Re: general questions about tesla coils
Steve Ward, Thu Nov 16 2006, 04:00AM

1. I see that many use varnish on the secondary and some use A LOT of varnish (>20 coats). Does it matter for performance? I believe that varnish adds to the stray capacitance of the coil. Is this a good thing?

It has no real effect on the self capacitance of the coil since the majority of the dielectric from coil to ground is air anyway. I usually use about 5 coats of varnish, its mostly for protection against scratches. More recently i have found epoxy to work out nicely. One brand in particular, Envirotex (i think) works very nicely. You want the stuff that is made for covering counter tops or furnature. Epoxy needs only 1 coat, and doesnt stink up the place if you have to do it indoors (good in winter time).

2. I have seen various primary designs: helix, spiral, cone. I can think for some reasons for this: available materials for the builder, coupling, flash-over, aesthetics. Is there a theoretical *best* design or is it just apples vs oranges?

The geometry of a helical primary allows for the highest coupling (which is far greater than we desire for TCs). The cone provides higher coupling, but also provides added voltage standoff between windings compared to the helical. Flat gives a lower coupling, but the best voltage standoff of all. Basically, you want the coupling high enough for good energy transfer (you want it to be rather quick), but you dont want racing sparks or sparking between the 2 coils. For SGTCs, flat primaries usually offer enough coupling, and its not worth the hassle of building a cone (though i think they look cool). You can always raise up the secondary to lower coupling if its too high.

3. My design (15kv/120ma primary with a synchronous 120bps spark gap) could use a resonant cap value of about 0.021uf and a LTR value of about 0.055uf. As long as Ls*Cs = Lp*Cp is it acceptable to have a tank capacitor *close* to one of these values or should it be as exact as possible. I have two 0.025uf Maxwell capacitors and if necessary I have some smaller ones I can chain in parallel/series to make up the difference; otherwise, I was going to use 0.05uf as my tank.

Provided you stay clear of the resonant value at 60hz (to avoid resonant rise between NST and tank cap) then you should be OK. The .055uF value was likely determined from simulation to provide the best power use of the transformer (maximum bang energy possible given the supply constraints). Its not critical to get exactly this value, .05uF will work just fine. Going larger will give you a lower charging voltage. Going with a cap thats too small will probably not raise the charging voltage, but the lower capacitance will mean smaller energy storage per bang. If using maxwell capacitors, be sure they are rated for 40kV, otherwise im not sure if i would risk using them. The peak charging voltage will be around 20kV, but the high primary Q causes a nearly 100% voltage reversal, so the net voltage rating should be 40kV for this type of use. Note that the MMCs we all use are pushed way beyond their ratings, and this seems to work ok. I have heard of people killing the 35kV variety of maxwells with 15kvAC power sources. Just a warning, as i know how much it would suck to blow up your nice caps.

4. I have seen a Freau formula for spark length as L (inches) = 1.7*sqrt(WallPlugPower)
Is the Wall Plug Power in watts the value without using PFC capacitors? Clearly, if I use PFC then the wall plug power significantly drops. Which value is used?

Note, power (watts) is not necessarily the the VA. Your power (watts) should be about the same with and without PFC, but your current draw will be reduced with PFC. I believe that John assumes one will use real power, and not the VA. I know that my 15kV 120mA powered coil drew about 14A in operation and produced about 66" sparks using a 120BPS SRSG and i think a .045uF MMC (ive long since sold it). With some extra tweaking it should have been capable of more i think, but it was good enough at the time.

Re: general questions about tesla coils
IamSmooth, Thu Nov 16 2006, 04:13AM

Steve, your reply clears a lot up. Thanks


When I connected the PFC caps to one 15kv/60ma NST I reduced the amps drawn from about 8.9 to 2.1 when I was drawing an arc. This is only about 240 watts, yet I keep hearing everyone talk about the NST being 900watts (when I really think they mean VA). So, for each 15kv/60ma what value should I use for the "power" in the formula?


As far as my maxwells they are each rated for 50kv so I am guessing I will be fine.

Based on your comments I might go for a small cone with a 5-10% angle.

Re: general questions about tesla coils
Steve Ward, Thu Nov 16 2006, 06:30AM

When I connected the PFC caps to one 15kv/60ma NST I reduced the amps drawn from about 8.9 to 2.1 when I was drawing an arc. This is only about 240 watts, yet I keep hearing everyone talk about the NST being 900watts (when I really think they mean VA). So, for each 15kv/60ma what value should I use for the "power" in the formula?

This is tricky. Capacitor charging is a bit different than drawing sparks. When drawing an arc with non PFC'd NST, the load looks mostly inductive. It doesnt help any that there are magnetic shunts, effectively there to limit the secondary current, but do not necessarily decrease the primary current draw, so you end up with a lot of reactive power, and not much real power, which you discovered. When people quote NST power, they probably should say VA.

Anyway, when a NST is charging a capacitor, you can get more real power out of it than when drawing an arc, this is due to the better impedance match. The capacitance cancels out some of the reactive component on the NST. I havent worked out the details for myself, but i believe you can actually extract about as much power (W) as they are rated for in VA. A lot of it has to do with the size of the tank capacitor and the BPS of the coil. When people say to use 2.5X the LTR value, i think this is to maximize the amount of power (W) used from the NST

I wouldnt really worry much about the power vs VA stuff right now unless you really must know how "efficient" your TC is. If you are looking for a value to plug into the Freau formula, id say just say that the VA rating is your input in watts, and that will probably be about the biggest spark you would expect. Please note that this formula is just an emperically derived guess at what the maximum spark length should be. Most coils dont meet the mark, and then again, some exceed it. Of course, not everyone measures power properly, if at all (they usually go by the VA rating of the NST). I should mention its possible to exceed VA rating of a NST given the proper setup, so some people might "cheat".

Re: general questions about tesla coils
Hazmatt_(The Underdog), Thu Nov 16 2006, 06:22PM

*** with regard to Cprimary ***

I am researching the sizing of the Cprimary capacitor, and if you are careful with your calculations you CAN find the value which gives you the nameplate VA. But you must be careful not to exceed this because the transformer will die.

Re: general questions about tesla coils
IamSmooth, Thu Nov 16 2006, 07:13PM

I have two more "general" questions if none of you mind:

1. I have a synchronous spark gap using an Oriental Motor. Will I have any trouble with magnetic or RF interferance if I place this underneath the primary? I plan to have the metal housing of the motor grounded to the earth.

2. I am going to use a 6 inch diameter secondary. The aspect ratio can be 1:3 to 1:5. I've seen some emperically derived formulas that suggest that certain ratios are better for a given input power level. I am aware of tuning Ls*(Cs+Ct) once I have picked the values to equal Lp*Cp. Is there any good reason to go 1:3, 1:4 or 1:5?

Re: general questions about tesla coils
Steve Ward, Thu Nov 16 2006, 10:55PM

I am researching the sizing of the Cprimary capacitor, and if you are careful with your calculations you CAN find the value which gives you the nameplate VA. But you must be careful not to exceed this because the transformer will die.

Great, id like to hear about your results. It should be possible to simulate this, provided you can model every aspect of the NST (i still dont know how to model core materials yet, so im not sure).

1. I have a synchronous spark gap using an Oriental Motor. Will I have any trouble with magnetic or RF interferance if I place this underneath the primary? I plan to have the metal housing of the motor grounded to the earth.

Shouldnt be a problem since the primary B-field is oscillating much faster than the 60hz field that the motor runs on. Also, the steel case of the motor should block most of those effects anyway. I never payed any attention to this sort of issue.

2. I am going to use a 6 inch diameter secondary. The aspect ratio can be 1:3 to 1:5. I've seen some emperically derived formulas that suggest that certain ratios are better for a given input power level. I am aware of tuning Ls*(Cs+Ct) once I have picked the values to equal Lp*Cp. Is there any good reason to go 1:3, 1:4 or 1:5?

Id say it depends on what spark length you expect. You might get up near 6 foot sparks with that supply. In my experience, you can push a secondary to produce sparks about 3X its winding length, but in general it seems more efficient if you aim for only 2X or so. So you may want to go with a 30-36" winding... 30" is probably just fine. 6" diameter sounds good too... again, just going from experience. Remember, you can basically make Ls and Cs whatever you want. Cp is pretty well defined for the power supply, so you just change Lp. Typically, its better to go with lower operating frequencies so that Lp is large, making Zsurge (sqrt(L/C)) large, keeping the peak primary current lower. This lowers the spark gap losses to some degree, and many find this to be noticable with spark length per input power.

Re: general questions about tesla coils
IamSmooth, Sun Nov 19 2006, 02:49AM

I have seen the spark gap parallel with the HV terminals of the transformer and in series, interchanged with the capacitor position. I have read that the former way is better as there is less "stress" on the transformer. Can anyone explain why this is the case?

Re: general questions about tesla coils
Colin 99, Sun Nov 19 2006, 09:40PM

You want to have the spark gap across the NSTs to prevent the voltage across them from getting too high, otherwise you could damage your NSTs. The primary capacitor forms a resonant circuit with the NST, If the capacitor is across the NST, the voltage can exceed the rating of the NST and damage them. This happens with Cltr also - producing a resonant rise across the capacitor and NST. With the spark gap across the NST (set right and safety gaps), it prevents the voltage from getting too high on each half cycle.

Also a Terry filter is suggested between your NSTs and spark gap to prevent high frequencies form getting back into the NSTs. The Terry filter is an RC filter with MOVs and safety gaps across them.

Shaun

Re: general questions about tesla coils
IamSmooth, Mon Nov 20 2006, 12:05AM

Thanks Colin, but maybe I wasn't clear in my question. My coil is going to have a synch spark gap with a safety static gap in case the synch motor fails. I am going to have a Terry Filter with a static gap for each end of the HV terminals of the NST with a ground in between.

My question is simply is there any difference in switching the position of the capacitor and the spark gap (my sych spark gap) if I have the other items in place? I guess the question can also be does the capacitor stress the NST more if it is parallel or series with the NST if I already have a safety spark gap across the HV ends of the NST to catch HV spikes?

Re: general questions about tesla coils
Hazmatt_(The Underdog), Mon Nov 20 2006, 02:08AM

What I have noticed in simulation is that your tank ring-down voltage peaks at roughly twice the input charging voltage. That means if have a 10KV NST and you are charging an LTR capacitor, the decaying ring-down will oscillate between +10kV and -10KV while decaying exponentionally.

You should also consider the current loops. During simulation you can expect several hundred amps through the tank circuit during ringdown. This is a low impedance path, but the transformer is a high impedance path (reative to primary). The current will go through both primary inductor and it will feed back through the transformer as well, ie. two current paths. You could think of it as the primary tank getting 99% and the transformer having 1% surging back through it, conceptually speaking.

Now if you wanted to put a safety gap in parallel with the transformer and the capacitor in parallel with the transformer, then that gap would fire shorting the transformer and the capacitor and close the circuit to the Lprimary. The ringdown would be reflected to the transformer as the current pulses back and forth through the capacitor. This could spell a lot of trouble for your transformer.

Generally you want to keep the high ringdown effects isolated from the transformer so that it operates as intended without a lot of stresses. The insulation class is pretty poor on these things. If you talk to Franceformer, you will find out that transformers are insulated for .707 Vpk because they're intended to be connected to a diode all the time (neon sign tube)

A good filter design can reduce these peaks back down to nominal plate ratings. I've already finalized my filter design and it looks great.

I would highly recommend to anyone to simulate their system and find the bugs before building. Simulation is a great tool to help you 'see' whats happening.

Re: general questions about tesla coils
Steve Conner, Mon Nov 20 2006, 05:34PM

The spark gap (both safety and regular) should always be across the NST terminals. Terry Fritz proved experimentally that this stresses the transformer least, because it doesn't see the oscillating RF voltage (as mentioned by hazmatt) that it would see across the capacitor.

Re: general questions about tesla coils
IamSmooth, Sun Dec 03 2006, 09:59PM

I have a question about proper grounding:

I know that the secondary RF ground should be separate. However, does the ground for the NST (Mains) have to be separate from the ground for the Terry Fritz filter or can they be combined? I am referring to the grounded center of the spark gap that is used to catch high-voltage spikes where each end of the spark gap is connected to the HV terminals of the NST.

Re: general questions about tesla coils
Hazmatt_(The Underdog), Sun Dec 03 2006, 11:26PM

The NST housing and midpoint on the non GFCI units share a ground that should be grounded to your AC line ground to prevent any voltage rise on the housing. I would use this midpoint ground as my 0 reference for safety gaps.

The newer GFCI units have a midpoint isolated from the housing that is about ~60 V above line ground. This voltage is on this small isolated bushing for the GFCI reference/fault condition detector. If you have a GFCI unit I would recommend using the isolated midpoint for your protective gap.

Re: general questions about tesla coils
IamSmooth, Mon Dec 04 2006, 01:54AM

thanks

Re: general questions about tesla coils
IamSmooth, Fri Dec 08 2006, 06:22PM

I want to start adjusting my static safety gap that I have with my sych gap. I also want to set the NST fail-safe gap on my filter. Is there a problem with running the NST attached to the primary LC circuit *without* the secondary at half power? Or, will this power not get dissapated fast enough and cause overload problems if there is no secondary to transfer the power for streamer production?

Second question: if one has a properly tuned coil, does it matter what is the value of the resonant frequency? That is, is there a difference in performance if the frequency is 250 khz or 80 khz? The only thing I considered was that at higher frequencies there could be higher RF losses.

Re: general questions about tesla coils
Hazmatt_(The Underdog), Fri Dec 08 2006, 07:55PM

I'm not understanding your first question the way you wrote it. But I can say this, if you are attempting to adjust a static or rotary gap and are varying the power input while connected to your LC primary tank, you pose a real threat to your transformer by means of dielectric breakdown. What do I mean by this? Well while your LC tank is connected to your transformer and say you have no switching of the gap, there is a voltage rise in the transformer above what it is rated for, so eventually you're going to cause a carbon track across one of the transformer secondaries and fault the transformer. I know this because I was doing some dumb things trying to see if my static gap was set far enough apart, and it was too far apart. The transformer autopsy revealed the carbon track and I finally knew why I had lost ~half my streamer length.

IN short, if you turn on the power to your primary circuit and you don't get any breakover then the gap is too wide and you might be burning the insulation on the winding.


Second question: The way I see it is a function of input power. When I try to design a coil I am trying to make it a good size for the power applied, which then determines the resonant frequency. Low power--> tiny coil--> high Fo. If you apply too much power to the coil then it will arc to itself, its strike rail, primary, possibly catch on fire, all that stuff.

Then with tuneing I just go for maximum streamer length. I did measure the inductance of all the connecting leads and tuned it for LpCp = LsCs, and it was right there in the ballpark the first time. So if you know the secondary characteristics, measure the primary leads, calculate, and you'll be really close to where you need to be.

Re: general questions about tesla coils
IamSmooth, Fri Dec 08 2006, 08:33PM

With regards to the first question: If I want to make sure my safety gap works and I am using a rotary gap, what is the safest way to test the static gap without frying my NST?