Marx generator powered tesla coil
Benjamin, Wed Nov 23 2016, 04:58AMI was reading through the TCML archives and came across the idea of using a marx generator to charge the tank cap of a tesla coil. However, in this configuration the spark gap is eliminated from the tank circuit and the tank cap is connected directly across the primary coil, eliminating the spark gap losses! The marx generator is connected across the tank cap through a spark gap. This circuit relies on the impedance of the primary coil to charge the tank cap in one quick pulse. The spark gap would have to quench before the tank circuit could start ringing. Would this work? The power supply wouldn't have to be a marx generator, it could really be any capacitor bank or pulsed hv transformer.
edit: here is the TCML post
Re: Marx generator powered tesla coil
hen918, Wed Nov 23 2016, 02:54PM
The problems with parallel LC circuits, as you are describing, are given here:
In short, parallel LC tanks are not suitable to be driven with a square-wave. It is, however, possible to use a Mazzilli driver to drive an LC tank with a sine wave, as is done here:
hen918, Wed Nov 23 2016, 02:54PM
The problems with parallel LC circuits, as you are describing, are given here:
In short, parallel LC tanks are not suitable to be driven with a square-wave. It is, however, possible to use a Mazzilli driver to drive an LC tank with a sine wave, as is done here:
Re: Marx generator powered tesla coil
Benjamin, Wed Nov 23 2016, 06:17PM
Here is a schematic of how I think it would be wired. Apparently It was done in 1964, that is why I posed the question.
]marx_powered_tesla_coil.pdf[/file]
Benjamin, Wed Nov 23 2016, 06:17PM
Here is a schematic of how I think it would be wired. Apparently It was done in 1964, that is why I posed the question.
]marx_powered_tesla_coil.pdf[/file]
Re: Marx generator powered tesla coil
Uspring, Thu Nov 24 2016, 11:42AM
Might well work, but there is an efficiency problem: Only at most 1/4 of the energy stored in the Marx will go into the tank cap. The rest will be dissipated in the spark gaps.
Uspring, Thu Nov 24 2016, 11:42AM
Might well work, but there is an efficiency problem: Only at most 1/4 of the energy stored in the Marx will go into the tank cap. The rest will be dissipated in the spark gaps.
Re: Marx generator powered tesla coil
Dr. Slack, Thu Nov 24 2016, 12:22PM
Suppose in Bejamin's pdf we did away with the tank capacitor. Now when the Marx fires, it's a number of C's in series with a number of gaps and the primary, basically an MMC with a multi-gap gap, with the MMC charged in parallel rather than series.
That eases the high voltage requirement on the power supply. If you really can't get an NST, then you could use a couple of MOTs and a marx configuration instead. One problem is that marx gaps tend to like higher voltages for stable breakdown, so there's a tension between stage voltage and output voltage, which is eased by reducing the number of stages.
For a mains-powered coil, this could be quite nice. Re-imagine the marx configuration slightly so that all stages *are* charged in parallel, rather than a ladder, so that all stages charge equally. Use charging inductors that are low impedance at mains but withstand the full primary operating voltage at the TC frequency, so no need for iron cores, they'd look like mini secondaries. Build the gaps like a multi-gap gap, setup with shims so they are all the same width so should self-fire reasonably consistently, with a common line of sight for UV so they all fire at the same time.
It's an interesting sledgehammer, it would probably crack the 'no NST' nut.
Dr. Slack, Thu Nov 24 2016, 12:22PM
Suppose in Bejamin's pdf we did away with the tank capacitor. Now when the Marx fires, it's a number of C's in series with a number of gaps and the primary, basically an MMC with a multi-gap gap, with the MMC charged in parallel rather than series.
That eases the high voltage requirement on the power supply. If you really can't get an NST, then you could use a couple of MOTs and a marx configuration instead. One problem is that marx gaps tend to like higher voltages for stable breakdown, so there's a tension between stage voltage and output voltage, which is eased by reducing the number of stages.
For a mains-powered coil, this could be quite nice. Re-imagine the marx configuration slightly so that all stages *are* charged in parallel, rather than a ladder, so that all stages charge equally. Use charging inductors that are low impedance at mains but withstand the full primary operating voltage at the TC frequency, so no need for iron cores, they'd look like mini secondaries. Build the gaps like a multi-gap gap, setup with shims so they are all the same width so should self-fire reasonably consistently, with a common line of sight for UV so they all fire at the same time.
It's an interesting sledgehammer, it would probably crack the 'no NST' nut.
Re: Marx generator powered tesla coil
Benjamin, Thu Nov 24 2016, 06:36PM
So something like this just with more stages and inductors instead of resistors. Would the air core charging inductors change the resonant frequency of the primary circuit?
Benjamin, Thu Nov 24 2016, 06:36PM
So something like this
just with more stages and inductors instead of resistors. Would the air core charging inductors change the resonant frequency of the primary circuit?Re: Marx generator powered tesla coil
Uspring, Fri Nov 25 2016, 02:35PM
As Dr. Slack said the charging inductors would be low impedance only at line frequencies. At the TC frequency their effect on the primary resonance is very small.
Uspring, Fri Nov 25 2016, 02:35PM
As Dr. Slack said the charging inductors would be low impedance only at line frequencies. At the TC frequency their effect on the primary resonance is very small.
Re: Marx generator powered tesla coil
Benjamin, Tue Nov 29 2016, 04:29AM
What inductance would be enough? I was thinking 500mH air cored with 24 gauge wire.
Benjamin, Tue Nov 29 2016, 04:29AM
What inductance would be enough? I was thinking 500mH air cored with 24 gauge wire.
Re: Marx generator powered tesla coil
johnf, Tue Nov 29 2016, 05:52AM
Benjamin please go and build a 500mH aircored inductor out of 24g wire and then see how big it is and how it behaves with line frequency
after doing all the above come back and report
johnf, Tue Nov 29 2016, 05:52AM
Benjamin please go and build a 500mH aircored inductor out of 24g wire and then see how big it is and how it behaves with line frequency
after doing all the above come back and report
Re: Marx generator powered tesla coil
Benjamin, Tue Nov 29 2016, 06:06AM
I can calculate all of those parameters. 1.5'' diameter by 5'' long, 230 turns, reactance at 60 hz is 188 milli ohms, so almost a short circuit. However what I'm interested in is how it will behave at high frequency connected across the tank cap.
Benjamin, Tue Nov 29 2016, 06:06AM
I can calculate all of those parameters. 1.5'' diameter by 5'' long, 230 turns, reactance at 60 hz is 188 milli ohms, so almost a short circuit. However what I'm interested in is how it will behave at high frequency connected across the tank cap.
Re: Marx generator powered tesla coil
Dr. Slack, Tue Nov 29 2016, 07:16AM
Hi Benjamin,
I think you have your m confused with your micro, m means milli to most of the planet. Use u for micro, when the unicode 0B3C character "μ" isn't handy.
Johnf was being facetious when he suggested you build a 500mH air core inductor, you could fit a small person inside it (50" x 15" x 2300 turns, inductance goes as volume, other things being equal).
A 5" x 1.5" 500uH (500μH) built as you suggest has an SRF of over 4MHz, so will be usable on any practical SGTC. The effect on primary tuning will be small, and will be tuned out / corrected for during the normal TC tuning process.
Dr. Slack, Tue Nov 29 2016, 07:16AM
Hi Benjamin,
I think you have your m confused with your micro, m means milli to most of the planet. Use u for micro, when the unicode 0B3C character "μ" isn't handy.
Johnf was being facetious when he suggested you build a 500mH air core inductor, you could fit a small person inside it (50" x 15" x 2300 turns, inductance goes as volume, other things being equal).
A 5" x 1.5" 500uH (500μH) built as you suggest has an SRF of over 4MHz, so will be usable on any practical SGTC. The effect on primary tuning will be small, and will be tuned out / corrected for during the normal TC tuning process.
Re: Marx generator powered tesla coil
Benjamin, Tue Nov 29 2016, 04:42PM
Ok sorry for the confusion, I meant micro not milli. How would one go about calculating the smallest usable value of inductance that would work in the marx tank circuit? I'm assuming that a larger inductance is better but what is the practical limit for say a 250kHz 1kW tesla coil and a 4 stage marx.
Benjamin, Tue Nov 29 2016, 04:42PM
Ok sorry for the confusion, I meant micro not milli. How would one go about calculating the smallest usable value of inductance that would work in the marx tank circuit? I'm assuming that a larger inductance is better but what is the practical limit for say a 250kHz 1kW tesla coil and a 4 stage marx.
Re: Marx generator powered tesla coil
Dr. Slack, Wed Nov 30 2016, 06:34PM
I'll give you a methodology, though not necessarily solid numbers.
Think of how much energy is stored in the primary coil itself, compared to how much is stored in other inductors. For instance the loop through the MMC and spark gap will store energy, as will current flowing through these charging chokes. The trick is to have *most* of your energy stored in the primary, where it's doing something useful.
It would probably be infeasible to get the energy stored in those non-primary inductors down to 1% of total. But if you had 50% stored elsewhere than the primary, that would make for a pretty poorly designed primary circuit. So it's going to be somewhere between the two extremes. 10%? 20% 5%. Whatever you can achieve and feel comfortable with.
Dr. Slack, Wed Nov 30 2016, 06:34PM
I'll give you a methodology, though not necessarily solid numbers.
Think of how much energy is stored in the primary coil itself, compared to how much is stored in other inductors. For instance the loop through the MMC and spark gap will store energy, as will current flowing through these charging chokes. The trick is to have *most* of your energy stored in the primary, where it's doing something useful.
It would probably be infeasible to get the energy stored in those non-primary inductors down to 1% of total. But if you had 50% stored elsewhere than the primary, that would make for a pretty poorly designed primary circuit. So it's going to be somewhere between the two extremes. 10%? 20% 5%. Whatever you can achieve and feel comfortable with.
Re: Marx generator powered tesla coil
Benjamin, Wed Nov 30 2016, 08:02PM
So if I was designing for 5% of the total energy stored in the inductors, would I want the combined inductance of all the charging chokes to be 20 times the inductance of the primary coil? How do I calculate the stored energy in the inductors? I know that a larger inductance will store more energy with the same current flowing through it, however a larger inductance also has a greater resistance to change in current.
Benjamin, Wed Nov 30 2016, 08:02PM
So if I was designing for 5% of the total energy stored in the inductors, would I want the combined inductance of all the charging chokes to be 20 times the inductance of the primary coil? How do I calculate the stored energy in the inductors? I know that a larger inductance will store more energy with the same current flowing through it, however a larger inductance also has a greater resistance to change in current.
Re: Marx generator powered tesla coil
jpsmith123, Thu Dec 01 2016, 05:03AM
Here's a paper describing a 500 KV, 500 J Marx generator that uses charging inductors. The coils are about 1 mh with 33 ohms resistance and they don't look like anything special.
You might be able to use some cheap audio crossover coils e.g. like this as charging inductors:
BTW sometimes the Marx capacitors are themselves the primary capacitance of a Tesla coil.
Edit: Imagine that the device described in this patent is your complete primary circuit:
jpsmith123, Thu Dec 01 2016, 05:03AM
Here's a paper describing a 500 KV, 500 J Marx generator that uses charging inductors. The coils are about 1 mh with 33 ohms resistance and they don't look like anything special.
You might be able to use some cheap audio crossover coils e.g. like this as charging inductors:
BTW sometimes the Marx capacitors are themselves the primary capacitance of a Tesla coil.
Edit: Imagine that the device described in this patent is your complete primary circuit:
Re: Marx generator powered tesla coil
Benjamin, Thu Dec 01 2016, 05:47AM
I am planning on using the marx caps as my tank caps. I would rather wind my own charging inductors because I already have a bunch of 24 gauge magnet wire. I just don't want to wind 8 coils before I know roughly the best inductance for them. I've attached a revised schematic of what I'm planning to build.
]marx_powered_tesla_coil.pdf[/file]
edit: This schematic is different from the other one even though it has the same name.
Benjamin, Thu Dec 01 2016, 05:47AM
I am planning on using the marx caps as my tank caps. I would rather wind my own charging inductors because I already have a bunch of 24 gauge magnet wire. I just don't want to wind 8 coils before I know roughly the best inductance for them. I've attached a revised schematic of what I'm planning to build.
]marx_powered_tesla_coil.pdf[/file]
edit: This schematic is different from the other one even though it has the same name.
Re: Marx generator powered tesla coil
Dr. Slack, Thu Dec 01 2016, 09:36AM
The reason I was encouraging you to think in the ratio of stored energy was that the energy in an inductor is 0.5LI2, directly proportional to L, for series connections when they're running at the same current. As you might imagine, even without doing the maths, for a parallel connection the energy goes as 1/L for the same voltage.
I guessed at some nominal figures for a biggish primary for javaTC, 10 turns with 10cm internal and 17cm external radii, and it's around 40uH. So with several inductors in shunt with that, you would not want to go much smaller than 500uH, and a tad bigger might be nice. Depending on the configuration, so all parallel, or inductors cascaded up the ladder, you will have different numbers of charging inductors being 'seen' by the primary. You will obviously have to draw it out in detail. A cascade along your marx (the conventional configuration) will act like a transformer so each of your inductors sees just part of the primary resonant voltage, that is, just the voltage on that part of the MMC.
I would be inclined to connect a Terry Filter between the MOTs and the Marx. MOTs do not have the HV insulation that NSTs do, and the thought of a whole bunch of high Q inductors connected to spark gaps and stuff has me feeling breakovers.
Dr. Slack, Thu Dec 01 2016, 09:36AM
The reason I was encouraging you to think in the ratio of stored energy was that the energy in an inductor is 0.5LI2, directly proportional to L, for series connections when they're running at the same current. As you might imagine, even without doing the maths, for a parallel connection the energy goes as 1/L for the same voltage.
I guessed at some nominal figures for a biggish primary for javaTC, 10 turns with 10cm internal and 17cm external radii, and it's around 40uH. So with several inductors in shunt with that, you would not want to go much smaller than 500uH, and a tad bigger might be nice. Depending on the configuration, so all parallel, or inductors cascaded up the ladder, you will have different numbers of charging inductors being 'seen' by the primary. You will obviously have to draw it out in detail. A cascade along your marx (the conventional configuration) will act like a transformer so each of your inductors sees just part of the primary resonant voltage, that is, just the voltage on that part of the MMC.
I would be inclined to connect a Terry Filter between the MOTs and the Marx. MOTs do not have the HV insulation that NSTs do, and the thought of a whole bunch of high Q inductors connected to spark gaps and stuff has me feeling breakovers.
Re: Marx generator powered tesla coil
Benjamin, Thu Dec 01 2016, 04:16PM
Thanks, that makes sense. I have another question, I was planning on using a single parallel copper pipe spark gap for each stage. Will this quench effectively enough or should I build a multiple air blast gap? The total effective capacitance of the cap bank is only 0.0105uF and the firing voltage should be greater than 16 kV.
Benjamin, Thu Dec 01 2016, 04:16PM
Thanks, that makes sense. I have another question, I was planning on using a single parallel copper pipe spark gap for each stage. Will this quench effectively enough or should I build a multiple air blast gap? The total effective capacitance of the cap bank is only 0.0105uF and the firing voltage should be greater than 16 kV.
Re: Marx generator powered tesla coil
Dr. Slack, Thu Dec 01 2016, 09:04PM
one gap per stage *is* a multi-gap for the whole MMC. Given the low voltage per stage, you might have difficulty setting up a single gap, let alone 2 in series. Don't forget to let them have a good sight-line to each other and keep them as close as electrically practical for UV to help triggering of all gaps at the same time.
Dr. Slack, Thu Dec 01 2016, 09:04PM
one gap per stage *is* a multi-gap for the whole MMC. Given the low voltage per stage, you might have difficulty setting up a single gap, let alone 2 in series. Don't forget to let them have a good sight-line to each other and keep them as close as electrically practical for UV to help triggering of all gaps at the same time.
Re: Marx generator powered tesla coil
Benjamin, Thu Dec 01 2016, 10:00PM
By multiple air blast gap I meant one gap per stage, four gaps total.
Benjamin, Thu Dec 01 2016, 10:00PM
By multiple air blast gap I meant one gap per stage, four gaps total.
Print this page