What would you recommend to meet these projects specs?

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
Large electric field and enough voltage to ignite plasma gas mixture tubes

Large electric field and the high voltage is the definition of a tesla coil so that works. Put a noble gas tube near a coil and boom.

The second part of this is the driving signal. It is the key part of this project. There is an arbitrary waveform specific to this device. How i've made it is simple, arbitrary waveform generator.

Now, I've experimented with amplifying a signal generator sine wave to resonate frequency of the coil and boom. Works just fine. The arbitrary waveform is not a sine wave.. I have heard of a arbitrary waveform not being a sine and therefore causing insane loss of efficiency. I get it. That's fine.

I've tested the same amplifier on the arbitrary waveform and quickly realized there was some insane distortion. I am assuming with my limited knowledge that that amplifier is no good and I should move on to a nice Class A amplifier or something to get the best re-creation of the signal. The amplifier I was using was only 5 watt. Still at resonate frequency it lit up a small incandescent light bulb. The idea is to go a bit further with 20 watt. Driving the primary-coil is basically a 0 ohm inductance load, right? Class A amplifiers get hot as well. Especially at those loads. Luckily, the device only needs to be on for 10 minutes and then can be off for any duration of cool down period.

Is a Class A amplifier the way to go?

Unfortunately we have to talk about RF laws. I get them a lot when asking. It's what people pick up the most when I tell them. It seems like there is a lot of hazyness around operating tesla coils. Like they are interference but are ok. Seems like there is way too many people operating them and control is too hard and the output is categorized as not a information carrying signal so its not that much of a big deal? I'm open to what to think. I'm willing to do anything to make it work, so if I need a permit or anything. I don't even know.

What are your thoughts? Are 20 watt class A amplifiers the way to go? Will they handle 10 minutes at resonate frequencies. Do you know any designs?
 

Papabravo

Joined Feb 24, 2006
16,817
The answers to most of your (rhetorical) questions is no.
Large electric field - imprecise. Do you know how to specify and measure the field you ar looking for?
Noble gasses are inert, they don't oxidize or burn
Amplifiers may or may not amplify arbitrary waveforms. It depends on whether the magnitude of the input allows the amplifier to stay in its linear range.
The notion of distortion for an arbitrary waveform is qualitative, hueristic, and mostly useless.
So what is it about a "nice Class A" amplifier that is so attractive aside from having the worst possible efficiency.
Driving the primary-coil is basically a 0 ohm inductance load, right? WRONG! Driving a primary-coil is a challenging undertaking.
Finally a yes. Class A amplifiers get hot as well. Yes they do!
I don't know how to approach helping you because you have no definition for what you want to do - not in any quantifiable sense.
My advice to you is to keep endlessly experimenting with whatever comes into your head until you either suffer a fatal accident (I'll nominate you for a Darwin Award), or you discover something interesting and useful. I'm not holding my breath.
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
The answers to most of your (rhetorical) questions is no.
Large electric field - imprecise. Do you know how to specify and measure the field you ar looking for?
Noble gasses are inert, they don't oxidize or burn
Amplifiers may or may not amplify arbitrary waveforms. It depends on whether the magnitude of the input allows the amplifier to stay in its linear range.
The notion of distortion for an arbitrary waveform is qualitative, hueristic, and mostly useless.
So what is it about a "nice Class A" amplifier that is so attractive aside from having the worst possible efficiency.
Driving the primary-coil is basically a 0 ohm inductance load, right? WRONG! Driving a primary-coil is a challenging undertaking.
Finally a yes. Class A amplifiers get hot as well. Yes they do!
I don't know how to approach helping you because you have no definition for what you want to do - not in any quantifiable sense.
My advice to you is to keep endlessly experimenting with whatever comes into your head until you either suffer a fatal accident (I'll nominate you for a Darwin Award), or you discover something interesting and useful. I'm not holding my breath.
Sorry, not looking to burn the gases. Just get them to ionize and glow in tubes.

Sorry, were you answering my question when you said "what about a nice class a amplifier is so attractive?" Thanks

Challenging how? Driving the primary, can you tell me what to look out for?

Going back to the amplifier, what do you recommend?

Not in any quantifiable sense? Is driving a primary at resonance with an arbitrary not making sense?
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
Ok must be on something you guys are not. Language bound again. Cant wait for new letters.... Anyway I meant resonance as in frequency must be 300khz to 500khz. Resonance in that sentence structure was me saying it must be at a frequency NUMBER that is resonance. I must drive at that frequency. Does that make sense?

Saying it again for maximum clarity

Resonance is a number say 330khz. I must drive this primary winding at that frequency of my arbitrary waveform (whatever it may be) at 20ish watts. And I need an amplifier (class a or otherwise, I'm open for suggestions here guys) that can handle this task completely

Does that make any sense

Next step here boys is going over syllables
 

atferrari

Joined Jan 6, 2004
4,425
What means "arbitrary waveforms" in this context? Have you generated some? Could you give concrete data?

What a lot of words you use, man!!
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
What means "arbitrary waveforms" in this context? Have you generated some? Could you give concrete data?

What a lot of words you use, man!!
It looks like this and contains 2047 "points" right now from an arbitrary waveform generator. 12 bit res. Actually because this arbitrary waveform gen only has like 1 million Samples per second (its a china brand) for the final project I'll get a SDG2042X and do 30 million points a second. Giving me resolution up to higher frequencies.
Capture+_2020-10-05-00-33-45.png





Hey man, words get you around, ok?
 

andrewmm

Joined Feb 25, 2011
1,757
breaking this down to bits that we can quantify.

You have a signal, is it a raised cosine pulse or something like ?
is it a single pulse, or repetative, what is the repetitio rate ?

Now down to the amplifier,
we need to know what bandwidth you need.
do you need down to DC or is AC coupled OK,

What is the upper frequency of the signal you want to send ?
any amplifier is going to limit the frequency range,
which will show as distortion as far as you see,
the input and the output are different shapes.

An amplifier needs to have a current and a voltage gain, as well as a frequency response.

If you need DC to 0.5 MHz, and you want FLAT thats a very expensive RF power amplifier.

You say about your load as being zero ohms,
any power into zero ohms, is an infinite current.
so hope that not true.

Engineering is about knowing what can be compromised on and what can't
 

Janis59

Joined Aug 21, 2017
1,348
If any gas pressure into "lamp" between 0,1 and 10 Torr, then for frequencies between 13 MHz and 130 MHz You need a 1...5 kV and 1...10 Amp. The simplest and at once a most trustee method to generate that is serial resonance, where lamp is staying electrodeless in midst of the resonant tank coil. The simplest generator is IXFH42N60 or faster APT6038. The citcuit consists of Drain to gnd caps matrix where best low-loss SMD caps matrix must be used. Then small coil in the Source to gnd wire with ca 20-30 turns 1...2 mm wire on 1 inch teflon core. Toward Gate some bias must be formed from 5V stabilizer via resistive voltage divider (1 Kohm "frog" trimmer) and serial inductance to stop the RF via the frog. Roughly 100-200 turns 0,1 mm wire over resistor. The tank is switched between Gate and Vcc, serial coil of ca 3-5 turns of pencil thick wire of 10-40 mm diameter as lamp demands. Capacitor must withstand abot 1 MVAR of reactive pover and kilovolts thus the best choice is Rogers pcb materials like one of Duroid (TM). Both metallized patches makes a capacitor over their teflon substrate. Few cm2 is needed only (20-200 pf according the frequency of choice). Thickess - best 3 mm but no less than 2 mm. The positive back-loop must not be organized, it happens automatically via the parasythic capacitances in MOSFET crystall. Vcc must be about 24 V but never over 36 V. With "frog"-trimmer adjust the beginning of oscillation and work current.
 

MisterBill2

Joined Jan 23, 2018
9,292
Driving a resonant circuit is how class "C" amplifiers are able to develop lots of power in a reasonably clean waveform. The "Q" factor of the resonant circuit does need to be high enough to create the whole sine wave, and so there is some effort required in setting the values. And of course the frequency used must be selected to avoid interfering with anybody who cares. That may not be easy. So a class "A" amplifier is not the best choice, use a class "C" circuit and drive the active device into saturation for a small portion of the cycle.
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
breaking this down to bits that we can quantify.

You have a signal, is it a raised cosine pulse or something like ?
is it a single pulse, or repetative, what is the repetitio rate ?

Now down to the amplifier,
we need to know what bandwidth you need.
do you need down to DC or is AC coupled OK,

What is the upper frequency of the signal you want to send ?
any amplifier is going to limit the frequency range,
which will show as distortion as far as you see,
the input and the output are different shapes.

An amplifier needs to have a current and a voltage gain, as well as a frequency response.

If you need DC to 0.5 MHz, and you want FLAT thats a very expensive RF power amplifier.

You say about your load as being zero ohms,
any power into zero ohms, is an infinite current.
so hope that not true.

Engineering is about knowing what can be compromised on and what can't
It holds specific inner frequencies, but does resemble raised cosine pulse.

Well its a tesla coil so its not going to operate at low frequencies or DC. Typical ranges for tesla coils are 50khz to 1Mhz. The coils ive been making range from 330khz and 750khz. Again since ive been driving them with sine waves to find resonance, that may change once I start with arbitrary. I aint going to expect the world from an amplifier. But that range is a start

Again I'm not very good with this stuff, but my 0 ohm is only in DC in an inductive load right? At higher frequencies its going to look like more resistance. So no, I was incorrect.

Which I guess makes me ask what the resistance is going to be? Will i know? I thought at the sweet point of resonance I'll get the least resistance?
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
If any gas pressure into "lamp" between 0,1 and 10 Torr, then for frequencies between 13 MHz and 130 MHz You need a 1...5 kV and 1...10 Amp. The simplest and at once a most trustee method to generate that is serial resonance, where lamp is staying electrodeless in midst of the resonant tank coil. The simplest generator is IXFH42N60 or faster APT6038. The citcuit consists of Drain to gnd caps matrix where best low-loss SMD caps matrix must be used. Then small coil in the Source to gnd wire with ca 20-30 turns 1...2 mm wire on 1 inch teflon core. Toward Gate some bias must be formed from 5V stabilizer via resistive voltage divider (1 Kohm "frog" trimmer) and serial inductance to stop the RF via the frog. Roughly 100-200 turns 0,1 mm wire over resistor. The tank is switched between Gate and Vcc, serial coil of ca 3-5 turns of pencil thick wire of 10-40 mm diameter as lamp demands. Capacitor must withstand abot 1 MVAR of reactive pover and kilovolts thus the best choice is Rogers pcb materials like one of Duroid (TM). Both metallized patches makes a capacitor over their teflon substrate. Few cm2 is needed only (20-200 pf according the frequency of choice). Thickess - best 3 mm but no less than 2 mm. The positive back-loop must not be organized, it happens automatically via the parasythic capacitances in MOSFET crystall. Vcc must be about 24 V but never over 36 V. With "frog"-trimmer adjust the beginning of oscillation and work current.
Interesting. I'm not sure but I feel I seen this in a youtube video. Are you sure this will allow me to pick the signal or is it optimized to find and self resonate? Like that slayer exciter
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
Driving a resonant circuit is how class "C" amplifiers are able to develop lots of power in a reasonably clean waveform. The "Q" factor of the resonant circuit does need to be high enough to create the whole sine wave, and so there is some effort required in setting the values. And of course the frequency used must be selected to avoid interfering with anybody who cares. That may not be easy. So a class "A" amplifier is not the best choice, use a class "C" circuit and drive the active device into saturation for a small portion of the cycle.
Well I feel a bit dumb on the whole class A thing.

Thanks a lot.

I'm new to this stuff, altho I really want to jump into a four year college in electronics. I'm a bit too old now maybe.. I'm 27. Idk.

Anyway, am I reading this right. A class c then has a very high frequency tuned oscillator that will be used in conjunction with the input and if everything lines up alright it will output my input signal relatively distortionless.

I'm really just looking for clearness on how to design because I want to ask someone who can to build it for me. I'm confident the fastest way is to just pay someone. Altho I cant be asking the person I'm paying to build a class A. I have to have an idea. You get the point. Thanks so much in helping me
 

MisterBill2

Joined Jan 23, 2018
9,292
The class "C" amplifier would be driven at the resonant frequency of the tuned circuits, it does nothing until there is a frequency drive applied. The active device in the amplifier switches on for quite a bit less than half a cycle and the resonant circuit is driven to complete that cycle. It works quite well. But the drive does not look like the picture of that pulse. There is a lot of good information on class "C" amplifiers and so I will not explain it. But they can work very well in a resonant Tesla coil system. Mosfets are more forgiving of high voltage spikes. Bipolar transistors work also.
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
The class "C" amplifier would be driven at the resonant frequency of the tuned circuits, it does nothing until there is a frequency drive applied. The active device in the amplifier switches on for quite a bit less than half a cycle and the resonant circuit is driven to complete that cycle. It works quite well. But the drive does not look like the picture of that pulse. There is a lot of good information on class "C" amplifiers and so I will not explain it. But they can work very well in a resonant Tesla coil system. Mosfets are more forgiving of high voltage spikes. Bipolar transistors work also.
I see. I see. I looked them up. They are super cool. All systems are really cool. Anyhow, I am getting the picture this wouldn't really recreate my waveform in that picture. Or something like it would be possible if the timing of the circuit was controlled in some deep mathematical way. The point was to exactly recreate that waveform.
 

MisterBill2

Joined Jan 23, 2018
9,292
OK, to create that waveform will require something different than a class "C" amplifier, because those depend on having a resonant circuit to produce the output, and from a resonant circuit one gets a sine wave, possibly with a bit of harmonic distortion, but really a sine wave. Worse yet, given that a Tesla coil system is a resonant tuned circuit it produces a sine wave, although there may be some distortion. Producing that waveform as shown in post #7 will be much more challenging.
What is it that you gain by using that specific shape?
 

Thread Starter

CosmicOrderMachines

Joined Oct 31, 2019
77
OK, to create that waveform will require something different than a class "C" amplifier, because those depend on having a resonant circuit to produce the output, and from a resonant circuit one gets a sine wave, possibly with a bit of harmonic distortion, but really a sine wave. Worse yet, given that a Tesla coil system is a resonant tuned circuit it produces a sine wave, although there may be some distortion. Producing that waveform as shown in post #7 will be much more challenging.
What is it that you gain by using that specific shape?
Yes the waveform is more important than the project. The concept is an electrical field disturbance that "paints" a picture. A picture that controls various substance placed in the field.
 
Top