Grounded strike target
Graham Armitage, Tue Apr 01 2014, 01:58AMBased on all I have read on this forum, it is suggested that the strike objects be grounded directly to the base of the secondary (also to ground wire in the 120v source) so as eliminate extraneous ground loops running through the house wiring. Makes sense in theory, but when I do that, a strike to the grounded object causes fluorescent lights to flicker, tv's goes haywire etc. When I let the arc strike to an object just sitting on the concrete floor it does not interfere with any of the house wiring.
Any explanation for why this seems to be opposite from what I read? I have the strike ring around the primary connected to the base of the secondary and ground, but I can see I would want that.
Re: Grounded strike target
Steve Conner, Tue Apr 01 2014, 07:50AM
The capacitance that stores the Tesla coil's energy is between the topload and the surroundings (walls of the room etc) Any sudden discharge of this capacitance will make a high transient current flow in the surroundings. It will invariably get into the house wiring as this is the most conductive part of the surroundings.
If the arc strikes an ungrounded object, there is no low-impedance return path to the surroundings, so the topload doesn't discharge suddenly.
You might think that grounding the strike object to the secondary base "completes the circuit" and all the discharge current circulates in that loop without getting into the wiring. This is only true for current at the coil's resonant frequency. The spike of current from a ground strike contains lots of high frequencies which can't pass through the secondary coil because of its high inductance. So they flow from the strike target straight out of the secondary base ground wire, and return to the topload via the surroundings.
Steve Conner, Tue Apr 01 2014, 07:50AM
The capacitance that stores the Tesla coil's energy is between the topload and the surroundings (walls of the room etc) Any sudden discharge of this capacitance will make a high transient current flow in the surroundings. It will invariably get into the house wiring as this is the most conductive part of the surroundings.
If the arc strikes an ungrounded object, there is no low-impedance return path to the surroundings, so the topload doesn't discharge suddenly.
You might think that grounding the strike object to the secondary base "completes the circuit" and all the discharge current circulates in that loop without getting into the wiring. This is only true for current at the coil's resonant frequency. The spike of current from a ground strike contains lots of high frequencies which can't pass through the secondary coil because of its high inductance. So they flow from the strike target straight out of the secondary base ground wire, and return to the topload via the surroundings.
Re: Grounded strike target
Graham Armitage, Tue Apr 01 2014, 10:00AM
So if a strike to an ungrounded object does not fully discharge the top-load that would suggest a lower current draw and as you say, less high frequency current returning via the surroundings. Two good things I imagine, so why is it always suggested that we ground the strike object? What is the benefit?
Graham Armitage, Tue Apr 01 2014, 10:00AM
So if a strike to an ungrounded object does not fully discharge the top-load that would suggest a lower current draw and as you say, less high frequency current returning via the surroundings. Two good things I imagine, so why is it always suggested that we ground the strike object? What is the benefit?
Re: Grounded strike target
Steve Conner, Tue Apr 01 2014, 10:09AM
I have no idea. When doing public demos I try to use a full Faraday cage, and if that's not possible I try to avoid arcing to ground altogether. Blazing away at a grounded target is a sure-fire way to generate lots of nasty EMI.
Steve Conner, Tue Apr 01 2014, 10:09AM
I have no idea. When doing public demos I try to use a full Faraday cage, and if that's not possible I try to avoid arcing to ground altogether. Blazing away at a grounded target is a sure-fire way to generate lots of nasty EMI.
Re: Grounded strike target
Graham Armitage, Tue Apr 01 2014, 11:02AM
Thanks Steve, that's good to hear, and to confirm what I was seeing and doing. How about the base of the secondary and strike ring - do you connect those together and ground them to the mains supply or do you use a separate RF ground?
Graham Armitage, Tue Apr 01 2014, 11:02AM
Thanks Steve, that's good to hear, and to confirm what I was seeing and doing. How about the base of the secondary and strike ring - do you connect those together and ground them to the mains supply or do you use a separate RF ground?
Re: Grounded strike target
Steve Conner, Tue Apr 01 2014, 11:09AM
Secondary base and strike ring grounding is a tricky issue. For protection of DRSSTC electronics, you want all of the grounds tied together, but for EMI you want to separate the RF grounds from the incoming power supply.
Nowadays I err on the side of protection, since as I explained above, separating the grounds doesn't really help EMI as most of the really bad stuff gets capacitively coupled straight into the surroundings from the topload anyway.
So, I tie all of the grounds together at the base of the coil. Secondary base ground, strike rails, IGBT heatsink, EMI line filter, and the AC line green wire. I also bypass the DC bus directly to the IGBT heatsink using a couple of 0.1uF capacitors, to give a direct return path for any primary strikes. Otherwise they can punch through your gate drive transformers or the IGBTs' isolation barriers and the resulting follow-through current from the DC bus will blow your setup to smithereens.
When running in a Faraday cage (more often than not nowadays) I also bond all of the above grounds to the cage.
Steve Conner, Tue Apr 01 2014, 11:09AM
Secondary base and strike ring grounding is a tricky issue. For protection of DRSSTC electronics, you want all of the grounds tied together, but for EMI you want to separate the RF grounds from the incoming power supply.
Nowadays I err on the side of protection, since as I explained above, separating the grounds doesn't really help EMI as most of the really bad stuff gets capacitively coupled straight into the surroundings from the topload anyway.
So, I tie all of the grounds together at the base of the coil. Secondary base ground, strike rails, IGBT heatsink, EMI line filter, and the AC line green wire. I also bypass the DC bus directly to the IGBT heatsink using a couple of 0.1uF capacitors, to give a direct return path for any primary strikes. Otherwise they can punch through your gate drive transformers or the IGBTs' isolation barriers and the resulting follow-through current from the DC bus will blow your setup to smithereens.
When running in a Faraday cage (more often than not nowadays) I also bond all of the above grounds to the cage.
Re: Grounded strike target
Graham Armitage, Tue Apr 01 2014, 11:19AM
Wow - that's good info. I like the idea of the protecting the circuits with the 0.1uF caps - that's a bonus. Other than that I do have a similar grounding config. Would love to see your faraday cage setup sometime.
Thanks
Graham Armitage, Tue Apr 01 2014, 11:19AM
Wow - that's good info. I like the idea of the protecting the circuits with the 0.1uF caps - that's a bonus. Other than that I do have a similar grounding config. Would love to see your faraday cage setup sometime.
Thanks
Re: Grounded strike target
Kizmo, Tue Apr 01 2014, 03:12PM
I can confirm that the grounding method Steve suggested works very well with big DRSSTC too. And by big i mean my BiggerDR (
)
Kizmo, Tue Apr 01 2014, 03:12PM
I can confirm that the grounding method Steve suggested works very well with big DRSSTC too. And by big i mean my BiggerDR (
)Re: Grounded strike target
Graham Armitage, Tue Apr 01 2014, 05:10PM
That's an impressive coil. Looking at the video it seems you have a cable from the ladder going back toward the coil. If that is not a grounding wire, as Steve advises against, what is it and where is it going?
Graham Armitage, Tue Apr 01 2014, 05:10PM
That's an impressive coil. Looking at the video it seems you have a cable from the ladder going back toward the coil. If that is not a grounding wire, as Steve advises against, what is it and where is it going?
Re: Grounded strike target
Steve Conner, Tue Apr 01 2014, 05:15PM
I think it is a grounding wire and Kizmo just doesn't care about EMI as he is in the middle of a forest in Finland.
Steve Conner, Tue Apr 01 2014, 05:15PM
I think it is a grounding wire and Kizmo just doesn't care about EMI as he is in the middle of a forest in Finland.
Re: Grounded strike target
Kizmo, Tue Apr 01 2014, 05:28PM
Yeah. Frozen ground is really poor conductor and even at TC voltages i wont get much ground strikes at all without that ground wire. And it does produce some nasty EMI indeed. I had to faraday cage my microphone to stop all those crack and pops during heavier ground strikes
Kizmo, Tue Apr 01 2014, 05:28PM
Yeah. Frozen ground is really poor conductor and even at TC voltages i wont get much ground strikes at all without that ground wire. And it does produce some nasty EMI indeed. I had to faraday cage my microphone to stop all those crack and pops during heavier ground strikes
Re: Grounded strike target
Graham Armitage, Tue Apr 01 2014, 05:34PM
Aaahhh - that makes sense. Forests don't care about EMI All of this helps me understand why I get hanging notes on the midi controller when it hits the strike ring. USB cable is picking up all sorts of noise. Good stuff - thanks guys.
Graham Armitage, Tue Apr 01 2014, 05:34PM
Aaahhh - that makes sense. Forests don't care about EMI
All of this helps me understand why I get hanging notes on the midi controller when it hits the strike ring. USB cable is picking up all sorts of noise. Good stuff - thanks guys.Re: Grounded strike target
Uspring, Wed Apr 02 2014, 07:40AM
Did anybody try putting a ferrite bead on the breakout rod? Might flash over during a ground strike, but perhaps the high frequency components don't carry that much voltage.
Uspring, Wed Apr 02 2014, 07:40AM
Did anybody try putting a ferrite bead on the breakout rod? Might flash over during a ground strike, but perhaps the high frequency components don't carry that much voltage.
Re: Grounded strike target
Steve Conner, Wed Apr 02 2014, 08:06AM
Arc Attack reported good results using a breakout point with an air-cored inductor built in, like a small Tesla secondary. It was enclosed in an oil-filled plexi tube with insulating ribs on the outside to discourage flashover.
It apparently made a substantial improvement in EMI and they now use it in all of their shows.
Steve Conner, Wed Apr 02 2014, 08:06AM
Arc Attack reported good results using a breakout point with an air-cored inductor built in, like a small Tesla secondary. It was enclosed in an oil-filled plexi tube with insulating ribs on the outside to discourage flashover.
It apparently made a substantial improvement in EMI and they now use it in all of their shows.
Re: Grounded strike target
Hydron, Wed Apr 02 2014, 10:11AM
USB is awfully susceptible to EMI - it uses single ended signalling for some functions (e.g. bus reset) and tends to form ground loops as well. This causes all sorts of problems when running in a tesla coil type EMI environment, especially with poorly made devices such as the $5 chinese MIDI cables (note that you don't get ground loops with MIDI - it uses optical isolation when done right).
I had issues with the connection dropping out when using a USB oscilloscope to make measurements of my DRSSTC - I improved it a bit with ferrite beads etc, but the real solution was to change the scope connection to ethernet, which is fully differential, AC coupled (via pulse transformers) and does not form ground loops when using UTP. An added benefit is that I can now plug it into a wifi router and use it remotely, enabling me to wirelessly scope the coil waveforms while it's running, or (with battery power) float the scope with only a few pF to ground.
Hydron, Wed Apr 02 2014, 10:11AM
Bushman wrote ...
Aaahhh - that makes sense. Forests don't care about EMI All of this helps me understand why I get hanging notes on the midi controller when it hits the strike ring. USB cable is picking up all sorts of noise. Good stuff - thanks guys.
Aaahhh - that makes sense. Forests don't care about EMI
All of this helps me understand why I get hanging notes on the midi controller when it hits the strike ring. USB cable is picking up all sorts of noise. Good stuff - thanks guys.USB is awfully susceptible to EMI - it uses single ended signalling for some functions (e.g. bus reset) and tends to form ground loops as well. This causes all sorts of problems when running in a tesla coil type EMI environment, especially with poorly made devices such as the $5 chinese MIDI cables (note that you don't get ground loops with MIDI - it uses optical isolation when done right).
I had issues with the connection dropping out when using a USB oscilloscope to make measurements of my DRSSTC - I improved it a bit with ferrite beads etc, but the real solution was to change the scope connection to ethernet, which is fully differential, AC coupled (via pulse transformers) and does not form ground loops when using UTP. An added benefit is that I can now plug it into a wifi router and use it remotely, enabling me to wirelessly scope the coil waveforms while it's running, or (with battery power) float the scope with only a few pF to ground.
Re: Grounded strike target
Graham Armitage, Wed Apr 02 2014, 11:34AM
These last two posts are really cool - great info that is not widely broadcast. I would love to know more about the configuration of the arc attack breakout. Any details on the web that you know of.
Glad to hear its not just me having trouble with USB. I do admit i have a cheap USB/MIDI cable, but didn't want to drop more $$ until I understood the problem. This helps - thanks.
Graham Armitage, Wed Apr 02 2014, 11:34AM
These last two posts are really cool - great info that is not widely broadcast. I would love to know more about the configuration of the arc attack breakout. Any details on the web that you know of.
Glad to hear its not just me having trouble with USB. I do admit i have a cheap USB/MIDI cable, but didn't want to drop more $$ until I understood the problem. This helps - thanks.
Print this page