Primary Current Feedback
Weston, Thu Apr 07 2016, 10:12PMI am working on a high frequency tesla coil and was curious why primary current feedback is not used for single resonant coils (SSTC, not DRSSTC).
I seem to remember reading in a thread a long time ago that primary current feedback is unstable or something similar, but I have been able to find any posts of that nature. However, I have also failed to find any non SSTCs that use primary current feedback...
It appears that depending on the loading of the coil, maximum voltage gain does not correspond to any specific phase relationship, but it is "close enough" where you could have it track some positive phase relationship and get decent voltage gain.
I am working on a class DE coil where I want to maintain an inductive load on the inverter, so I want to directly control the phase between primary current and inverter output voltage anyways.
I just wanted to check if there is any reason why primary current feedback on a non DR coil will not work, such as additional resonator parasitic which would introduce multiple zones that the PLL could lock on to and potentially cause instability.
Re: Primary Current Feedback
Sigurthr, Thu Apr 07 2016, 11:55PM
It's not used in SR coils because there's nothing to force the primary to oscillate at the f0 of the secondary. In a SR topology primary feedback would cause oscillation at the primary's natural resonance point, not the secondary's. If you make the primary f0 match the secondary f0 you've got a DR coil, not a SR coil. If you make the primary resonant with LC but not the secondary, you've got a fancy air-core ZVS, not a TC.
Sigurthr, Thu Apr 07 2016, 11:55PM
It's not used in SR coils because there's nothing to force the primary to oscillate at the f0 of the secondary. In a SR topology primary feedback would cause oscillation at the primary's natural resonance point, not the secondary's. If you make the primary f0 match the secondary f0 you've got a DR coil, not a SR coil. If you make the primary resonant with LC but not the secondary, you've got a fancy air-core ZVS, not a TC.
Re: Primary Current Feedback
Weston, Fri Apr 08 2016, 12:36AM
What do you mean by the the primary's natural resonance point? There is none if its only single resonant, the only thing in can lock to is the secondaries resonant point.
Based on the simulation of a resonant secondary coupled to a primary I get a single pole and a single zero. It should be possible to lock to these with a PLL. Its ~ the same as the upper pole in a DRSSTC.
Is the issue related to Q? During a ground arc the system could become over damped and fall out of oscillation / blow up. This will not happen with a DRSSTC system as you still have the primary resonance to lock on to.
Weston, Fri Apr 08 2016, 12:36AM
What do you mean by the the primary's natural resonance point? There is none if its only single resonant, the only thing in can lock to is the secondaries resonant point.
Based on the simulation of a resonant secondary coupled to a primary I get a single pole and a single zero. It should be possible to lock to these with a PLL. Its ~ the same as the upper pole in a DRSSTC.
Is the issue related to Q? During a ground arc the system could become over damped and fall out of oscillation / blow up. This will not happen with a DRSSTC system as you still have the primary resonance to lock on to.
Re: Primary Current Feedback
GrantX, Sat Apr 09 2016, 07:33AM
The primary coil will have a self-resonant frequency, due to the parasitic inter-winding and environmental capacitance (same as secondary and any other coil). But it's probably in the region of tens of MHz at least, several orders of magnitude higher than the secondary. Would it be possible to use a low pass filter on the output of the primary feedback CT? It could be set up so everything beyond the secondary's first harmonic is smothered out.
GrantX, Sat Apr 09 2016, 07:33AM
The primary coil will have a self-resonant frequency, due to the parasitic inter-winding and environmental capacitance (same as secondary and any other coil). But it's probably in the region of tens of MHz at least, several orders of magnitude higher than the secondary. Would it be possible to use a low pass filter on the output of the primary feedback CT? It could be set up so everything beyond the secondary's first harmonic is smothered out.
Re: Primary Current Feedback
Sigurthr, Sat Apr 09 2016, 07:55AM
Weston, there are ALWAYS resonances; to everything from everything when dealing with RF. Even the distance between your hands will have a differential capacitance, and this gap will have its own resonance frequency (and a frequency each for between each hand and the TC). The key to working well with RF is to look for and consider resonances outside of your desired one.
GrantX, it would be interesting to see how well that works out. I wonder though if you'd really be seeing primary current unless you shielded the feedback extremely well. I think that poor or no shielding would result in capacitive feedback from the secondary "leaking in", since it is likely to be orders of magnitude stronger than the weak primary current's signal (remember the primary impedance would degrade the amplitude of the secondary f0). You'd have an antenna with a bad phase alteration in that case I think. I think you'd have to filter at the probe too, not just filter at the driver, as the filter would pass the capacitively-leaked-in secondary feedback.
Sigurthr, Sat Apr 09 2016, 07:55AM
Weston, there are ALWAYS resonances; to everything from everything when dealing with RF. Even the distance between your hands will have a differential capacitance, and this gap will have its own resonance frequency (and a frequency each for between each hand and the TC). The key to working well with RF is to look for and consider resonances outside of your desired one.
GrantX, it would be interesting to see how well that works out. I wonder though if you'd really be seeing primary current unless you shielded the feedback extremely well. I think that poor or no shielding would result in capacitive feedback from the secondary "leaking in", since it is likely to be orders of magnitude stronger than the weak primary current's signal (remember the primary impedance would degrade the amplitude of the secondary f0). You'd have an antenna with a bad phase alteration in that case I think. I think you'd have to filter at the probe too, not just filter at the driver, as the filter would pass the capacitively-leaked-in secondary feedback.
Re: Primary Current Feedback
Uspring, Sun Apr 10 2016, 01:17PM
Phase shift between primary voltage and current in a SSTC is mostly 90 degrees except for a dip around secondary resonance. For low couplings and low secondary Q, e.g. at strong arc loading, this dip can be small. Phase shift between primary voltage and secondary voltage or current depends much stronger on frequency, so the choice of the PLL reference phase is less critical.
The PLL also needs to know, whether to increase or decrease the frequency. That's much more easy for a non dipping frequency - phase relationship.
Uspring, Sun Apr 10 2016, 01:17PM
Phase shift between primary voltage and current in a SSTC is mostly 90 degrees except for a dip around secondary resonance. For low couplings and low secondary Q, e.g. at strong arc loading, this dip can be small. Phase shift between primary voltage and secondary voltage or current depends much stronger on frequency, so the choice of the PLL reference phase is less critical.
The PLL also needs to know, whether to increase or decrease the frequency. That's much more easy for a non dipping frequency - phase relationship.
Print this page