Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sat Nov 19 2011, 05:44PMBy now every lover of Tesla coils probably knows about Steve Ward's QCW. If not you ought to check this out, it is awesome.
The QCW is a dual resonant solid-state design. The long, elegant sword sparks are made using a long burst of RF, some 10-15ms. A highish frequency (over 300kHz) is used, and a ramp-shaped modulation envelope is applied.
The modulation is currently done by ramping the DC bus voltage using a synchronous buck converter, similar to a Class-D audio amplifier but much more powerful. (Steve's QCW used a bigass IGBT brick for the modulator.)
Anyway, enough background. My idea is to combine the Class-D modulator into the existing SSTC bridge. Instead of pulsing an IGBT in a modulator on and off, I intend to just pulse the RF on and off, and let the Q of the Tesla resonator smooth it all out, at least enough for government work.
I call it sigma-delta modulation because as far as I can see, that's basically what it is.My Mk2 PLL driver already has a current limiter that works like this, so it is basically a case of replacing the current limit trimpot with a ramp generator. I've decided to use my Mjollnir DRSSTC as a guinea pig for this.
^Mjollnir's driver board removed and hacked to add ramp generator.
^A "Ramp Time" dial added to the remote control.
^ Ramp voltages from the generator. (I use a window comparator instead of rectifying the primary current, so I need two voltages.)
^Low power test run in the lab, with DC bus reduced to 160V and secondary short-circuited, as I don't want to make any sparks right now. if one hit the LeCroy I would be in big trouble- it's not mine
^At the start of the ramp, the primary current is discontinuous- a series of "banglets" of 1 and 2 cycles. Hopefully when the secondary is resonating, it will blend them together into one bang.
^By the end it has sorted itself out and you can see a nice ramp up to approximately 250A peak here. The beginning is actually better than it looks. Most of the amplitude is just "hair" from switching spikes picked up by the CT. And, you can see the DC bus capacitor running out of energy at the end, collapsing the voltage waveform.
On Monday we will take it to the uni HV lab, cameras will be rolling and a fire truck on standby!
(Monday...)
Well here are the HV lab test results as promised...
The weird thing is that the sparks fundamentally don't look that different to how they did without the ramp! Either the ramp generator is really bad (the scope shows the primary current, and as you can see it is quite bad
) or the 220kHz resonant frequency affects the spark appearance so strongly that no amount of envelope shaping can undo it.They look pretty big and gnarly though, and no components blew up, so I think it's at least somewhat of a result. I intend to go on and improve the modulation somewhat, then tighten the coupling and add more DC bus caps and see if the sparks can be got any bigger that way.
Then I'll build a new resonator with a frequency above 320kHz, which according to Steve Ward is the magic frequency where the sparks suddenly go swordlike. If I don't get sword sparks, this will prove that the modulator isn't good enough.
The tank cap used in this coil is 2 strings of 10 470nF, 1kV PP film/foil caps, for a total of 0.1uF. I've managed to get hold of 10 conduction cooled mica caps rated at 22nF, 1kV RMS, 75A RMS, to make a new "MMC From Hell".
Stay tuned for more updates, if you pardon the pun!
Schematic of ramp modulated driver
]dr111127_ramp.pdf[/file]
Re: Sigma delta modulated QCW- bang without the buck?
Avalanche, Sat Nov 19 2011, 06:59PM
Great looking setup! I think that has answered all the questions I had floating round in my head, I was getting confused by different QCW topologies on the other thread.
So if you're applying delta-sigma modulation to the gate drive instead of modulating the bus voltage during the burst, what effect does primary/secondary coupling have - especially when the modulation is low? I would have thought that the H dridge, in it's off state would try to damp the output from the resonator, making the output spark size very much non-linear with regards to modulation (except at 100% when you're basically at CW)
Probably, maybe, I dunno
Edit: do you have by any chance a scope shot like the last one, but with B & A zoomed on the beginning of the pulse? ...or where abouts is the second to last scope shot in relation to the 10mS pulse?
Avalanche, Sat Nov 19 2011, 06:59PM
Great looking setup! I think that has answered all the questions I had floating round in my head, I was getting confused by different QCW topologies on the other thread.
So if you're applying delta-sigma modulation to the gate drive instead of modulating the bus voltage during the burst, what effect does primary/secondary coupling have - especially when the modulation is low? I would have thought that the H dridge, in it's off state would try to damp the output from the resonator, making the output spark size very much non-linear with regards to modulation (except at 100% when you're basically at CW)
Probably, maybe, I dunno
Edit: do you have by any chance a scope shot like the last one, but with B & A zoomed on the beginning of the pulse? ...or where abouts is the second to last scope shot in relation to the 10mS pulse?
Re: Sigma delta modulated QCW- bang without the buck?
dude_500, Sat Nov 19 2011, 07:14PM
Is the resistor in the top right corner of the board that looks like it's about to catch on fire a problem? :P
dude_500, Sat Nov 19 2011, 07:14PM
Is the resistor in the top right corner of the board that looks like it's about to catch on fire a problem? :P
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sat Nov 19 2011, 07:20PM
Hey Avalanche
There are so many unknowns! I don't really know the answers to these questions but I'm just going to fire it up and see what happens!
The way I see it, when the H bridge is on it functions as an inverter. When it's off, it functions as a rectifier. The power flow is the same magnitude, because the primary current can't change quickly because of the resonator Q. it just reverses direction. So you have +1, -1 lumps of RF going to the resonator, exactly like a "1-bit" sigma delta DAC in a CD player or whatever.
This is untrue for levels so low that the primary current dies out altogether. In that case the bridge doesn't rectify anything. But I think the error caused by this is in a direction that would improve the results.
Here's a shot from the very beginning of the burst. I should have done all of them like the last one, but it took me until the last one to remember how the LeCroy's zoom worked. You're seeing 1 cycle of "motor" followed by three-quarters of a cycle of "regen".
Dude500, that resistor got crisped years ago! I had trouble with the PLL chip latching up and sending DC to the gate drive transformer. That's the damping resistor across the DC block cap, so the DC cooked it. I got the latchup fixed, but when I measured the resistor it still seemed OK, so I didn't replace it.
Steve Conner, Sat Nov 19 2011, 07:20PM
Hey Avalanche
There are so many unknowns! I don't really know the answers to these questions but I'm just going to fire it up and see what happens!
The way I see it, when the H bridge is on it functions as an inverter. When it's off, it functions as a rectifier. The power flow is the same magnitude, because the primary current can't change quickly because of the resonator Q. it just reverses direction. So you have +1, -1 lumps of RF going to the resonator, exactly like a "1-bit" sigma delta DAC in a CD player or whatever.
This is untrue for levels so low that the primary current dies out altogether. In that case the bridge doesn't rectify anything. But I think the error caused by this is in a direction that would improve the results.
Here's a shot from the very beginning of the burst. I should have done all of them like the last one, but it took me until the last one to remember how the LeCroy's zoom worked.
You're seeing 1 cycle of "motor" followed by three-quarters of a cycle of "regen".
Dude500, that resistor got crisped years ago! I had trouble with the PLL chip latching up and sending DC to the gate drive transformer. That's the damping resistor across the DC block cap, so the DC cooked it. I got the latchup fixed, but when I measured the resistor it still seemed OK, so I didn't replace it.
Re: Sigma delta modulated QCW- bang without the buck?
Dr. Dark Current, Sat Nov 19 2011, 08:59PM
Interesting project, hope you get it working well
Dr. Dark Current, Sat Nov 19 2011, 08:59PM
Interesting project, hope you get it working well
Re: Sigma delta modulated QCW- bang without the buck?
Goodchild, Sat Nov 19 2011, 09:24PM
AWESOME. I'm glad some one finally tried it rather than our techno theoretical babble hehe
Steve this looks good so far, the ripple at the lowish frequency is still present as predicted by simulation. I'm curious how this will turn out with the secondary thrown into the mix.
The tricky part is going to be how this reacts with a higher coupling. I have a feeling that coupling is going to play a big role in how the final system operates.
What is your current resonator configuration? It seems like it's rather low coupling low Z and low freq like a regular DR?
Great work, keep us updated I'm itching to see how it turns out!
Eric
Goodchild, Sat Nov 19 2011, 09:24PM
AWESOME. I'm glad some one finally tried it rather than our techno theoretical babble hehe
Steve this looks good so far, the ripple at the lowish frequency is still present as predicted by simulation. I'm curious how this will turn out with the secondary thrown into the mix.
The tricky part is going to be how this reacts with a higher coupling. I have a feeling that coupling is going to play a big role in how the final system operates.
What is your current resonator configuration? It seems like it's rather low coupling low Z and low freq like a regular DR?
Great work, keep us updated
I'm itching to see how it turns out! Eric
Re: Sigma delta modulated QCW- bang without the buck?
JimmyH, Sun Nov 20 2011, 05:39AM
Do you have a guess at what the loaded Q is going to be?
I think Ward's loaded Q is pretty low, and the simulations we did showed a lot of ripple once you have some big sparks. I'm not sure at all how much ripple is acceptable, but if the goal is long straight sparks, I wouldn't be surprised if it had to be pretty smooth.
It seemed possible to do a multi-layer secondary (parallel) with an oil cap inside (only 60kv or so!) to really get the energy storage up there, but even that seemed borderline.
Doing some rough numbers, I think Wards coil was about 2amps at 60kv or 120kVA and 60kw (300V 200A in). at 400khz, thats ~65mj per half cycle, which is the energy stored in a 35pF cap at 60kv.
These numbers don't suggest a high loaded Q =\
I'm excited to see how it works!
JimmyH, Sun Nov 20 2011, 05:39AM
Do you have a guess at what the loaded Q is going to be?
I think Ward's loaded Q is pretty low, and the simulations we did showed a lot of ripple once you have some big sparks. I'm not sure at all how much ripple is acceptable, but if the goal is long straight sparks, I wouldn't be surprised if it had to be pretty smooth.
It seemed possible to do a multi-layer secondary (parallel) with an oil cap inside (only 60kv or so!) to really get the energy storage up there, but even that seemed borderline.
Doing some rough numbers, I think Wards coil was about 2amps at 60kv or 120kVA and 60kw (300V 200A in). at 400khz, thats ~65mj per half cycle, which is the energy stored in a 35pF cap at 60kv.
These numbers don't suggest a high loaded Q =\
I'm excited to see how it works!
Re: Sigma delta modulated QCW- bang without the buck?
Goodchild, Sun Nov 20 2011, 09:01AM
WOAH Jimmy is posting to 4hv that hasn't happened in 10,000 years :p
I agree that the Q of the system is going to suck rather hard when making sparks, I would figure this with most DRSSTC as well though.
As you mentioned my only concern with this drive is going to be that ripple! At <10KHz that seems nasty for sparks growth. Even with my system small glitches in class D ramp output (maybe 10 or 20V at the most) could really mess up your sparks! Branching is usually the thing that results.
I'm hopefully though because there is nothing I would like more than to get rid of all that extra hardware!
Goodchild, Sun Nov 20 2011, 09:01AM
JimmyH wrote ...
Do you have a guess at what the loaded Q is going to be?
I think Ward's loaded Q is pretty low, and the simulations we did showed a lot of ripple once you have some big sparks. I'm not sure at all how much ripple is acceptable, but if the goal is long straight sparks, I wouldn't be surprised if it had to be pretty smooth.
It seemed possible to do a multi-layer secondary (parallel) with an oil cap inside (only 60kv or so!) to really get the energy storage up there, but even that seemed borderline.
Doing some rough numbers, I think Wards coil was about 2amps at 60kv or 120kVA and 60kw (300V 200A in). at 400khz, thats ~65mj per half cycle, which is the energy stored in a 35pF cap at 60kv.
These numbers don't suggest a high loaded Q =\
I'm excited to see how it works!
Do you have a guess at what the loaded Q is going to be?
I think Ward's loaded Q is pretty low, and the simulations we did showed a lot of ripple once you have some big sparks. I'm not sure at all how much ripple is acceptable, but if the goal is long straight sparks, I wouldn't be surprised if it had to be pretty smooth.
It seemed possible to do a multi-layer secondary (parallel) with an oil cap inside (only 60kv or so!) to really get the energy storage up there, but even that seemed borderline.
Doing some rough numbers, I think Wards coil was about 2amps at 60kv or 120kVA and 60kw (300V 200A in). at 400khz, thats ~65mj per half cycle, which is the energy stored in a 35pF cap at 60kv.
These numbers don't suggest a high loaded Q =\
I'm excited to see how it works!
WOAH Jimmy is posting to 4hv that hasn't happened in 10,000 years :p
I agree that the Q of the system is going to suck rather hard when making sparks, I would figure this with most DRSSTC as well though.
As you mentioned my only concern with this drive is going to be that ripple! At <10KHz that seems nasty for sparks growth. Even with my system small glitches in class D ramp output (maybe 10 or 20V at the most) could really mess up your sparks! Branching is usually the thing that results.
I'm hopefully though because there is nothing I would like more than to get rid of all that extra hardware!
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sun Nov 20 2011, 10:59AM
WB, Jimmy!
To answer some of the questions:
The coil previously was set up as a DRSSTC. It operates on the upper pole at 220kHz, it has a 0.1uF tank cap, about 6 primary turns, and the tightest coupling possible without flashovers.
This is quite a high frequency and impedance as DRSSTCs go, and I found that it already seemed to like long bursts. I ran them up to 2ms with the current controller limiting to 400A, and got some really thick, gnarly sparks.
I take on board the comments about ripple, Q and so on, and I guess perfect sword sparks are too much to hope for. But I think it's reasonable to expect that the long burst sparks will straighten out and lengthen at least a bit, or branch less than they did before, or maybe make some cool new spark effects that nobody was expecting.
Another advantage, as Steve Ward pointed out to me, is that the controlled ramp-up should reduce the voltage stress on everything, allowing the coupling to be tightened. But maybe that's not a good thing!
The limiting factor in the long bursts was always the single 3300uF DC bus cap. Even charged to the full 320V, it was mostly empty by the end. I'm getting another 4!
Steve Conner, Sun Nov 20 2011, 10:59AM
WB, Jimmy!
To answer some of the questions:
The coil previously was set up as a DRSSTC. It operates on the upper pole at 220kHz, it has a 0.1uF tank cap, about 6 primary turns, and the tightest coupling possible without flashovers.
This is quite a high frequency and impedance as DRSSTCs go, and I found that it already seemed to like long bursts. I ran them up to 2ms with the current controller limiting to 400A, and got some really thick, gnarly sparks.
I take on board the comments about ripple, Q and so on, and I guess perfect sword sparks are too much to hope for. But I think it's reasonable to expect that the long burst sparks will straighten out and lengthen at least a bit, or branch less than they did before, or maybe make some cool new spark effects that nobody was expecting.
Another advantage, as Steve Ward pointed out to me, is that the controlled ramp-up should reduce the voltage stress on everything, allowing the coupling to be tightened. But maybe that's not a good thing!
The limiting factor in the long bursts was always the single 3300uF DC bus cap. Even charged to the full 320V, it was mostly empty by the end. I'm getting another 4!
Re: Sigma delta modulated QCW- bang without the buck?
teravolt, Mon Nov 21 2011, 08:39PM
Do you plan to build another tesla with a higher f res if you get brancing sparks? The two things that have been mensioned is F res 350-500khz and a ramping power envilope to make sword sparks. what type of caps does ^Mjollnir's use? good work Steve
teravolt, Mon Nov 21 2011, 08:39PM
Do you plan to build another tesla with a higher f res if you get brancing sparks? The two things that have been mensioned is F res 350-500khz and a ramping power envilope to make sword sparks. what type of caps does ^Mjollnir's use? good work Steve
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Mon Nov 21 2011, 09:42PM
Well here are the HV lab test results as promised...
The weird thing is that the sparks fundamentally don't look that different to how they did without the ramp! Either the ramp generator is really bad (the scope shows the primary current, and as you can see it is quite bad ) or the 220kHz resonant frequency affects the spark appearance so strongly that no amount of envelope shaping can undo it.
They look pretty big and gnarly though, and no components blew up, so I think it's at least somewhat of a result. I intend to go on and improve the modulation somewhat, then tighten the coupling and add more DC bus caps and see if the sparks can be got any bigger that way.
Then I'll build a new resonator with a frequency above 320kHz, which according to Steve Ward is the magic frequency where the sparks suddenly go swordlike. If I don't get sword sparks, this will prove that the modulator isn't good enough.
The tank cap used in this coil is 2 strings of 10 470nF, 1kV PP film/foil caps, for a total of 0.1uF. I've managed to get hold of 10 conduction cooled mica caps rated at 22nF, 1kV RMS, 75A RMS, to make a new "MMC From Hell".
Stay tuned for more updates, if you pardon the pun!
(Moved pics and video into top post as per Projects forum rules.)
Steve Conner, Mon Nov 21 2011, 09:42PM
Well here are the HV lab test results as promised...
The weird thing is that the sparks fundamentally don't look that different to how they did without the ramp! Either the ramp generator is really bad (the scope shows the primary current, and as you can see it is quite bad
) or the 220kHz resonant frequency affects the spark appearance so strongly that no amount of envelope shaping can undo it.They look pretty big and gnarly though, and no components blew up, so I think it's at least somewhat of a result. I intend to go on and improve the modulation somewhat, then tighten the coupling and add more DC bus caps and see if the sparks can be got any bigger that way.
Then I'll build a new resonator with a frequency above 320kHz, which according to Steve Ward is the magic frequency where the sparks suddenly go swordlike. If I don't get sword sparks, this will prove that the modulator isn't good enough.
The tank cap used in this coil is 2 strings of 10 470nF, 1kV PP film/foil caps, for a total of 0.1uF. I've managed to get hold of 10 conduction cooled mica caps rated at 22nF, 1kV RMS, 75A RMS, to make a new "MMC From Hell".
Stay tuned for more updates, if you pardon the pun!
(Moved pics and video into top post as per Projects forum rules.)
Re: Sigma delta modulated QCW- bang without the buck?
Marko, Mon Nov 21 2011, 09:47PM
Lol that makes a good bang... reminds far more of zillipoper coils than a QCW... can see more clearly now - your primary current seems to be rising really fast, more DRSSTC-like, notching and then settling to stable level... hm, control circuit malfunction?
Marko
Marko, Mon Nov 21 2011, 09:47PM
Lol that makes a good bang... reminds far more of zillipoper coils than a QCW... can see more clearly now - your primary current seems to be rising really fast, more DRSSTC-like, notching and then settling to stable level... hm, control circuit malfunction?
Marko
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Mon Nov 21 2011, 09:58PM
Well, I'd class Zilipopper's coil as a QCW too.
The control circuit isn't necessarily "malfunctioning", it's just not doing what I expected! I can sometimes see the primary current drop almost to zero at the end of the burst and then pop up again.
I have to find out whether it's some sort of instability caused by nonlinear spark loading, or whether noise is just getting into it and tripping it prematurely, which would be a malfunction.
Edit: The best explanation I can think of is that the DC bus cap is emptying out, and a half cycle of mains is coming along and recharging it partly.
Steve Conner, Mon Nov 21 2011, 09:58PM
Well, I'd class Zilipopper's coil as a QCW too.
The control circuit isn't necessarily "malfunctioning", it's just not doing what I expected! I can sometimes see the primary current drop almost to zero at the end of the burst and then pop up again.
I have to find out whether it's some sort of instability caused by nonlinear spark loading, or whether noise is just getting into it and tripping it prematurely, which would be a malfunction.
Edit: The best explanation I can think of is that the DC bus cap is emptying out, and a half cycle of mains is coming along and recharging it partly.
Re: Sigma delta modulated QCW- bang without the buck?
Avalanche, Tue Nov 22 2011, 01:24PM
Awesome, yes it certainly is loud! The only time I get that kind of sound from my SSTCs is when they explode
Look forward to seeing what happens when you add more DC bus capacitance, should straighten out the ramp a bit especially as the loop is closed on current.
Avalanche, Tue Nov 22 2011, 01:24PM
Awesome, yes it certainly is loud! The only time I get that kind of sound from my SSTCs is when they explode
Look forward to seeing what happens when you add more DC bus capacitance, should straighten out the ramp a bit especially as the loop is closed on current.
Re: Sigma delta modulated QCW- bang without the buck?
teravolt, Wed Nov 23 2011, 03:01PM
a tube is a natural chioce for a high voltage high frequency modulator. would it be pausible to do delta sigma with a VTTC
teravolt, Wed Nov 23 2011, 03:01PM
a tube is a natural chioce for a high voltage high frequency modulator. would it be pausible to do delta sigma with a VTTC
Re: Sigma delta modulated QCW- bang without the buck?
Marko, Wed Nov 23 2011, 03:36PM
Well, I was thinking of either a delta sigma topology (sensing the primary current with a CT, and controlling the tube either in "staccato" by a mosfet or by toggling the screen grid voltage);
or, in case this still fails, screen could just be modulated linearly with a ramp to produce the same effect! Not as efficient but I see no reason why wouldn't it work!
Marko
Marko, Wed Nov 23 2011, 03:36PM
teravolt wrote ...
a tube is a natural chioce for a high voltage high frequency modulator. would it be pausible to do delta sigma with a VTTC
a tube is a natural chioce for a high voltage high frequency modulator. would it be pausible to do delta sigma with a VTTC
Well, I was thinking of either a delta sigma topology (sensing the primary current with a CT, and controlling the tube either in "staccato" by a mosfet or by toggling the screen grid voltage);
or, in case this still fails, screen could just be modulated linearly with a ramp to produce the same effect! Not as efficient but I see no reason why wouldn't it work!
Marko
Re: Sigma delta modulated QCW- bang without the buck?
Dr. Dark Current, Thu Nov 24 2011, 09:29AM
Yes, the modulation of the screen should work. The parallel resonant primary circuit behaves close to a constant voltage sink (as opposed to constant current sink for a series resonant primary), so the modulation shouldn't be that inefficient.
Dr. Dark Current, Thu Nov 24 2011, 09:29AM
Yes, the modulation of the screen should work. The parallel resonant primary circuit behaves close to a constant voltage sink (as opposed to constant current sink for a series resonant primary), so the modulation shouldn't be that inefficient.
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Thu Nov 24 2011, 10:39AM
It doesn't matter if the modulation is inefficient anyway. The plate of a vacuum tube should be able to handle a huge pulse of heat for a short time- it's not like a silicon die that could crack from thermal shock, or whatever it is that kills them.
I think linear modulation with a tetrode is a good idea, and will be pursuing it as a Plan B
Steve Conner, Thu Nov 24 2011, 10:39AM
It doesn't matter if the modulation is inefficient anyway. The plate of a vacuum tube should be able to handle a huge pulse of heat for a short time- it's not like a silicon die that could crack from thermal shock, or whatever it is that kills them.
I think linear modulation with a tetrode is a good idea, and will be pursuing it as a Plan B
Re: Sigma delta modulated QCW- bang without the buck?
Nicko, Sat Nov 26 2011, 03:29AM
Well, it is made in Scotland...
Nicko, Sat Nov 26 2011, 03:29AM
Steve McConner wrote ...
...
...
"May contain traces of nuts" ??? (look closely)...
...
Well, it is made in Scotland...
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sun Nov 27 2011, 03:32PM
All my Tesla coil boards have some sort of nut reference in them!
Now publishing a schematic of the driver, since people are asking me for it.
Steve Conner, Sun Nov 27 2011, 03:32PM
All my Tesla coil boards have some sort of nut reference in them!
Now publishing a schematic of the driver, since people are asking me for it.
Re: Sigma delta modulated QCW- bang without the buck?
Steve Ward, Sun Dec 04 2011, 10:25AM
Hey guys,
I was thinking that i mislead you into believing that the Q of my QCW tesla coil was really low, this was because of my invalid math and not taking the time to figure it out! I finally did the numbers based on some figures i wrote down about its operation at the peak spark length:
Buck output voltage = 300V, output current = 100A, power = 30kW
Tesla primary current = 150Apk
Tank impedance (20nF at 340khz) = 23.4 ohms, peak voltage = 150*23.4 = 3510v.
Peak stored energy in primary is .5*20n*3510^2 = .1232J
**The mistake i made was thinking Q = 340khz*.1232J / 30kW = 1.4 **
I forgot the factor of 2pi, so its really Q = 2*pi*f*Estore/Pdiss = 2*3.14*340k*.1232/30k = 8.8.
In fact, this is only the energy stored in the primary during operation, i think the secondary energy must also be included (as its energy is stored in phase with the primary), this should raise the Q even higher i suppose. Lets consider the energy stored in the secondary and streamer capacitance of 23pF at a voltage of 66kV, the stored energy is another .05J. Is it coincidental that .05/.1232 = .4, which is the coupling coefficient of the system? I suppose then the total Q of the system is really:
Q = 2*pi*340k*(.1232+.0500)/30k = 12.3.
Gosh was i way off in my interpretation of this before! I was thinking my Q was like 6.28X less :P. Anyway, these are pretty rough numbers, but they are at least in the ballpark.
I need to set up the machine again with a few different secondaries and see what the benefit of the lower secondary impedance was, i unfortunately cant seem to find any good notes on that test, though i recall the primary current was much lower for the same spark length.
Anyway, i think this also means a base-fed resonator (as Marko was thinking of doing with some vacuum tubes) might not be a bad way to go. You probably still end up with a Q of 10 or so for the system (means you need a higher Q secondary, by design), but you no longer need a tank cap, you are offloading all that energy storage into a bigger toroid. I think the double resonant thing still works good for low voltage drives since the overall system Q is pretty much the same, its essentially just using resonance as a tool for impedance matching, the question is, how you gonna store the energy?? will it be in expensive, elusive mica capacitors on the primary, or in a huge field around the toroid? Might even go as far as considering an oil-based capacitor built into the secondary coil (this was Jimmy's idea anyway), or maybe a stack of doorknob caps to achieve something like 50pF. But its starting to seem like no matter what the Q of any tesla coil in this frequency regime is gonna have a total Q around 10.
Steve
Steve Ward, Sun Dec 04 2011, 10:25AM
Hey guys,
I was thinking that i mislead you into believing that the Q of my QCW tesla coil was really low, this was because of my invalid math and not taking the time to figure it out! I finally did the numbers based on some figures i wrote down about its operation at the peak spark length:
Buck output voltage = 300V, output current = 100A, power = 30kW
Tesla primary current = 150Apk
Tank impedance (20nF at 340khz) = 23.4 ohms, peak voltage = 150*23.4 = 3510v.
Peak stored energy in primary is .5*20n*3510^2 = .1232J
**The mistake i made was thinking Q = 340khz*.1232J / 30kW = 1.4 **
I forgot the factor of 2pi, so its really Q = 2*pi*f*Estore/Pdiss = 2*3.14*340k*.1232/30k = 8.8.
In fact, this is only the energy stored in the primary during operation, i think the secondary energy must also be included (as its energy is stored in phase with the primary), this should raise the Q even higher i suppose. Lets consider the energy stored in the secondary and streamer capacitance of 23pF at a voltage of 66kV, the stored energy is another .05J. Is it coincidental that .05/.1232 = .4, which is the coupling coefficient of the system? I suppose then the total Q of the system is really:
Q = 2*pi*340k*(.1232+.0500)/30k = 12.3.
Gosh was i way off in my interpretation of this before! I was thinking my Q was like 6.28X less :P. Anyway, these are pretty rough numbers, but they are at least in the ballpark.
I need to set up the machine again with a few different secondaries and see what the benefit of the lower secondary impedance was, i unfortunately cant seem to find any good notes on that test, though i recall the primary current was much lower for the same spark length.
Anyway, i think this also means a base-fed resonator (as Marko was thinking of doing with some vacuum tubes) might not be a bad way to go. You probably still end up with a Q of 10 or so for the system (means you need a higher Q secondary, by design), but you no longer need a tank cap, you are offloading all that energy storage into a bigger toroid. I think the double resonant thing still works good for low voltage drives since the overall system Q is pretty much the same, its essentially just using resonance as a tool for impedance matching, the question is, how you gonna store the energy?? will it be in expensive, elusive mica capacitors on the primary, or in a huge field around the toroid? Might even go as far as considering an oil-based capacitor built into the secondary coil (this was Jimmy's idea anyway), or maybe a stack of doorknob caps to achieve something like 50pF. But its starting to seem like no matter what the Q of any tesla coil in this frequency regime is gonna have a total Q around 10.
Steve
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sun Dec 04 2011, 11:45AM
Thanks for the info Steve! it's interesting that the loaded Q ended up somewhere about 10 again.
Progress has been slow, but I'll hopefully get some busbars to make my DC bus cap bank today.
Base driving a resonator with tubes seems like a good idea in theory, but it could be a big engineering challenge in practice. Marko recommended a halfbridge with floating drive for the high side tube, but a single tube with a loading capacitor across it would probably work too. I built a tube plasma globe driver like this, and it's also how the tetrode "plasma tweeter" works, except the frequency of that is so high that the tube's own plate capacitance makes an acceptable loading capacitor.
Steve Conner, Sun Dec 04 2011, 11:45AM
Thanks for the info Steve!
it's interesting that the loaded Q ended up somewhere about 10 again.Progress has been slow, but I'll hopefully get some busbars to make my DC bus cap bank today.
Base driving a resonator with tubes seems like a good idea in theory, but it could be a big engineering challenge in practice. Marko recommended a halfbridge with floating drive for the high side tube, but a single tube with a loading capacitor across it would probably work too. I built a tube plasma globe driver like this, and it's also how the tetrode "plasma tweeter" works, except the frequency of that is so high that the tube's own plate capacitance makes an acceptable loading capacitor.
Re: Sigma delta modulated QCW- bang without the buck?
Marko, Sun Dec 04 2011, 09:03PM
Hi guys
Well, I still can't wrap my brain around the nature of spark loading yet, and for base drive secondary Q will be the thing that matters!
Projections of Q in between 1-2 just seemed bizarre to me though. It would almost turn out that the ultimate tesla coil is no tesla coil, and that you could attain the same result by using a big linear tube amplifier operating at 50kV or so and driving a breakout point directly!
Q of 10 should allow some pretty epic sparkage with base feed voltages of 10-20kV.
I would venture into something like that if had some good pulse tubes around, and, more importantly, some good big energy storage... I was thinking that a several kilojoule, 20-30kV can crushing cap would be in order here.
Once I finish all my major projects (which are kind of losing point more and more over time) I might try looking into this dreamware. Unless someone does it before me ofcourse :)
A "plasma tweeter" circuit on steroids could indeed be an easy thing to try for someone who has a GMI 7 and some HV energy storage caps lying around... ramp could be applied to grid. Not saying that a normal VTTC wouldn't work, but this circuit would also be much simpler and save hassle with primary coil and tank caps!
Marko
Marko, Sun Dec 04 2011, 09:03PM
Hi guys
Well, I still can't wrap my brain around the nature of spark loading yet, and for base drive secondary Q will be the thing that matters!
Projections of Q in between 1-2 just seemed bizarre to me though. It would almost turn out that the ultimate tesla coil is no tesla coil, and that you could attain the same result by using a big linear tube amplifier operating at 50kV or so and driving a breakout point directly!
Q of 10 should allow some pretty epic sparkage with base feed voltages of 10-20kV.
I would venture into something like that if had some good pulse tubes around, and, more importantly, some good big energy storage... I was thinking that a several kilojoule, 20-30kV can crushing cap would be in order here.
Once I finish all my major projects (which are kind of losing point more and more over time) I might try looking into this dreamware. Unless someone does it before me ofcourse :)
A "plasma tweeter" circuit on steroids could indeed be an easy thing to try for someone who has a GMI 7 and some HV energy storage caps lying around... ramp could be applied to grid. Not saying that a normal VTTC wouldn't work, but this circuit would also be much simpler and save hassle with primary coil and tank caps!
Marko
Re: Sigma delta modulated QCW- bang without the buck?
Dr. Dark Current, Sun Dec 04 2011, 09:09PM
Marko,
I was planning a "standard" class-C oscillator VTTC with the GMI-7 tube, screen QCW modulation and a defibrillator cap for energy storage, but someone outbid me on the tube
at least I can finish all the other projects...
Dr. Dark Current, Sun Dec 04 2011, 09:09PM
Marko,
I was planning a "standard" class-C oscillator VTTC with the GMI-7 tube, screen QCW modulation and a defibrillator cap for energy storage, but someone outbid me on the tube
at least I can finish all the other projects... Re: Sigma delta modulated QCW- bang without the buck?
Marko, Sun Dec 04 2011, 10:04PM
Hi Jan,
Dr. H. on here seems to own a batch of GMI-7's, I thought to ask him whether he would sell some. A base fed coil in a tweeter-like configuration would be interesting and very easy to attempt if anything.
I do have a big microwave tube at home, external air cooled anode... also it's the type with oxide filament and sensitive grids, which might blow off if pulsed operation was attempted due to their low thermal capacity.
Marko
Marko, Sun Dec 04 2011, 10:04PM
Hi Jan,
Dr. H. on here seems to own a batch of GMI-7's, I thought to ask him whether he would sell some. A base fed coil in a tweeter-like configuration would be interesting and very easy to attempt if anything.
I do have a big microwave tube at home, external air cooled anode... also it's the type with oxide filament and sensitive grids, which might blow off if pulsed operation was attempted due to their low thermal capacity.
Marko
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sat Dec 10 2011, 03:31PM
Update... I fired the coil again with 13,200uF DC bus capacitance instead of 3300. (Operating voltage about 320V as before.) The sparks got thicker and brighter, bright enough to light up the lab, but they didn't get any longer. The spark seems to have reached its final length and structure by about 5ms, and any extra energy just makes it hotter and more meandery.
I didn't blow anything, except the 32A breaker I was running off, when I tried to charge my new monster capbank too quickly.
I discussed the low Q thing with Steve Ward. He says he got his math wrong and the Q was actually about 6!
Next experiment: tighten the coupling and see if the ramp helps to prevent flashovers.
Steve Conner, Sat Dec 10 2011, 03:31PM
Update... I fired the coil again with 13,200uF DC bus capacitance instead of 3300. (Operating voltage about 320V as before.) The sparks got thicker and brighter, bright enough to light up the lab, but they didn't get any longer. The spark seems to have reached its final length and structure by about 5ms, and any extra energy just makes it hotter and more meandery.
I didn't blow anything, except the 32A breaker I was running off, when I tried to charge my new monster capbank too quickly.
I discussed the low Q thing with Steve Ward. He says he got his math wrong and the Q was actually about 6!
Next experiment: tighten the coupling and see if the ramp helps to prevent flashovers.
Re: Sigma delta modulated QCW- bang without the buck?
Marko, Sat Dec 10 2011, 08:08PM
Hi Steve
Any pics n vids of those big, bangy sparks?
Many ideas fall into water with magic frequency required being so high - we can pretty much forget about using bricks, and even the fastest igbt's suffer great losses in this application which makes scaling up the design to much more than what Ward did difficult.
The options we have left are paralleling igbt's or igbt inverters, using tubes, or cryogenic devices (rather experimental for now). Guess what's my favourite!
I think your burst length is getting so great that the coil is running pretty much in the CW mode at the end of the burst.
You said that peak power of your coil as DRSSTC was low, on order of 10-20kW... and since we know how CW sparks tend to bunch up I wouldn't be surprised that your coil won't be able to make greater arc length... I actually believe they might get shorter if you add even more energy and prolong the burst! (If that would make any sense as the igbt's are already pushed pretty hard).
Note that my Big Bad the SSTC only produces about 50cm CW spark with 10kW input.
Time for a 350kHz secondary I guess!
Marko
Marko, Sat Dec 10 2011, 08:08PM
Hi Steve
Any pics n vids of those big, bangy sparks?
Many ideas fall into water with magic frequency required being so high - we can pretty much forget about using bricks, and even the fastest igbt's suffer great losses in this application which makes scaling up the design to much more than what Ward did difficult.
The options we have left are paralleling igbt's or igbt inverters, using tubes, or cryogenic devices (rather experimental for now). Guess what's my favourite!
I think your burst length is getting so great that the coil is running pretty much in the CW mode at the end of the burst.
You said that peak power of your coil as DRSSTC was low, on order of 10-20kW... and since we know how CW sparks tend to bunch up I wouldn't be surprised that your coil won't be able to make greater arc length... I actually believe they might get shorter if you add even more energy and prolong the burst! (If that would make any sense as the igbt's are already pushed pretty hard).
Note that my Big Bad the SSTC only produces about 50cm CW spark with 10kW input.
Time for a 350kHz secondary I guess!
Marko
Re: Sigma delta modulated QCW- bang without the buck?
Steve Ward, Sun Jan 01 2012, 02:59AM
Watt the Fork, did you measure the secondary base current before and after the bus cap modification? Perhaps because of your modulation scheme, the sagging bus cap didnt matter a whole lot and you just stopped skipping as many cycles in order to maintain the target primary current?
Or, maybe the impedance of the system is limiting the peak power delivered to the sparks. What happens if you try to ring it up even higher? Will it do it? or will it just saturate the control and not ring up more primary amps in the long term?
Also, have you checked what the power factor from your H-bridge is when you are at the peak power level? Hopefully you are pretty much driving the whole time, and not recycling, as that would effectively lower the power factor of the inverter (those recycle cycles average in as negative 1 PF), which is lame. That would mean either 1) turn it up higher, or 2) increase the primary impedance, or maybe lower coupling (need to think about that again to be sure).
Steve Ward, Sun Jan 01 2012, 02:59AM
Watt the Fork, did you measure the secondary base current before and after the bus cap modification? Perhaps because of your modulation scheme, the sagging bus cap didnt matter a whole lot and you just stopped skipping as many cycles in order to maintain the target primary current?
Or, maybe the impedance of the system is limiting the peak power delivered to the sparks. What happens if you try to ring it up even higher? Will it do it? or will it just saturate the control and not ring up more primary amps in the long term?
Also, have you checked what the power factor from your H-bridge is when you are at the peak power level? Hopefully you are pretty much driving the whole time, and not recycling, as that would effectively lower the power factor of the inverter (those recycle cycles average in as negative 1 PF), which is lame. That would mean either 1) turn it up higher, or 2) increase the primary impedance, or maybe lower coupling (need to think about that again to be sure).
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Sun Jan 01 2012, 10:23AM
Happy new year all! Good questions, no idea of the answers will check it out when we go back after the Christmas break. I'll get out the Picoscope and the 13 year old sacrificial laptop.
I think the control scheme worked to keep the ramp the same until the DC bus cap "ran out" at which point the RF current suddenly collapsed. With the extra capacitance this wouldn't happen, but I think the streamer had reached its final length before even the small capacitor ran out of charge.
Steve Conner, Sun Jan 01 2012, 10:23AM
Happy new year all! Good questions, no idea of the answers
will check it out when we go back after the Christmas break. I'll get out the Picoscope and the 13 year old sacrificial laptop.I think the control scheme worked to keep the ramp the same until the DC bus cap "ran out" at which point the RF current suddenly collapsed. With the extra capacitance this wouldn't happen, but I think the streamer had reached its final length before even the small capacitor ran out of charge.
Re: Sigma delta modulated QCW- bang without the buck?
Gregory, Fri Jan 06 2012, 11:20PM
And, what about use a linear tube amplifier driven the base of a QCW ressonator?
Gregory, Fri Jan 06 2012, 11:20PM
directly!
Projections of Q in between 1-2 just seemed bizarre to me though. It would almost turn out that the ultimate tesla coil is no tesla coil, and that you could attain the same result by using a big linear tube amplifier operating at 50kV or so and driving a breakout point
And, what about use a linear tube amplifier driven the base of a QCW ressonator?
Re: Sigma delta modulated QCW- bang without the buck?
Steve Ward, Tue Jan 10 2012, 05:07AM
Well, it would be useful to know how much real power the bridge delivers throughout your bang. I recently started messing with my QCW system and found a tuning that actually limited the power of the system quite a bit, where the primary current actually went down with increasing drive voltage! I suppose it might be possible, with current regulation, that your pulse skipper starts jumping more pulses and makes the effective voltage lower, hence the sparks dont grow any longer... maybe? The nature of your control makes it somewhat difficult to know what the real power is, since you have to average all the driving cycles vs recycling cycles to know what the actual power factor is.
Lucky for me and my bus modulation scheme, i can assume my power factor is unity (zero current switching, V and I are in phase) and just look at the volts and amps going to the DRSSTC bus. Of course, im currently investigating other techniques to get rid of the bus modulation, namely class DE switching applied to a phase-shifted bridge. Initially this might look a bit lousy since my existing QCW system needs to modulate the voltage from 50V to 350V. More investigation of the behavior of a double resonant transformer shows there are in fact 3 modes that can give zero current switching: the upper and lower poles*, and a zero frequency somewhere in between**. When operating at the zero, the voltage gain is simply due to the transformer ratio of the system, where as when you operate at either pole, the voltage gain is limited by output loading, so the voltage will just keep going as high as the spark lets it. Anyway, operating at the zero gives a more direct control of the toroid voltage and the whole thing acts like a voltage source. I measured a 5 foot straight spark and found that i need a starting voltage of 56kV and a top voltage of 65kV, so only about 15% modulation depth, which suggests to me that class DE switching is probably not a bad candidate. Of course the control may end up quite a bit more sophisticated.
* in my typical primary feedback schemes, the system will jump to one of these pole frequencies because they satisfy the zero phase requirement for ZCS. Depending on the tuning, one pole frequency gets more energy in it than the other and takes over once this dynamic settles out.
** It takes proper control to get the thing to oscillate at the zero frequency because there is not much gain there compared to the pole frequencies. Its worth noting that a system oscillating at pole frequency will actually converge to the zero frequency as the Q of the system drops sufficiently, and the system effectively looks like a voltage source. I believe my original QCW setup transitions from upper pole operation to zero operation, which is marked by a sharp increase in primary current (vs voltage) near the end of my ramp, that is, the power goes up more than linearly. It also causes my sparks to branch. Tests with a lower impedance resonator avoided this issue, likely because the lower impedance resonator maintains operation at the pole frequency.
Id like to write up a paper about all this QCW DRSSTC theory once i get it all worked out on my own. Turns out there are very many ways to tune a system, including some that are actually intentionally out of tune to exploit specific behaviors.
Steve Ward, Tue Jan 10 2012, 05:07AM
I think the control scheme worked to keep the ramp the same until the DC bus cap "ran out" at which point the RF current suddenly collapsed. With the extra capacitance this wouldn't happen, but I think the streamer had reached its final length before even the small capacitor ran out of charge.
Well, it would be useful to know how much real power the bridge delivers throughout your bang. I recently started messing with my QCW system and found a tuning that actually limited the power of the system quite a bit, where the primary current actually went down with increasing drive voltage! I suppose it might be possible, with current regulation, that your pulse skipper starts jumping more pulses and makes the effective voltage lower, hence the sparks dont grow any longer... maybe? The nature of your control makes it somewhat difficult to know what the real power is, since you have to average all the driving cycles vs recycling cycles to know what the actual power factor is.
Lucky for me and my bus modulation scheme, i can assume my power factor is unity (zero current switching, V and I are in phase) and just look at the volts and amps going to the DRSSTC bus. Of course, im currently investigating other techniques to get rid of the bus modulation, namely class DE switching applied to a phase-shifted bridge. Initially this might look a bit lousy since my existing QCW system needs to modulate the voltage from 50V to 350V. More investigation of the behavior of a double resonant transformer shows there are in fact 3 modes that can give zero current switching: the upper and lower poles*, and a zero frequency somewhere in between**. When operating at the zero, the voltage gain is simply due to the transformer ratio of the system, where as when you operate at either pole, the voltage gain is limited by output loading, so the voltage will just keep going as high as the spark lets it. Anyway, operating at the zero gives a more direct control of the toroid voltage and the whole thing acts like a voltage source. I measured a 5 foot straight spark and found that i need a starting voltage of 56kV and a top voltage of 65kV, so only about 15% modulation depth, which suggests to me that class DE switching is probably not a bad candidate. Of course the control may end up quite a bit more sophisticated.
* in my typical primary feedback schemes, the system will jump to one of these pole frequencies because they satisfy the zero phase requirement for ZCS. Depending on the tuning, one pole frequency gets more energy in it than the other and takes over once this dynamic settles out.
** It takes proper control to get the thing to oscillate at the zero frequency because there is not much gain there compared to the pole frequencies. Its worth noting that a system oscillating at pole frequency will actually converge to the zero frequency as the Q of the system drops sufficiently, and the system effectively looks like a voltage source. I believe my original QCW setup transitions from upper pole operation to zero operation, which is marked by a sharp increase in primary current (vs voltage) near the end of my ramp, that is, the power goes up more than linearly. It also causes my sparks to branch. Tests with a lower impedance resonator avoided this issue, likely because the lower impedance resonator maintains operation at the pole frequency.
Id like to write up a paper about all this QCW DRSSTC theory once i get it all worked out on my own. Turns out there are very many ways to tune a system, including some that are actually intentionally out of tune to exploit specific behaviors.
Re: Sigma delta modulated QCW- bang without the buck?
Goodchild, Tue Jan 10 2012, 03:10PM
Interesting you are trying class DE, I did a lot of simulation on DE a while back don't see any reason it won't work other than you kinda have to use MOSFETs in order for it to actually ZVS. but with some nice large MOSFETs this may not be a problem. A phase shifted bridge also "should" be more efficient than the buck/ZCS setup.
Steve did you ever nail down the spark model for the QCW? If I'm remembering right you said you had to modify it somewhat to get the same primary current profile. I ask because I have a working working DE model, just never had a proper QCW model to test it on.
Goodchild, Tue Jan 10 2012, 03:10PM
Steve Ward wrote ...
Well, it would be useful to know how much real power the bridge delivers throughout your bang. I recently started messing with my QCW system and found a tuning that actually limited the power of the system quite a bit, where the primary current actually went down with increasing drive voltage! I suppose it might be possible, with current regulation, that your pulse skipper starts jumping more pulses and makes the effective voltage lower, hence the sparks dont grow any longer... maybe? The nature of your control makes it somewhat difficult to know what the real power is, since you have to average all the driving cycles vs recycling cycles to know what the actual power factor is.
Lucky for me and my bus modulation scheme, i can assume my power factor is unity (zero current switching, V and I are in phase) and just look at the volts and amps going to the DRSSTC bus. Of course, im currently investigating other techniques to get rid of the bus modulation, namely class DE switching applied to a phase-shifted bridge. Initially this might look a bit lousy since my existing QCW system needs to modulate the voltage from 50V to 350V. More investigation of the behavior of a double resonant transformer shows there are in fact 3 modes that can give zero current switching: the upper and lower poles*, and a zero frequency somewhere in between**. When operating at the zero, the voltage gain is simply due to the transformer ratio of the system, where as when you operate at either pole, the voltage gain is limited by output loading, so the voltage will just keep going as high as the spark lets it. Anyway, operating at the zero gives a more direct control of the toroid voltage and the whole thing acts like a voltage source. I measured a 5 foot straight spark and found that i need a starting voltage of 56kV and a top voltage of 65kV, so only about 15% modulation depth, which suggests to me that class DE switching is probably not a bad candidate. Of course the control may end up quite a bit more sophisticated.
* in my typical primary feedback schemes, the system will jump to one of these pole frequencies because they satisfy the zero phase requirement for ZCS. Depending on the tuning, one pole frequency gets more energy in it than the other and takes over once this dynamic settles out.
** It takes proper control to get the thing to oscillate at the zero frequency because there is not much gain there compared to the pole frequencies. Its worth noting that a system oscillating at pole frequency will actually converge to the zero frequency as the Q of the system drops sufficiently, and the system effectively looks like a voltage source. I believe my original QCW setup transitions from upper pole operation to zero operation, which is marked by a sharp increase in primary current (vs voltage) near the end of my ramp, that is, the power goes up more than linearly. It also causes my sparks to branch. Tests with a lower impedance resonator avoided this issue, likely because the lower impedance resonator maintains operation at the pole frequency.
Id like to write up a paper about all this QCW DRSSTC theory once i get it all worked out on my own. Turns out there are very many ways to tune a system, including some that are actually intentionally out of tune to exploit specific behaviors.
I think the control scheme worked to keep the ramp the same until the DC bus cap "ran out" at which point the RF current suddenly collapsed. With the extra capacitance this wouldn't happen, but I think the streamer had reached its final length before even the small capacitor ran out of charge.
Well, it would be useful to know how much real power the bridge delivers throughout your bang. I recently started messing with my QCW system and found a tuning that actually limited the power of the system quite a bit, where the primary current actually went down with increasing drive voltage! I suppose it might be possible, with current regulation, that your pulse skipper starts jumping more pulses and makes the effective voltage lower, hence the sparks dont grow any longer... maybe? The nature of your control makes it somewhat difficult to know what the real power is, since you have to average all the driving cycles vs recycling cycles to know what the actual power factor is.
Lucky for me and my bus modulation scheme, i can assume my power factor is unity (zero current switching, V and I are in phase) and just look at the volts and amps going to the DRSSTC bus. Of course, im currently investigating other techniques to get rid of the bus modulation, namely class DE switching applied to a phase-shifted bridge. Initially this might look a bit lousy since my existing QCW system needs to modulate the voltage from 50V to 350V. More investigation of the behavior of a double resonant transformer shows there are in fact 3 modes that can give zero current switching: the upper and lower poles*, and a zero frequency somewhere in between**. When operating at the zero, the voltage gain is simply due to the transformer ratio of the system, where as when you operate at either pole, the voltage gain is limited by output loading, so the voltage will just keep going as high as the spark lets it. Anyway, operating at the zero gives a more direct control of the toroid voltage and the whole thing acts like a voltage source. I measured a 5 foot straight spark and found that i need a starting voltage of 56kV and a top voltage of 65kV, so only about 15% modulation depth, which suggests to me that class DE switching is probably not a bad candidate. Of course the control may end up quite a bit more sophisticated.
* in my typical primary feedback schemes, the system will jump to one of these pole frequencies because they satisfy the zero phase requirement for ZCS. Depending on the tuning, one pole frequency gets more energy in it than the other and takes over once this dynamic settles out.
** It takes proper control to get the thing to oscillate at the zero frequency because there is not much gain there compared to the pole frequencies. Its worth noting that a system oscillating at pole frequency will actually converge to the zero frequency as the Q of the system drops sufficiently, and the system effectively looks like a voltage source. I believe my original QCW setup transitions from upper pole operation to zero operation, which is marked by a sharp increase in primary current (vs voltage) near the end of my ramp, that is, the power goes up more than linearly. It also causes my sparks to branch. Tests with a lower impedance resonator avoided this issue, likely because the lower impedance resonator maintains operation at the pole frequency.
Id like to write up a paper about all this QCW DRSSTC theory once i get it all worked out on my own. Turns out there are very many ways to tune a system, including some that are actually intentionally out of tune to exploit specific behaviors.
Interesting you are trying class DE, I did a lot of simulation on DE a while back don't see any reason it won't work other than you kinda have to use MOSFETs in order for it to actually ZVS. but with some nice large MOSFETs this may not be a problem. A phase shifted bridge also "should" be more efficient than the buck/ZCS setup.
Steve did you ever nail down the spark model for the QCW? If I'm remembering right you said you had to modify it somewhat to get the same primary current profile. I ask because I have a working working DE model, just never had a proper QCW model to test it on.
Re: Sigma delta modulated QCW- bang without the buck?
Steve Ward, Wed Jan 11 2012, 05:02PM
For a spark model, i found that using a ~20k resistor in series with a 36kV TVS diode (a diode with a forward and reverse voltage of 36kV for example) works pretty good... it at least gets me in the ballpark. These values were just guesses and dont handle the capacitive aspect of the streamer so i cant say its perfect, but its pretty reasonable i think. Really, the model should be more frequency dependent, so maybe a parallel capacitance with the zener would work ok. I havent tried it yet since the capacitance ought to be varying with streamer length and its somewhat confusing how much capacitance should actually be there.
Steve Ward, Wed Jan 11 2012, 05:02PM
For a spark model, i found that using a ~20k resistor in series with a 36kV TVS diode (a diode with a forward and reverse voltage of 36kV for example) works pretty good... it at least gets me in the ballpark. These values were just guesses and dont handle the capacitive aspect of the streamer so i cant say its perfect, but its pretty reasonable i think. Really, the model should be more frequency dependent, so maybe a parallel capacitance with the zener would work ok. I havent tried it yet since the capacitance ought to be varying with streamer length and its somewhat confusing how much capacitance should actually be there.
Re: Sigma delta modulated QCW- bang without the buck?
teravolt, Wed Jan 11 2012, 07:57PM
I dont know if this a stupid question or not when there is frequency splitting is it so that the upper pole is voltage or inductivly dominant and the lower current or capacitivly dominant? does the the spark behave differently when ever you use the lower pole or the upper pole?
teravolt, Wed Jan 11 2012, 07:57PM
I dont know if this a stupid question or not when there is frequency splitting is it so that the upper pole is voltage or inductivly dominant and the lower current or capacitivly dominant? does the the spark behave differently when ever you use the lower pole or the upper pole?
Re: Sigma delta modulated QCW- bang without the buck?
Steve Conner, Wed Jan 11 2012, 10:56PM
Steve, are you saying you actually found a practical use for the zero in a Tesla coil transfer function? :O
Antonio suggested driving it years ago, but I wasn't interested because it's a kind of anti-resonance so you can't lock a PLL to it. I thought the gain of it was literally zero in the steady state, not the transformer ratio.
I agree with everything else in your post. Under heavy loading the system doesn't have two resonant frequencies any more, just a hump like a bandpass filter. Radio design handbooks show how to make double-tuned RF transformers (radio ham's name for a DRSSTC :) ) and there is a critical coupling coefficient, a function of the loaded Q, that defines the boundary between double humped and single humped responses. I can't remember the formula right now.
Steve Conner, Wed Jan 11 2012, 10:56PM
Steve, are you saying you actually found a practical use for the zero in a Tesla coil transfer function? :O
Antonio suggested driving it years ago, but I wasn't interested because it's a kind of anti-resonance so you can't lock a PLL to it. I thought the gain of it was literally zero in the steady state, not the transformer ratio.
I agree with everything else in your post. Under heavy loading the system doesn't have two resonant frequencies any more, just a hump like a bandpass filter. Radio design handbooks show how to make double-tuned RF transformers (radio ham's name for a DRSSTC :) ) and there is a critical coupling coefficient, a function of the loaded Q, that defines the boundary between double humped and single humped responses. I can't remember the formula right now.
Re: Sigma delta modulated QCW- bang without the buck?
Steve Ward, Thu Jan 12 2012, 10:10PM
When you drive a DRSSTC at its zero frequency, the inverter output voltage is in phase with the secondary output voltage, so the whole thing looks like a *stiff* voltage source and the secondary is parallel resonant. With infinite Q on primary and secondary, the primary current settles out to zero once you get the secondary voltage up to its maximum (hence making this mode hard to achieve with current-feedback based switching). As the Q is lowered, the inverter has to deliver whatever power necessary to maintain the same secondary voltage. Though i have this nagging feeling that im forgetting about something bad that happens when the thing goes out of tune from streamers... i think its possible for the zero to disappear (as in, you cant operate there without hard switching) because not all the reactances cancel out perfectly anymore, thus the load appears capacitive.
This makes me think of another idea i wanted to try, you can think of it as the reverse of the SSTC. Consider how on a SSTC you have minimal energy storage in the primary capacitance (the DC blocker cap), and so we claim its a single-resonant system (damn near most of the energy is stored in the secondary). Well, you can achieve a very similar thing by using a tuned primary and an "untuned" secondary, basically by making the secondary capacitance *too small* for resonance. This gives a similar "voltage source" like behavior since the secondary voltage is more in-phase with the inverter output voltage (well the secondary C makes for some phase error here). Its basically a resonant primary with a "step up transformer" rather than a "tesla coil". Putting most of the energy storage in the primary side lets it have more control over the operating frequency (good for when the sparks hit ground), and also allows for a much tinier secondary since its not storing much energy. Its feasible, i think, to make something like a 2 inch long winding for a secondary (likely under oil) that does 60 inch sparks (in air). I still need to work on this idea some more to figure out if its a dead end or not... lets just say im not completely sold on needing double resonance :P.
Steve Ward, Thu Jan 12 2012, 10:10PM
When you drive a DRSSTC at its zero frequency, the inverter output voltage is in phase with the secondary output voltage, so the whole thing looks like a *stiff* voltage source and the secondary is parallel resonant. With infinite Q on primary and secondary, the primary current settles out to zero once you get the secondary voltage up to its maximum (hence making this mode hard to achieve with current-feedback based switching). As the Q is lowered, the inverter has to deliver whatever power necessary to maintain the same secondary voltage. Though i have this nagging feeling that im forgetting about something bad that happens when the thing goes out of tune from streamers... i think its possible for the zero to disappear (as in, you cant operate there without hard switching) because not all the reactances cancel out perfectly anymore, thus the load appears capacitive.
This makes me think of another idea i wanted to try, you can think of it as the reverse of the SSTC. Consider how on a SSTC you have minimal energy storage in the primary capacitance (the DC blocker cap), and so we claim its a single-resonant system (damn near most of the energy is stored in the secondary). Well, you can achieve a very similar thing by using a tuned primary and an "untuned" secondary, basically by making the secondary capacitance *too small* for resonance. This gives a similar "voltage source" like behavior since the secondary voltage is more in-phase with the inverter output voltage (well the secondary C makes for some phase error here). Its basically a resonant primary with a "step up transformer" rather than a "tesla coil". Putting most of the energy storage in the primary side lets it have more control over the operating frequency (good for when the sparks hit ground), and also allows for a much tinier secondary since its not storing much energy. Its feasible, i think, to make something like a 2 inch long winding for a secondary (likely under oil) that does 60 inch sparks (in air). I still need to work on this idea some more to figure out if its a dead end or not... lets just say im not completely sold on needing double resonance :P.
Print this page