High Voltage Crystal Oscillator
Enceladus, Thu Feb 09 2017, 07:17AMEver since I leaned that Q values in excess of a million are possible with crystal resonators I have wondered if there might be a way to design a crystal power oscillator that would operate something like a Tesla coil with the conventional secondary circuit replaced or supplemented by a crystal with potentially several orders of magnitude greater Q factor. I've tried simulating a few schemes where I basically placed the crystal below and in series with the topload of a conventional coil with very underwhelming results. I know essentially how to simulate the behavior of a crystal but I'm still not grasping something here.
Can any of you think of any way this sort of thing could work? Is there even any merit to the idea?
-J
Re: High Voltage Crystal Oscillator
Erlend^SE, Thu Feb 09 2017, 08:23PM
Well, check piezo-transformers.
It's the real deal. Piezo crystal with multiple connections and resonant driver.
I may have a LCD backlight driver around somewhere based on them.
Erlend^SE, Thu Feb 09 2017, 08:23PM
Well, check piezo-transformers.
It's the real deal. Piezo crystal with multiple connections and resonant driver.
I may have a LCD backlight driver around somewhere based on them.
Re: High Voltage Crystal Oscillator
Antonio, Sat Feb 11 2017, 01:14PM
Tesla coils don't depend on the quality factor of the resonator. Too low quality factor creates losses, of course, but after a quite low value there is no more difference.
Antonio, Sat Feb 11 2017, 01:14PM
Tesla coils don't depend on the quality factor of the resonator. Too low quality factor creates losses, of course, but after a quite low value there is no more difference.
Re: High Voltage Crystal Oscillator
Dr. Slack, Sun Feb 12 2017, 12:06PM
@Antonio, as in, once they break out, the streamer loading is so heavy, that a high unloaded Q is irrelevant
Dr. Slack, Sun Feb 12 2017, 12:06PM
@Antonio, as in, once they break out, the streamer loading is so heavy, that a high unloaded Q is irrelevant
Re: High Voltage Crystal Oscillator
Enceladus, Mon Feb 13 2017, 05:52PM
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
Maybe you and I just have different ideas on what constitutes a low Q value. The only ways it could be <<100 would be:
A) If your secondary wire is too small, or not copper
B) If your secondary turns are widely spaced or too few in number
C) If your secondary coil has insufficient height or diameter
D) If you deliberately add series resistance or shunt conductance to broaden the bandwidth or otherwise limit performance
TC secondaries actually have similar inductance and ESR to quartz crystals in the ~1MHz operating range, but crystals generally have series capacitance measured in fF.
The only type of crystal I can think of that could stand up to such extreme electrical and mechanical stress would be the igniter type, but I have no clue what kind of Q factors they'd have. Judging by their size and shape, they probably resonate at too high a frequency to be useful anyways. I would probably have to cut or grow some special.
Perhaps a cavity resonator would be better suited to stand in as a tertiary magnifier circuit. I'm sure it's been tried, but as for the crystals, there is very little information.
Any links to info on piezo-transformers or anything else along these lines would be appreciated.
Enceladus, Mon Feb 13 2017, 05:52PM
Antonio wrote ...
Tesla coils don't depend on the quality factor of the resonator. Too low quality factor creates losses, of course, but after a quite low value there is no more difference.
Tesla coils don't depend on the quality factor of the resonator. Too low quality factor creates losses, of course, but after a quite low value there is no more difference.
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
Maybe you and I just have different ideas on what constitutes a low Q value. The only ways it could be <<100 would be:
A) If your secondary wire is too small, or not copper
B) If your secondary turns are widely spaced or too few in number
C) If your secondary coil has insufficient height or diameter
D) If you deliberately add series resistance or shunt conductance to broaden the bandwidth or otherwise limit performance
TC secondaries actually have similar inductance and ESR to quartz crystals in the ~1MHz operating range, but crystals generally have series capacitance measured in fF.
The only type of crystal I can think of that could stand up to such extreme electrical and mechanical stress would be the igniter type, but I have no clue what kind of Q factors they'd have. Judging by their size and shape, they probably resonate at too high a frequency to be useful anyways. I would probably have to cut or grow some special.
Perhaps a cavity resonator would be better suited to stand in as a tertiary magnifier circuit. I'm sure it's been tried, but as for the crystals, there is very little information.
Any links to info on piezo-transformers or anything else along these lines would be appreciated.
Re: High Voltage Crystal Oscillator
Sulaiman, Mon Feb 13 2017, 08:06PM
from memory of my coils;
. before breakout "Q" = 100 to 500
. during breakout ... really not sure but guestimated at 3 to 10
Here I use the definition of Q = (reactive power) / (real power)
sometimes I use Q=(resonant frequency)/(3dB bandwidth)
so if you have a piezo/resonant transformer
with a loaded/working Q of 10,000 giving a hv output of 1kW
then the reactive power would be 10 MVAR
or if you think of it as Q=(reactive resonant current)/(resistive load current)
a 1A output current (arc) would require a resonating current of 10kA for a loaded Q of 10,000
the physical constraints are quite stringent.
Sulaiman, Mon Feb 13 2017, 08:06PM
from memory of my coils;
. before breakout "Q" = 100 to 500
. during breakout ... really not sure but guestimated at 3 to 10
Here I use the definition of Q = (reactive power) / (real power)
sometimes I use Q=(resonant frequency)/(3dB bandwidth)
so if you have a piezo/resonant transformer
with a loaded/working Q of 10,000 giving a hv output of 1kW
then the reactive power would be 10 MVAR
or if you think of it as Q=(reactive resonant current)/(resistive load current)
a 1A output current (arc) would require a resonating current of 10kA for a loaded Q of 10,000
the physical constraints are quite stringent.
Re: High Voltage Crystal Oscillator
Antonio, Tue Feb 14 2017, 01:03AM
No. The maximum voltage gain in a conventional Tesla coil, modeled as two coupled LC circuits, is given, by energy conservation, as que square root of the ratio between primary and secondary capacitances, or, equivalently, by the ratio between secondary and primary inductances. High Q in the coils can't increase this gain. Low Q can reduce it, of course.
Antonio, Tue Feb 14 2017, 01:03AM
Enceladus wrote ...
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
No. The maximum voltage gain in a conventional Tesla coil, modeled as two coupled LC circuits, is given, by energy conservation, as que square root of the ratio between primary and secondary capacitances, or, equivalently, by the ratio between secondary and primary inductances. High Q in the coils can't increase this gain. Low Q can reduce it, of course.
Re: High Voltage Crystal Oscillator
Enceladus, Tue Feb 14 2017, 09:23AM
I think I get what your saying. I certainly have a lot still to learn, but a lot of what I said above is actually consistent with what you're saying here.
So in a real world coil, what is a typical critical value, where the correlation stops? I could use some examples. What about slayer exiters? They often don't even include a primary cap, and basically just pump the secondary circuit continuously. Also, what about 3-coil TC's? Doesn't the tertiary circuit effectiveness depend on it having high Q?
Enceladus, Tue Feb 14 2017, 09:23AM
Antonio wrote ...
No. The maximum voltage gain in a conventional Tesla coil, modeled as two coupled LC circuits, is given, by energy conservation, as que square root of the ratio between primary and secondary capacitances, or, equivalently, by the ratio between secondary and primary inductances. High Q in the coils can't increase this gain. Low Q can reduce it, of course.
Enceladus wrote ...
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
What are you talking about dude? That's like saying a Tesla coil resonator doesn't depend on resonance. A TC secondary circuit is always designed to have the highest inductance, lowest series resistance, and thus the highest Q factor possible. Voltage gain from input to output, neglecting losses will depend on the root of the ratio of Ls to Lp and the voltage gain within the secondary circuit when driven at resonance is equal to its Q factor. Thus, the secondary Q should be as high as practical. Q factor IS a factor of the final gain product. That is, ignoring dynamic loading effects, a coil with a secondary circuit Q of 200 will have twice the voltage gain of a coil with a Q of 100.
No. The maximum voltage gain in a conventional Tesla coil, modeled as two coupled LC circuits, is given, by energy conservation, as que square root of the ratio between primary and secondary capacitances, or, equivalently, by the ratio between secondary and primary inductances. High Q in the coils can't increase this gain. Low Q can reduce it, of course.
I think I get what your saying. I certainly have a lot still to learn, but a lot of what I said above is actually consistent with what you're saying here.
So in a real world coil, what is a typical critical value, where the correlation stops? I could use some examples. What about slayer exiters? They often don't even include a primary cap, and basically just pump the secondary circuit continuously. Also, what about 3-coil TC's? Doesn't the tertiary circuit effectiveness depend on it having high Q?
Re: High Voltage Crystal Oscillator
jpsmith123, Tue Feb 14 2017, 07:30PM
Maybe you are asking something like: Being that there is a Q dependent voltage gain possible with a series LC circuit, then is it somehow possible to get a very high voltage gain from a crystal (such as used in a crystal oscillator), which has a very high Q?
And I think an answer is no, because, when you look at a crystal equivalent circuit, you cannot get access to a set of terminals where there would be a voltage gain as in a discrete LC circuit.
jpsmith123, Tue Feb 14 2017, 07:30PM
Maybe you are asking something like: Being that there is a Q dependent voltage gain possible with a series LC circuit, then is it somehow possible to get a very high voltage gain from a crystal (such as used in a crystal oscillator), which has a very high Q?
And I think an answer is no, because, when you look at a crystal equivalent circuit, you cannot get access to a set of terminals where there would be a voltage gain as in a discrete LC circuit.
Enceladus wrote ...
Ever since I leaned that Q values in excess of a million are possible with crystal resonators I have wondered if there might be a way to design a crystal power oscillator that would operate something like a Tesla coil with the conventional secondary circuit replaced or supplemented by a crystal with potentially several orders of magnitude greater Q factor. I've tried simulating a few schemes where I basically placed the crystal below and in series with the topload of a conventional coil with very underwhelming results. I know essentially how to simulate the behavior of a crystal but I'm still not grasping something here.
Can any of you think of any way this sort of thing could work? Is there even any merit to the idea?
-J
Ever since I leaned that Q values in excess of a million are possible with crystal resonators I have wondered if there might be a way to design a crystal power oscillator that would operate something like a Tesla coil with the conventional secondary circuit replaced or supplemented by a crystal with potentially several orders of magnitude greater Q factor. I've tried simulating a few schemes where I basically placed the crystal below and in series with the topload of a conventional coil with very underwhelming results. I know essentially how to simulate the behavior of a crystal but I'm still not grasping something here.
Can any of you think of any way this sort of thing could work? Is there even any merit to the idea?
-J
Re: High Voltage Crystal Oscillator
woodchuck, Tue Feb 14 2017, 10:27PM
Yup. It's a case of taking lumped parameter models too far. All models (as all analogies) fall apart at some point. But if anyone wanted to do some crazy experimentation, a large bar or disk of PZT might be a place to start.
woodchuck, Tue Feb 14 2017, 10:27PM
Yup. It's a case of taking lumped parameter models too far. All models (as all analogies) fall apart at some point. But if anyone wanted to do some crazy experimentation, a large bar or disk of PZT might be a place to start.
Re: High Voltage Crystal Oscillator
Antonio, Wed Feb 15 2017, 01:01AM
Antonio, Wed Feb 15 2017, 01:01AM
Enceladus wrote ...
I think I get what your saying. I certainly have a lot still to learn, but a lot of what I said above is actually consistent with what you're saying here.
So in a real world coil, what is a typical critical value, where the correlation stops? I could use some examples. What about slayer exiters? They often don't even include a primary cap, and basically just pump the secondary circuit continuously. Also, what about 3-coil TC's? Doesn't the tertiary circuit effectiveness depend on it having high Q?
In the picture is a simulation of the output of a Tesla coil, that ideally should generate 100 kV, oscillating at ~500 kHz. Four curves are shown, for secondary Q= infinity, 1000, 100, and 10. It can be seen that only Q=10 results in serious loss. In circuits having continuous excitation, and not the single capacitor discharge of a Tesla coil, Q really plays a role in determining the maximum voltage before breakout.I think I get what your saying. I certainly have a lot still to learn, but a lot of what I said above is actually consistent with what you're saying here.
So in a real world coil, what is a typical critical value, where the correlation stops? I could use some examples. What about slayer exiters? They often don't even include a primary cap, and basically just pump the secondary circuit continuously. Also, what about 3-coil TC's? Doesn't the tertiary circuit effectiveness depend on it having high Q?
Re: High Voltage Crystal Oscillator
Conundrum, Wed Feb 15 2017, 05:39AM
Yup, piezo transformers do work.
PLEASE NOTE: this may or may not be patented, I have no idea but some discussions on this subject were recently removed from this and other forums, suspect that they infringed on something. If so then it would be useful to know for future reference.
Conundrum, Wed Feb 15 2017, 05:39AM
Yup, piezo transformers do work.
PLEASE NOTE: this may or may not be patented, I have no idea but some discussions on this subject were recently removed from this and other forums, suspect that they infringed on something. If so then it would be useful to know for future reference.
Re: High Voltage Crystal Oscillator
Enceladus, Wed Feb 15 2017, 07:54PM
That's exactly what I see when trying to simulate the concept; but consider a crystal used in a gas igniter. When it is subjected to a rapid change in mechanical stress it develops a high voltage pulse at its terminals. Now if that stress were instead created by an alternating electric field at the crystals resonant frequency rather than a mechanical shock, how does the crystal know the difference? How would you model an igniter crystal differently?
Enceladus, Wed Feb 15 2017, 07:54PM
jpsmith123 wrote ...
And I think an answer is no, because, when you look at a crystal equivalent circuit, you cannot get access to a set of terminals where there would be a voltage gain as in a discrete LC circuit.
And I think an answer is no, because, when you look at a crystal equivalent circuit, you cannot get access to a set of terminals where there would be a voltage gain as in a discrete LC circuit.
That's exactly what I see when trying to simulate the concept; but consider a crystal used in a gas igniter. When it is subjected to a rapid change in mechanical stress it develops a high voltage pulse at its terminals. Now if that stress were instead created by an alternating electric field at the crystals resonant frequency rather than a mechanical shock, how does the crystal know the difference? How would you model an igniter crystal differently?
Re: High Voltage Crystal Oscillator
jpsmith123, Wed Feb 15 2017, 09:53PM
It seems you're close to re-discovering the concept of a piezoelectric transformer
.
jpsmith123, Wed Feb 15 2017, 09:53PM
It seems you're close to re-discovering the concept of a piezoelectric transformer
. Re: High Voltage Crystal Oscillator
Enceladus, Fri Feb 17 2017, 08:06AM
Haha, yes, Ereland mentioned them in passing above and I asked for links but never saw a reply. I only posted this thread because all of my attempts at researching the concept seemed to veer toward work done by John Hutchison and other similarly pseudoscientific researchers. I attribute it to certain peoples' perception that crystals are somehow "magical". I love TI technical docs though. This is exactly what I was looking for. Thanks for this link.
Enceladus, Fri Feb 17 2017, 08:06AM
jpsmith123 wrote ...
It seems you're close to re-discovering the concept of a piezoelectric transformer.
It seems you're close to re-discovering the concept of a piezoelectric transformer
. Haha, yes, Ereland mentioned them in passing above and I asked for links but never saw a reply. I only posted this thread because all of my attempts at researching the concept seemed to veer toward work done by John Hutchison and other similarly pseudoscientific researchers. I attribute it to certain peoples' perception that crystals are somehow "magical". I love TI technical docs though. This is exactly what I was looking for. Thanks for this link.
Re: High Voltage Crystal Oscillator
Conundrum, Fri Feb 17 2017, 11:20AM
Thermodynamics still applies, PZTs (google "Rosen transformer") still have losses but they are not as severe as with a conventional wound transformer. I did read somewhere that graphene "windings" have been suggested and some products already use them.
I also found that Transoner (tm) does work but the yields are much lower than they first thought and the failure mode is total so bad units can't even be used for low power applications.
Its worth mentioning that the problems might be fixable by growing them using MOCVD or a related mechanism similar to how some of the newer EL panels are made.
Conundrum, Fri Feb 17 2017, 11:20AM
Thermodynamics still applies, PZTs (google "Rosen transformer") still have losses but they are not as severe as with a conventional wound transformer. I did read somewhere that graphene "windings" have been suggested and some products already use them.
I also found that Transoner (tm) does work but the yields are much lower than they first thought and the failure mode is total so bad units can't even be used for low power applications.
Its worth mentioning that the problems might be fixable by growing them using MOCVD or a related mechanism similar to how some of the newer EL panels are made.
Print this page