ZVS long cable freq stabilization

[kreasteve]

Hi all, I'm new here and to high voltage.

I wanted to build a weed killer. By the way it already exists: https://zasso.com/

Take a look on my picture. (Test with external 24V/400W supply)

I bought a ZVS from Austria with a flyback transformer, improved the heat stability and built some parts around it, included two 1.5m long 40kV unshielded cables. The building process isn't finished: Safety circuits, frontpanel and a few other things are in development.

The problem is, that the generated frequency (normally 15kHz as you might expect) turns to 75kHz if both long cables are connected. I think there is some capacitive coupling over the ground. Can I somehow stabilize the freqency?

Other things I want to say:
-The three parallel SMPS are not the best way to produce ripple free voltage. But they are laying around.
-I should have started with highvoltage earlier. Its very interesting as I notices during the tests. A tesla coil is on my ToDo list.


Tanks in advance!

Best
Steven

[T3sl4co1l]

Try more capacitance across the primary?  Idunno, it's a dumb oscillator (if it's the usual simple circuit), it can do whatever it wants, and transformers are complex networks with many characteristic frequencies.

Tim

[kreasteve]

Hi Tim,

thanks for your reply!

Yes the frequency of the ZVS depends on the load.
I had allready tried double the capacity.

I'm going to build an H-Bridge on my own to enforce the frequency I want.

Steven

[fwblack]

Steven, did you ever get this working?  I've been researching and trying to build a weed fryer based on this: https://rootwave.com/portfolio-item/rootwave-pro/ and have not made much progress.  I had not found the Zasso product you linked to, it looks much simpler.
Fred

[davekni]

Hadn't paid attention to this thread until yesterday's post.  Prompted me to research electric weed zapping a bit.  Also decided to play a bit with driveway weeds this afternoon.  Not at all ideal conditions.  Reading suggests weed zapping works best when top layer of soil is dry, causing more current to flow deeper into weed roots.  Ground is quite damp here and it was starting to sprinkle slightly.  So likely poor depth today.  Still a bit of fun:
   

/>This playing uses the 100kHz supply in my Jacob's ladder posted here:
    https://highvoltageforum.net/index.php?topic=831.msg5491#msg5491
Supply uses a ZVS oscillator feeding DIY ferrite transformer.

The problem is, that the generated frequency (normally 15kHz as you might expect) turns to 75kHz if both long cables are connected.

Doubt the original poster will be back to see a reply, but the issue appears to be little or no gap in HV transformer ferrite core.  As I discuss in above Jacob's ladder post, coupling factor needs to be below 86% to keep a ZVS oscillator running under all arc load conditions.  With coupling under 0.86, frequency with secondary loaded will be at most 1.95x higher than open-secondary frequency.  Reported 5x change suggests high coupling, which suggests little or no gap.

[kreasteve]

Steven, did you ever get this working?  I've been researching and trying to build a weed fryer based on this: https://rootwave.com/portfolio-item/rootwave-pro/ and have not made much progress.  I had not found the Zasso product you linked to, it looks much simpler.
Fred

Yes, I‘m using it with an 18V makita accumulator. The power is not as powerful as I want to.

Doubt the original poster will be back to see a reply, but the issue appears to be little or no gap in HV transformer ferrite core.  As I discuss in above Jacob's ladder post, coupling factor needs to be below 86% to keep a ZVS oscillator running under all arc load conditions.  With coupling under 0.86, frequency with secondary loaded will be at most 1.95x higher than open-secondary frequency.  Reported 5x change suggests high coupling, which suggests little or no gap.

That sounds interesting. Which gap do you mean?

[davekni]

That sounds interesting. Which gap do you mean?

Ferrite cores for typical flyback transformers have a gap or two, small non-magnetic spacer between the two ferrite C-cores.  This allows operation with a DC bias as in flyback circuits, or in an L/C resonant circuit such as these ZVS oscillators.  Cores without a gap need to be driven with symmetric AC waveforms such as from an H-bridge.  Any significant magnetization current (primary current not cancelled by secondary current) will cause cure saturation without a gap.
Also, gap reduces coupling factor between primary and secondary.  If coupling factor is above 0.86, Q can get too low for ZVS oscillator to oscillate for certain arc load resistances.  ZVS oscillator DC input current ramps up very high when oscillation stops.