nicholson wrote:

Yes dispersion spreads the down-going transient although
it could still build up quite a high volts/turn near the
base.   This has been discussed before as one possible
cause of racing arcs.  Here is an animation of secondary
voltage distributions (July 2009, modeled on Marco
Denicolai's Thor system),



A normal ring-up of the secondary occurs until at 40uS
the topload is suddenly and completely discharged to
ground.

A slow-mo of the discharge,

http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Steve Ward

  • High Voltage Enthusiast
  • *
  • Posts: 15
  • Karma: +0/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #12 on: September 19, 2017, 03:20:29 AM »
Hi Guys!  I've been off the forums for awhile, but Mads tipped me off on this thread.  Good stuff here!
 
I don't have much to add yet, just a few observations.

What type of resistor was used for the current shunts?  Hopefully some sort of low inductance metal film type?

If you get to try with higher bandwidth, it's expected that there are high current events in the nS time scale that would give more insight to the streamer growth itself.  If you look into how corona detection works, and the theory of stepped leaders, there's an expectation that the streamer grows in sort of discrete steps, and as that charge jumps out into space it shows up as current pulses sourced from the toroid capacitance.  Just thinking on where this could go...  On QCW coils i expect all the streamer growth to happen during the negative voltage portion of the oscillation (this is my 'theory' for un-branching streamers), but i bet for regular DRSSTCs it happens on both positive and negative cycles.

Quote
As a guess I'd go with the first theory - maybe the ground strike pulls the system out of tune enough that less energy is transferred from the primary, allowing it to ring up to a level high enough to trip the OCD as suggested. It definitely requires more analysis and to look at the other data sets rather than just the ground strike one.

Ground arcs are really low resistance compared to the impedance of the secondary coil so it's expected that the secondary no longer acts as an LC, but just an L with a parallel resistance, which can be boiled down to a single-resonant transformer.

Im surprised there does not appear to be much phase difference in spark vs toroid current.  This suggests that the spark's resistance looks small compared to its capacitive reactance.

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #13 on: September 19, 2017, 01:16:47 PM »
What type of resistor was used for the current shunts?  Hopefully some sort of low inductance metal film type?

Resistors were what I found in the junk box that looked least likely to be wirewound (hence the different values for each one - couldn't find 2x suitable 1R resistors). I did a quick check of their inductance and was happy at the time (though I cannot remember the exact result due to it being a fair while ago), but I may not have considered frequencies as high as may be been present when ground strikes occur. In the future I'd use a better resistor or a high-freq Pearson.

If you get to try with higher bandwidth, it's expected that there are high current events in the nS time scale that would give more insight to the streamer growth itself.  If you look into how corona detection works, and the theory of stepped leaders, there's an expectation that the streamer grows in sort of discrete steps, and as that charge jumps out into space it shows up as current pulses sourced from the toroid capacitance.  Just thinking on where this could go...  On QCW coils i expect all the streamer growth to happen during the negative voltage portion of the oscillation (this is my 'theory' for un-branching streamers), but i bet for regular DRSSTCs it happens on both positive and negative cycles.

I will be doing some more measurements on a QCW coil once it's built, but that could take a fair while - it's half done but I've got a lot of pcb and software work to do designing/building a new phase-shift QCW driver and controller from scratch (I may be interested in details of your UD3 work as inspiration for this - at the very least it might make sense to copy your driver/controller comms protocol rather than dream up a new incompatible one!).

Any new floating topload measurements would be at 100MSa/s (max of the scope I'm using) allowing for up to ~40MHz single-shot bandwidth, but I would also make some simultaneous higher-bandwidth measurements of the primary (am thinking bridge V, primary I, secondary base I, E-field, plus some digital channels for interrupter/gate drive/ocd).

Unfortunately I don't have access to anything other than the half-complete QCW coil here in the UK, otherwise I'd do some more rigorous measurements now. Hopefully when the QCW is done I can get some info on your theory, and potentially also normal DRSSTC operation if I'm able to run it in non-QCW mode without flashovers (maybe with a flat primary?)

I'm surprised there does not appear to be much phase difference in spark vs toroid current.  This suggests that the spark's resistance looks small compared to its capacitive reactance.

I never ended up digging deep into the data, but it looked to me like the ground strike current was mostly in phase with the topload voltage, while pre-ground strike it was not. I've attached a waveform that I've done some maths to - the numbers will be a bit rough and I had to apply a high-pass filter to the voltage (which would result in some phase shift) but it gives some sensible looking results, with the streamer load going from capacitive to resistive (~25k ohms at it's lowest) once the strike channel is established (Ch A is topload voltage, B is breakout current in A):
« Last Edit: September 19, 2017, 01:19:11 PM by Hydron »

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #14 on: September 19, 2017, 02:31:02 PM »
I've been stopped short by some preliminary results analyzing the data. In order to calculate top load voltage I integrated the net current going into the top load, i.e. channela + 0.5*channelb. The + sign is because channelb is reversed in polarity. The integral is the top load charge, which has to be divided by its capacitance to get the voltage. Here is what I got:

Somehow the x axis numbering got screwed. The whole diagram time is 1ms.  On the right side the effect of a small offset is visible, but during the action at the left side the baseline is obviously tilted down. I'd have to add a huge offset to correct for this, but this doesn't make sense to me. I also considered a different sensitivity of the scope for positive and negative voltages. That makes the diagram more acceptable, but still not good enough to extract reliable phase information between voltage and arc current, apart from the fact, that it is implausible, that the scope has a problem like this. Does anybody has an idea?
The first diagram was made from big_gnd_strike_0.txt. I made another one from big_gnd_strike_99.txt. This one has a ground strike.

It appears, that the ground strike removes most voltage from the top, if one considers the skewed baseline. The current waveform around the spike does not appear to show much ringing, which would be expected, as the current rises and drops fast there. This puts a limit on the inductance of the resistor measuring the arc current. The duration of the ground strike seems to be around 1-2 us.

Sadly, the failure to obtain a reliable top voltage implies a difficulty to calculate top voltage - arc current phase relationships. But the arc indeed looks more capacitive than resistive.

I don't know if high time resolution measurements can clear up the step like growth of arcs. The arc is like a resistive/capacitive low pass network, which smoothes out any high frequency details coming from its tip. The diagram in Hydrons first post show initially a non sinusoidal arc current. Perhaps that still persists during later cycles but is invisible in the arc current plot.

Hydron, can you give some details to the maths applied? Edit: I see, that you high pass filtered the data. Clever. Phase shifts due to this should almost cancel if done on both channels.




« Last Edit: September 19, 2017, 08:03:14 PM by Uspring »

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #15 on: September 19, 2017, 11:32:16 PM »
I don't have the exact details at hand on the maths I used (it's on another PC), but basically I used a high pass filter (fc of 15kHz or so) to knock off the worst of the skew. While there is no big issue getting rid of any DC component the error is unfortunately not a pure DC signal, so a filter was the what I needed to use, with it's drawbacks of eliminating some real signal and it's associated phase shift.

I agree that there's something a little odd going with the captures - while integrating a signal will make small offsets/non-linearities stand out strongly, I'm really not sure why it's this bad (and why it's not a DC offset). There are also some small discontinuities (jumps) in some of the current data which I'm not sure are real either.
One potential reason is that I think I put some TVS diodes in parallel with the current sense resistors for scope input protection, assuming that their capacitance would be small enough relative to 1-2 ohms at the TC frequency to ignore (2nF at 75kHz is approx 1k ohm), but there may have been some non-linear effects introduced that have skewed the results. When I do future measurements I'll skip the TVS diodes (though I'm not 100% sure I used them here, memory fails me).

Hopefully the knowledge of the physical system will allow for some correction to be done other than a brute force filter, e.g. looking at each half cycle in isolation, or determining if the apparent error is dependent on the magnitude of each current peak (as could happen if positive going response wasn't quite equal to the negative).

I'll try to post some more info on the maths I used tomorrow, but it's nothing too interesting, just an IIR filter that was picked fairly arbitrarily to make the waveform look good without too much phase shift.
« Last Edit: September 20, 2017, 09:10:14 AM by Hydron »

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #16 on: September 21, 2017, 11:28:21 AM »
Thank you for the info. I was able to make my diagrams look like yours by adding a simple one order C/R type 15kHz high pass. Including a scaling down of negative scope values by 0.93 helped also somewhat. I'll experiment with some parabolic slope. Hard to tell where the non linearity comes from.

I found it interesting, that the current voltage phase reverts to 0 after the ground strike. Expected, though.

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #17 on: March 20, 2018, 03:32:26 PM »
It's been a while since Hydron posted his arc current measurements, so I'll shortly describe, what he obtained. He scoped the current between the secondary top and the top load and also the current between top load and the breakout point. Alltogether it's a set of five runs, 2 made at the max power and 3 more at decreasing power levels. The lowest one didn't cause any breakout, but presumably some corona. The runs were 1s long at 100 bursts per second.
I've made similar measurements some years ago on a lower powered coil. http://4hv.org/e107_plugins/forum/forum_viewtopic.php?153922

From the net current to the top, i.e. the difference between secondary to top and top to arc current, the net charge transferred to the toroid can be calculated by summing the net current over time. The top voltage is the ratio of charge to toroid capacitance. I've assumed a toroid C of 31pF. The voltage is not completely accurate, since it is influenced by other nearby charges, e.g. in the top windings of the secondary and also the arc itself. But it is difficult to account for that.

Summing the input currents is very sensitive to non linearities due perhaps to the scope or the TVS diodes across the measuring resistors. I tried to model the non linearities, but that proved to be only partially successful. So I also high pass filtered the currents in order to remove any DC contributions from the non linearity. Still the results weren't completely satisfactory, but I left it at that. My own measurements also show this effect.

Here are shown 2 voltage and current diagrams from the first 2 burst at max power. The voltage scale is shown at the left (Volts) and the current on the right (Amps). Voltages are in blue, currents in red. Nicely visible in the first diagram is the effect of heating up the spark channel. At about 40us the voltage reaches 300kv but the current only about 0.5A. At about 200us the voltage is only about 200kv, but the current peaks at about 3A.





Even between bursts some heat still remains. The second diagram shows a reduced peak voltage. This is probably due to the earlier rise of current causing a larger loading in the second burst. Hot air expands, which makes it thinner and thus reduces the breakdown voltage. The following later burst during the run don't show significant reductions in voltage anymore.

For a comparison of power levels, I've made some diagrams in decreasing order of power, each taken near the end of a run.  The max currents show a strong dependence on voltage. Between the first and third plot the peak currents decrease from about 3A to 1A, while peak voltages don't differ very much. The rapid rise of current at increasing voltage probably pretty much clamps the voltage. The limited voltage does not imply a short arc. I believe the current is a much better indicator of arc length than voltage. A high current transfers large charges somewhere and large charges can be only taken up by large capacitances of the arc. So we need long arcs in order to account for big currents. The last diagram just shows some corona. Peak current is only about 15mA.









With the exception of the lowest power level, the arc load is not purely resistive. It has also a capacitive component, which can be seen by the phase shift between voltage and current. This corresponds to the well known effect of the lowering of secondary resonance frequency after arc breakout. The diagram below shows the phase shift (blue) in degrees and the coil frequency (red) for the max power burst. The phase shifts are a bit larger than my own measurements (~50-60 degrees). From the phase shift, current and voltage and frequency it is possible to calculate the capacitance and resistance of the arc. At the point of max current at 180us time into the burst the capacitance reaches 29pF, which is almost that of the toroid. The secondary tank doesn't get completely out of tune because there is also a strong resistive load of about 210kOhm, which causes the secondary Q to drop to around 4. The resonance curve becomes quite flat then. The largest rms power consumed by the arc is about 120kW.



In the last diagram I've collected some statistics. Each point corresponds to one burst. Shown are the max net currents into the toroid, i.e. toroid current minus arc current on the x-axis. 1A is roughly equivalent to 50kV on the toroid. On the y-axis there is the max arc current for the burst. The max power, higher power and medium power runs are combined here. There are 3 blobs visible the medium power one around 1A arc current, the higher power one at 2.5A and the max power version above 3A. The max power blob has a number of high arc current outliers produced by ground arcs. I should add, that the point of max arc current in a burst usually not coincides with that of max voltage due to the dynamical character of arc growth.


« Last Edit: March 20, 2018, 03:43:04 PM by Uspring »

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 239
  • Karma: +7/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #18 on: March 22, 2018, 11:51:49 PM »
Thanks for the analysis, will read through it properly when i get a chance (busy travelling at the moment).

I will be in the position to do some more measurements in a week or so with a few improvements:
- sampling a bit quicker (100MSa/s)
- probably sampling at 14 bit rather than 10
- will use a couple of pearson model 2877 CTs for toroid current measurements rather than the resistors used earlier
- simultanously measure primary/secondary-base currents with a second scope (triggered via fibre link from first scope - will need to calibrate the delay)
- better characterise the coil prior to measurement, i.e. do better physical measurements, check resonant frequencies of secondary (with/without toroid) and primary (with/without secondary) and also check MMC and primary tap points

If there are any other suggestions of how to improve the measurements then I'd be happy to hear them.
I'm assuming that measuring primary/secondary-base currents is the best use of the second scope rather than using one of the channels to look at bridge voltage output? (Might be able to have a 1-bit value recorded of that anyway via the ext trigger input).

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 105
  • Karma: +8/-0
    • View Profile
Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #19 on: March 23, 2018, 11:32:39 AM »
Looks like a good idea to use CTs for current measurements. Do you have an idea about their bandwidth considering e.g. ground strikes? It still irks me, that I have no good conception of what is causing the drifts of the current integrals. An advantage of using resistors is, that you can catch possible polarity dependent effects of the arc. A CT might filter that out.
More information about your coil, particularly the primary tank would also be useful.

In the last few months I've been thinking about my arc model. A major problem is the dynamical nature of arc growth, mostly determined by heat production, heat capacitance and heat conduction. A lot simpler would be the investigation of more or less stationary arcs produced e.g. in a QCW TC. A post about an improved version of the model applied to your coil is in the works.


High Voltage Forum

Re: Hydron's 160mm DRSSTC and topload current measurements
« Reply #19 on: March 23, 2018, 11:32:39 AM »

 


* Recent Topics and Posts

post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
Today at 11:19:45 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
Today at 03:36:44 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
DICKEYBIRD
Today at 03:24:38 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
Today at 02:23:29 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
DICKEYBIRD
Today at 01:06:44 AM
post Re: Floppiclene! A Electronics Cleaning Relic of the Past.
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Justin
July 18, 2019, 10:29:32 PM
post Re: Floppiclene! A Electronics Cleaning Relic of the Past.
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Mads Barnkob
July 18, 2019, 08:52:35 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
July 18, 2019, 05:49:27 AM
post A new quick way to get into tune a DRSSTC
[Dual Resonant Solid State Tesla coils]
VNTC
July 18, 2019, 03:13:27 AM
post Re: Floppiclene! A Electronics Cleaning Relic of the Past.
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Twospoons
July 18, 2019, 12:06:53 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
July 17, 2019, 07:56:41 PM
post Floppiclene! A Electronics Cleaning Relic of the Past.
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Mads Barnkob
July 17, 2019, 11:25:22 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Netzpfuscher
July 16, 2019, 07:39:33 AM
post Re: What is the true in tuning guide?
[Dual Resonant Solid State Tesla coils]
VNTC
July 16, 2019, 05:11:21 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
profdc9
July 16, 2019, 04:05:07 AM
post Re: DRSSTC in brust mode
[Dual Resonant Solid State Tesla coils]
SAMGO
July 16, 2019, 02:25:27 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Hydron
July 16, 2019, 01:54:30 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 15, 2019, 02:51:26 PM
post Re: What is the true in tuning guide?
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 15, 2019, 02:48:02 PM
post Re: A problem solved with hand tools
[General chatting]
Mads Barnkob
July 15, 2019, 01:48:08 PM
post Re: Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 15, 2019, 01:45:00 PM
post Re: A problem solved with hand tools
[General chatting]
Hydron
July 15, 2019, 11:21:29 AM
post What is the true in tuning guide?
[Dual Resonant Solid State Tesla coils]
VNTC
July 15, 2019, 10:01:21 AM
post A problem solved with hand tools
[General chatting]
klugesmith
July 15, 2019, 07:52:41 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
klugesmith
July 15, 2019, 07:33:30 AM
post Re: tesla coil varnish is dumb
[Solid state Tesla coils]
Hydron
July 14, 2019, 10:49:03 PM
post Re: tesla coil varnish is dumb
[Solid state Tesla coils]
T3sl4co1l
July 14, 2019, 09:59:40 PM
post Re: tesla coil varnish is dumb
[Solid state Tesla coils]
Mads Barnkob
July 14, 2019, 08:20:52 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Netzpfuscher
July 14, 2019, 06:06:30 PM
post tesla coil varnish is dumb
[Solid state Tesla coils]
Fumeaux
July 14, 2019, 05:29:10 PM
post Led Closeups with Mobile phone and lens .
[Smart phones]
DashApple
July 14, 2019, 03:00:09 PM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 14, 2019, 02:06:26 PM
post Re: DRSSTC in brust mode
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 14, 2019, 02:04:22 PM
post Re: Twin System Build
[Dual Resonant Solid State Tesla coils]
oneKone
July 14, 2019, 01:27:18 AM
post Re: DRSSTC in brust mode
[Dual Resonant Solid State Tesla coils]
johnf
July 14, 2019, 01:22:15 AM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
oneKone
July 13, 2019, 03:48:54 PM
post DRSSTC in brust mode
[Dual Resonant Solid State Tesla coils]
SAMGO
July 13, 2019, 04:17:55 AM
post Re: Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
donnersm
July 12, 2019, 08:54:54 PM
post Re: Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
profdc9
July 12, 2019, 07:16:40 PM
post Re: Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
profdc9
July 12, 2019, 06:33:00 PM
post Smooth corona or heavy spark, RF wave radiation frequency of Tesla coil drops ve
[Dual Resonant Solid State Tesla coils]
VNTC
July 12, 2019, 11:09:28 AM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 12, 2019, 09:56:38 AM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
oneKone
July 12, 2019, 01:23:58 AM
post Re: Completed my pulsed power generator
[Capacitor banks]
MRMILSTAR
July 11, 2019, 05:48:27 PM
post Re: Completed my pulsed power generator
[Capacitor banks]
Mads Barnkob
July 11, 2019, 02:58:13 PM
post Re: Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
July 11, 2019, 02:45:24 PM
post Kaizer half bridge driver
[Dual Resonant Solid State Tesla coils]
donnersm
July 11, 2019, 06:59:01 AM
post Completed my pulsed power generator
[Capacitor banks]
MRMILSTAR
July 11, 2019, 05:43:31 AM
post Re: Twin System Build
[Dual Resonant Solid State Tesla coils]
fnordest
July 11, 2019, 01:05:42 AM
post Re: "LightningStorm" VTTC project
[Vacuum tube Tesla coils]
Avenger
July 09, 2019, 07:31:37 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30