Choosing a Tank Capacitor for my Tesla coil

[radicalsniper]

Hello! I am currently having a hard time choosing what will be my Tank capacitor. I am currently using a ZVS Driver with a Flyback transformer type 4B.

My questions are:

1) Can I use a bucket capacitor for the tank capacitor? I know that it isn't efficient when compared to MMCs but I would gladly choose lower cost low efficiency side vs higher cost high efficiency side as long as it is safe.

2) Is it possible to measure the voltage rating of a bucket capacitor? If possible, how?

Hoping for answers to my questions. Thanks! (I will update the post with the picture of the driver when I will have a good connection)

 

[DerStrom8]

Hi radicalsniper, welcome to ETO!

I have seen Tesla Coils built using ZVS drivers and bucket capacitors, though as you say it's not very efficient and there is a chance of it causing the Tesla coil to operate very poorly. An alternate option which I would prefer over a bucket capacitor is an array of beer bottle capacitors. Their construction is fairly consistent (glass thickness, etc) and they are able to withstand around 40kV (though I would not suggest going above 15kV for Tesla coils). Bucket capacitors may work fine, though the size of the bucket, type of plastic it's made out of, impurities in the plastic, etc can greatly affect the operation.

I used beer bottle caps in one of my Tesla coils and found that, if built properly, one bottle can give you between 0.5 and 0.7 nanofarads of capacitance, with a ~40kV rating.

You can calculate dielectric breakdown voltage using the following formula:

where "Vbd" is the breakdown voltage, "Eds" is the dielectric strength of the material (you can find charts online, like at this link), and "d" is the distance between the plates (which, assuming the plates are directly against the dielectric, is simply the thickness of the dielectric). Note that the voltage is in kV, Eds is in kV/mm, and d is in mm. So, for example, let's say you make a capacitor with a sheet of 2mm thick polypropylene:

According to the chart, polypropylene has a dielectric strength of 23.6 kV/mm.
Thickness of the polypropylene is 2mm.

Therefore, the breakdown voltage of the dielectric is:

It is important to know, however, that you should NEVER drive a capacitor at, or even remotely near, its breakdown voltage. Most capacitors are derated significantly to prevent the user from applying voltage anywhere near its breakdown voltage. For example, you may have a capacitor rated for 10V but it may not break down until you apply three or four times that. When creating your own capacitors, you don't have the luxury of having a preset voltage rating that takes this deration into account. You only have the breakdown voltage. Personally, I calculate the required breakdown voltage as follows:

Take the voltage you plan to apply to the capacitor (let's say, 10kV). Multiply that by 2 to get the normal operating voltage rating of the capacitor (20kV). Then, multiply that by 2 again to get the breakdown voltage you need to shoot for (40kV). In order to drive a Tesla coil at 10kV, your homemade capacitor should be designed to withstand at least 40kV. This is just for safety reasons. Obviously you have a little bit of play, so don't stress too much if you can only get a 35kV rating, but keep it as close to or above 40kV as possible if driving at 10kV.

One thing to note about any homemade capacitor, make sure it's well-vented. Beer bottles have been known to explode if they are tightly sealed. Internal arcing can cause the air inside the bottle to heat up and expand, causing significant pressure buildup. Make sure to leave a vent hole in the cap. The same goes for any other type of homemade capacitor.

I hope this helps! Feel free to let me know if you have any further questions. Good luck!

Regards,
Matt

 

[radicalsniper]

Thank you for your reply! Actually I have another question. According to this site, type 4b flybacks might give us 80kV which we assumed to be the output of our flyback transformer for safety purposes.

Making its output bigger by upgrading it to a ZVS driver, we don't know what its output voltage will be so we planned to make bucket capacitors in series. Now that you've suggested a better alternative, we'll just build bottle capacitors in series but we don't know how many.

Question: Do you actually know the typical range of the output of a type 4B flyback transformer? I personally don't believe that it is 80kV (correct me if I'm wrong).

Sorry for bothering you with many questions, but we just want to create a Tesla coil in one try without blowing anything up.

We used this schematic in creating the driver

And this is our Tesla coil

**broken link removed**

 

[DerStrom8]

Thank you for your reply! Actually I have another question. According to this site, type 4b flybacks might give us 80kV which we assumed to be the output of our flyback transformer for safety purposes.

Making its output bigger by upgrading it to a ZVS driver, we don't know what its output voltage will be so we planned to make bucket capacitors in series. Now that you've suggested a better alternative, we'll just build bottle capacitors in series but we don't know how many.

Question: Do you actually know the typical range of the output of a type 4B flyback transformer? I personally don't believe that it is 80kV (correct me if I'm wrong).

In all honesty, it would be MUCH easier for you if you chose a standard AC transformer such as a neon sign transformer or an oil burner ignition transformer. They are far easier to use, and their output voltage is fairly constant. That would also allow you to make the proper calculations for your coil, and figure out the required tank capacitance. Sure, you may get lucky by just throwing some beer bottle capacitors together, but it is not very likely.

Sorry for bothering you with many questions, but we just want to create a Tesla coil in one try without blowing anything up.

It's no bother, it's what we're here for. The important thing you need to know is that it is unlikely that the Tesla coil will work the first time (in one try). TCs must be perfectly tuned in order to work properly, and without doing all of the calculations with the exact values (inductance and parasitic capacitance of the secondary, capacitance of the topload, inductance and parasitic capacitance of the primary, capacitance of the tank cap, etc) it will almost definitely not be tuned right off the bat.

Do you know how a Tesla coil works? Explain to me (in your own words) the theory of operation and it will give me an idea of where you are. It'll tell me where I need to start.

Regards,
Matt

 

[radicalsniper]

Actually, we calculated its parameters. We used the equations that is also used by Connick and Behrend and also tips from Jadwani in constructing the Tesla coil.

We then compared it with the software called TeslaMap.

Theory of operation of a Tesla coil is that when the spark gap is open the supply charges the tank capacitor and then when the capacitor's voltage is very high, breakdown will occur; bridging the gap between thus connecting the capacitor across the primary winding. Because of magnetic coupling between the primary and secondary, energy is being transferred from the primary to the secondary and there will also be a point in time that the top load cannot retain the charge anymore, causing breakdown in the form of corona discharge.

I could send to you our documentation of our project so that maybe you can read and point out what is wrong.

We actually considered to have NSTs but here in our place, Flyback transformers are the most common thing. We searched for available NSTs online but time is our enemy and we will be late if we ordered it online.

So, the output voltage of a ZVS driver can only be measured by a high voltage probe? There is no specifications or whatsoever about its rated output voltage?

 

[DerStrom8]

Theory of operation of a Tesla coil is that when the spark gap is open the supply charges the tank capacitor and then when the capacitor's voltage is very high, breakdown will occur; bridging the gap between thus connecting the capacitor across the primary winding. Because of magnetic coupling between the primary and secondary, energy is being transferred from the primary to the secondary and there will also be a point in time that the top load cannot retain the charge anymore, causing breakdown in the form of corona discharge.

That's a very cut-and-dried answer, I see it all over the internet all the time and it suggests you don't actually fully understand what's going on. Seeing as you don't even mention resonant frequencies, it makes me think you're not completely clear on the theory of operation. I suggest reading through my post here: https://www.electro-tech-online.com/blog-entries/building-a-drsstc-pt-1-background.753/
It is originally written for my DRSSTC build, but it gives a little bit of background on the operation of a SGTC.

Basically, the capacitor charges during the first half cycle of the AC input. If the capacitor is matched properly to the transformer, the capacitor reaches a voltage high enough to break down the air in the spark gap half way though the cycle, and that sends all of the energy stored in the capacitor into the primary coil. This forms a resonant LC (inductive and capacitive) circuit. When you connect an inductor and a capacitor together, they charge and discharge each other alternately (a discharging inductor charges the capacitor, and a discharging capacitor charges the inductor). This creates an alternating current oscillation. Simply saying "because of magnetic coupling between the primary and secondary" is not enough. You need to understand that this alternating current in the primary tank circuit (the tank capacitor and primary coil) creates an expanding and collapsing magnetic field, and that is what induces current in the secondary coil. The goal is to match the resonant frequency of the primary tank circuit (primary capacitor and primary coil) with the resonant frequency of the secondary tank circuit (the secondary coil and the topload).

I could send to you our documentation of our project so that maybe you can read and point out what is wrong.

That would be great. I'd be interested to see your work so far.

We actually considered to have NSTs but here in our place, Flyback transformers are the most common thing. We searched for available NSTs online but time is our enemy and we will be late if we ordered it online.

My concern is that you don't know what the output voltage of your flyback transformer is, and it will be somewhat difficult to control. Also, the output is not a clean sine wave, so your calculations for the tank capacitor may not be correct -- The efficiency will likely be much lower, and this may cause the coil to not work at all. Will it be considered "late" if it doesn't work at all due to improper calculations? I suggest looking for local neon sign repair shops or HVAC supply/repair shops. You may be able to find a secondhand transformer (either a neon sign transformer or an oil burner ignition transformer) from one of them for cheap or free. I can guarantee it would be a much better option (and far more likely to work) than a ZVS flyback driver.

So, the output voltage of a ZVS driver can only be measured by a high voltage probe? There is no specifications or whatsoever about its rated output voltage?

Many high voltage probes can't measure anywhere near the 80kV value, and the ones that can are very expensive. You can get a rough estimate of the output voltage by assuming 1kV for every mm that the spark can jump. So, if the spark jumps when you're 10mm away, it's probably roughly 10kV. This varies significantly with humidity, air pollution, altitude, etc. though, so this cannot be used as an actual measurement. It's just a very rough estimate.

 

[radicalsniper]

My concern is that you don't know what the output voltage of your flyback transformer is, and it will be somewhat difficult to control. Also, the output is not a clean sine wave, so your calculations for the tank capacitor may not be correct

This is also what is bothering me.

Basically, the capacitor charges during the first half cycle of the AC input. If the capacitor is matched properly to the transformer, the capacitor reaches a voltage high enough to break down the air in the spark gap half way though the cycle, and that sends all of the energy stored in the capacitor into the primary coil. This forms a resonant LC (inductive and capacitive) circuit. When you connect an inductor and a capacitor together, they charge and discharge each other alternately (a discharging inductor charges the capacitor, and a discharging capacitor charges the inductor). This creates an alternating current oscillation. Simply saying "because of magnetic coupling between the primary and secondary" is not enough. You need to understand that this alternating current in the primary tank circuit (the tank capacitor and primary coil) creates an expanding and collapsing magnetic field, and that is what induces current in the secondary coil. The goal is to match the resonant frequency of the primary tank circuit (primary capacitor and primary coil) with the resonant frequency of the secondary tank circuit (the secondary coil and the topload).

Nice explanation! I actually now see the need to read more about Tesla coils. I read it somewhere that it need to be in resonance so that it will produce the longest arcs possible.

And also, is there a formula that will calculate a Tesla coil's efficiency? What I only read is how Fritz calculated the efficiency, and he said that we must first define what efficiency are we talking about. It can be read here.

Maybe you can send to me your email so that I can send the documentation of our project. Thanks!

 

[DerStrom8]

Nice explanation! I actually now see the need to read more about Tesla coils. I read it somewhere that it need to be in resonance so that it will produce the longest arcs possible.

And also, is there a formula that will calculate a Tesla coil's efficiency? What I only read is how Fritz calculated the efficiency, and he said that we must first define what efficiency are we talking about. It can be read here.

This is a great point, "efficiency" can be defined as the efficiency of the transfer of energy stored in the capacitor passed on to the secondary coil, or the efficiency of the transformer, or the efficiency of the entire setup. The simplest formula is Pout = Pin. You're going to lose a lot of power in the form of heat, mechanical vibration, etc, so the electrical power output will be quite a bit less than the electrical power input. There is not a simple formula for calculating power in the entire system, as there are far too many variables to consider. The best you can do is choose materials that have as little loss as possible. For example, let's suppose you used black PVC pipe for your secondary coil. Black PVC tubing tends to have relatively high carbon content. Since carbon is a resistive material, it will act as a resistor between the secondary coil turns, leading to losses. Another example, if your primary tank capacitor has a high ESR (that is, equivalent series resistance, which acts as a resistor in series with the capacitor), you will experience a voltage drop across that resistor. The current flowing through it will also cause it to heat up, leading to losses in the form of heat. Also, capacitors have a leakage current, which means there is effectively a resistor in parallel with the capacitor, so some current gets through even when it theoretically shouldn't. This current through the resistor also causes it to heat up, leading to further losses.

As you can see, there are many, many variables at play, and figuring them all out is practically impossible. Instead, think of "efficiency" as a relative term. Example: A white PVC secondary will allow the coil to be "relatively" more efficient than if you use a black PVC secondary. A tank capacitor with a low ESR is "relatively" more efficient than one with a high ESR. This can go on and on. The trick is just going with the most (relatively) efficient options, rather than trying to calculate it all. Does this make sense? I hope I'm not being too confusing

Regards,
Matt

 

[radicalsniper]

I see! I think I understand where you are going. Thanks for the answers!

Maybe after several readings I may comment here again if some concepts confuse me, I'll be glad if you'll answer!

Thank you and good day!