non-DR QCW's
Dr. ISOTOP, Mon Apr 02 2012, 06:53PMDo QCW coils need to be double-resonant to get good spark-to-power ratio? From Steve Ward's measurements we know that the topload voltages are only ~50KV, so it appears that the resonant voltage and current buildup that makes a DRSSTC so wonderful is not necessary.
Re: non-DR QCW's
teravolt, Tue Apr 03 2012, 01:03AM
hmmmm
teravolt, Tue Apr 03 2012, 01:03AM
hmmmm
Re: non-DR QCW's
Gregory, Tue Apr 03 2012, 03:49AM
I think that you need a a drsstc because the ressonant swicthing, and, in a sstc you will have some reatance reflected on the primary. This question come in good hour I'm making one QCW, the buck is working good and now I'm working on a sstc to test the buck, so, lets see the results
Gregory, Tue Apr 03 2012, 03:49AM
I think that you need a a drsstc because the ressonant swicthing, and, in a sstc you will have some reatance reflected on the primary. This question come in good hour I'm making one QCW, the buck is working good and now I'm working on a sstc to test the buck, so, lets see the results
Re: non-DR QCW's
Goodchild, Tue Apr 03 2012, 06:05AM
From my experiments I would say you need the DR effect in a QCW. YOU can look at two similar examples a SSTC drive with half rec mains and a VTTC with half rec mains. Both systems are using the same type of input waveform into there main switching section, but the VTTC has one difference from the SSTC, a resonant tank capacitor. I personally think that the higher currents seen in the primary of a QCW are important in the generation of the long sparks because of there inherent current drive nature. i.e. length is more a function of drive current and time than voltage.
Although the currents seen in a VTTC are much lower than a QCW it has a much higher drive voltage to make up the difference.
I don't see this effect being possible with a SSTC unless a insanely high voltage was used or a supper low primary Z. In ether of these cases the SSTC would become a resonant system because the parasitics would then become the resonant component of the primary.
I'm probably totally wrong here so yeah um, yeah! :P
Goodchild, Tue Apr 03 2012, 06:05AM
From my experiments I would say you need the DR effect in a QCW. YOU can look at two similar examples a SSTC drive with half rec mains and a VTTC with half rec mains. Both systems are using the same type of input waveform into there main switching section, but the VTTC has one difference from the SSTC, a resonant tank capacitor. I personally think that the higher currents seen in the primary of a QCW are important in the generation of the long sparks because of there inherent current drive nature. i.e. length is more a function of drive current and time than voltage.
Although the currents seen in a VTTC are much lower than a QCW it has a much higher drive voltage to make up the difference.
I don't see this effect being possible with a SSTC unless a insanely high voltage was used or a supper low primary Z. In ether of these cases the SSTC would become a resonant system because the parasitics would then become the resonant component of the primary.
I'm probably totally wrong here so yeah um, yeah! :P
Re: non-DR QCW's
Dr. ISOTOP, Tue Apr 03 2012, 07:42AM
I wonder whether QCW's like high primary currents or long pulses more, i.e. for a given modulator capacity, should one aim for reasonably long ramps that hit a very high Ipk, or long, slow ramps with low slope.
I'm actually sort of inclined to believe its the latter right now, given how low primary risetimes seem to facilitate spark growth. Has anyone tried playing with this?
Dr. ISOTOP, Tue Apr 03 2012, 07:42AM
I wonder whether QCW's like high primary currents or long pulses more, i.e. for a given modulator capacity, should one aim for reasonably long ramps that hit a very high Ipk, or long, slow ramps with low slope.
I'm actually sort of inclined to believe its the latter right now, given how low primary risetimes seem to facilitate spark growth. Has anyone tried playing with this?
Re: non-DR QCW's
Avalanche, Tue Apr 03 2012, 08:07AM
I'm building a QCW as an SSTC, or I was - it's been on hold since Christmas!
My theory is that with an SSTC, streamer output is pretty much proportional to DC bus voltage anyway, so my QCW will simply modulate the DC bus voltage (no current feedback necessary). Because streamer growth will be in the rising part of the voltage envelope, I can get away without a feedback loop at all - as it doesn't matter what happens to the DC bus voltage after the streamer has reached maximum length (as long as it dies back down before the next streamer).
Might work
Edit - I should clarify, by "streamer output" I mean total streamer mass - or total volume of plasma. By streamer growth, I mean the shape given to that given volume of streamer mass caused by the shape of the voltage envelope, or it's rate of rise.
Avalanche, Tue Apr 03 2012, 08:07AM
I'm building a QCW as an SSTC, or I was - it's been on hold since Christmas!
My theory is that with an SSTC, streamer output is pretty much proportional to DC bus voltage anyway, so my QCW will simply modulate the DC bus voltage (no current feedback necessary). Because streamer growth will be in the rising part of the voltage envelope, I can get away without a feedback loop at all - as it doesn't matter what happens to the DC bus voltage after the streamer has reached maximum length (as long as it dies back down before the next streamer).
Might work
Edit - I should clarify, by "streamer output" I mean total streamer mass - or total volume of plasma. By streamer growth, I mean the shape given to that given volume of streamer mass caused by the shape of the voltage envelope, or it's rate of rise.
Re: non-DR QCW's
Steve Conner, Tue Apr 03 2012, 09:06AM
Steve Ward's original QCW used DC bus voltage modulation without any current feedback, so that is the approved way of doing it. My experiments with current feedback were none too impressive, and I plan to stop using it.
Steve Conner, Tue Apr 03 2012, 09:06AM
Steve Ward's original QCW used DC bus voltage modulation without any current feedback, so that is the approved way of doing it. My experiments with current feedback were none too impressive, and I plan to stop using it.
Re: non-DR QCW's
Goodchild, Tue Apr 03 2012, 03:26PM
Well basically if you give the QCW to much PW and not enough voltage you get very fat stubby sparks. but on the other hand if you don't give it enough PW at high voltage the spark with branch. It is a carful balancing act between voltage and PW as you want to make the spark longer you add voltage, but at the same time you also need to increase PW to give the spark the necessary time to heat up the air or it will simply branch out. This is the only way I know of to keep the plasma column straight.
Well correction, you can also use a larger e-field (bigger toroid) to keep it straight, this also needs less voltage.
Goodchild, Tue Apr 03 2012, 03:26PM
bwang wrote ...
I wonder whether QCW's like high primary currents or long pulses more, i.e. for a given modulator capacity, should one aim for reasonably long ramps that hit a very high Ipk, or long, slow ramps with low slope.
I'm actually sort of inclined to believe its the latter right now, given how low primary risetimes seem to facilitate spark growth. Has anyone tried playing with this?
I wonder whether QCW's like high primary currents or long pulses more, i.e. for a given modulator capacity, should one aim for reasonably long ramps that hit a very high Ipk, or long, slow ramps with low slope.
I'm actually sort of inclined to believe its the latter right now, given how low primary risetimes seem to facilitate spark growth. Has anyone tried playing with this?
Well basically if you give the QCW to much PW and not enough voltage you get very fat stubby sparks. but on the other hand if you don't give it enough PW at high voltage the spark with branch. It is a carful balancing act between voltage and PW as you want to make the spark longer you add voltage, but at the same time you also need to increase PW to give the spark the necessary time to heat up the air or it will simply branch out. This is the only way I know of to keep the plasma column straight.
Well correction, you can also use a larger e-field (bigger toroid) to keep it straight, this also needs less voltage.
Re: non-DR QCW's
Gregory, Tue Apr 03 2012, 04:07PM
I will stop thinking about and try it. yestardey I made a full bridge with two iifp460 in parallel for each switch. I plan to use 30 to 300v or more, last see what hapens.
Gregory, Tue Apr 03 2012, 04:07PM
I will stop thinking about and try it. yestardey I made a full bridge with two iifp460 in parallel for each switch. I plan to use 30 to 300v or more, last see what hapens.
Re: non-DR QCW's
Gregory, Thu Apr 05 2012, 03:37AM
I tryed it and I saw that the modulator was making his job bad. I think that is because the reative load of the sstc primary
Gregory, Thu Apr 05 2012, 03:37AM
I tryed it and I saw that the modulator was making his job bad. I think that is because the reative load of the sstc primary
Re: non-DR QCW's
Steve Ward, Sat Apr 07 2012, 05:14AM
The DR part is mainly just a convenient way to get a higher system Q without having to store all the energy in the secondary... that is to say, you could make a single resonant coil, but its gotta be much lower impedance to work as well as a DRSSTC would.
I think a regular SSTC (single resonant) would work OK if designed right. The secondary impedance needs to be a lot lower than what most people use, that means lots of toroid capacitance and very little secondary inductance. Id probably try getting the reactance at resonance to be less than 20K ohms as a start (it might benefit from being even a lot lower depending on the drivers output impedance). Keeping the Fres above 320khz or so seems to allow for the branchless spark growth, which seems to primarily be due to lower voltage spark growth. Going higher than maybe 360khz seemed to have no benefit in my experiments, though may be useful because it lowers the required output voltage per spark length. My straight sparks needed about 65kV peak when grown to 4 feet, at 365khz.
I think i got lucky that the bus modulation worked so well at controlling the spark power. If the DRSSTC was a stiffer voltage source, it wouldnt work well at all, and then i think current regulator would be better. Anyway, dont stop using your control technique until you try the higher frequency coil first! Or at least mess with the ramp parameters to be sure it doesnt work.
Steve Ward, Sat Apr 07 2012, 05:14AM
The DR part is mainly just a convenient way to get a higher system Q without having to store all the energy in the secondary... that is to say, you could make a single resonant coil, but its gotta be much lower impedance to work as well as a DRSSTC would.
I think a regular SSTC (single resonant) would work OK if designed right. The secondary impedance needs to be a lot lower than what most people use, that means lots of toroid capacitance and very little secondary inductance. Id probably try getting the reactance at resonance to be less than 20K ohms as a start (it might benefit from being even a lot lower depending on the drivers output impedance). Keeping the Fres above 320khz or so seems to allow for the branchless spark growth, which seems to primarily be due to lower voltage spark growth. Going higher than maybe 360khz seemed to have no benefit in my experiments, though may be useful because it lowers the required output voltage per spark length. My straight sparks needed about 65kV peak when grown to 4 feet, at 365khz.
Steve Ward's original QCW used DC bus voltage modulation without any current feedback, so that is the approved way of doing it. My experiments with current feedback were none too impressive, and I plan to stop using it.
I think i got lucky that the bus modulation worked so well at controlling the spark power. If the DRSSTC was a stiffer voltage source, it wouldnt work well at all, and then i think current regulator would be better. Anyway, dont stop using your control technique until you try the higher frequency coil first! Or at least mess with the ramp parameters to be sure it doesnt work.
Re: non-DR QCW's
hvman, Sat Apr 07 2012, 02:06PM
I tested my half bridge SSTC with a 580 kHz seconadary and I achieved up to 1 foot sword-like sparks. I don't use modulator, coil run with a synced interrupter at half wave rectified mains.
hvman, Sat Apr 07 2012, 02:06PM
I tested my half bridge SSTC with a 580 kHz seconadary and I achieved up to 1 foot sword-like sparks. I don't use modulator, coil run with a synced interrupter at half wave rectified mains.
Re: non-DR QCW's
Mads Barnkob, Sat Apr 07 2012, 02:55PM
I have been and am working on a DRSSTC version of the same, a synched interupter to conduct for the first 5ms of the 50Hz period, but its trapped in a box while im moving :)
Mads Barnkob, Sat Apr 07 2012, 02:55PM
hvman wrote ...
I tested my half bridge SSTC with a 580 kHz seconadary and I achieved up to 1 foot sword-like sparks. I don't use modulator, coil run with a synced interrupter at half wave rectified mains.
I tested my half bridge SSTC with a 580 kHz seconadary and I achieved up to 1 foot sword-like sparks. I don't use modulator, coil run with a synced interrupter at half wave rectified mains.
I have been and am working on a DRSSTC version of the same, a synched interupter to conduct for the first 5ms of the 50Hz period, but its trapped in a box while im moving :)
Print this page