I'm back, but now with government funding :D
Tim Koene, Mon Sept 25 2006, 07:30PMHello all, some of you might remember me ;)
I have been asked to make a Tesla Coil for the EE HV lab at university, to use as a demo piece.
The equipment I can use:
-36KVrms 4KW potential transformer
-Several 2nf 40KV very expensive murata ceramic capacitors
-access to lathes and winding tools and wire
-full funding for the project, so with proper motivation anything can be bought.
Now, I have been gone from the hobby for a long time but this seems the golden way to get back in.
The thing I want from you good people is hints tips and pointers. What would you build, when you were in my shoes?
Thanks for any input you can give me.
Tim
Re: I'm back, but now with government funding :D
Marko, Mon Sept 25 2006, 08:01PM
hi..
afaik, 36kV is pretty high in desired TC voltages. You may come to problems with making a capacitor able to withstand that, and in your place I would really hate to ruin those pretty caps!
Marko, Mon Sept 25 2006, 08:01PM
hi..
afaik, 36kV is pretty high in desired TC voltages. You may come to problems with making a capacitor able to withstand that, and in your place I would really hate to ruin those pretty caps!
Re: I'm back, but now with government funding :D
Reaching, Mon Sept 25 2006, 08:10PM
yeah, you will need 2 in series maybe with corona surpressors or so. how many caps do you have? for a big TC with those caps the mmc will have 100 or more caps in it. looks like the caps arent suitable for a 4kw monster transformer. but if i were in your shoes i would buy a few maxwell pulse caps, a sync motor a spun toroid and some other things and would build a very big sgtc with sync rotary spark gap etc, a big scary monster
Reaching, Mon Sept 25 2006, 08:10PM
yeah, you will need 2 in series maybe with corona surpressors or so. how many caps do you have? for a big TC with those caps the mmc will have 100 or more caps in it. looks like the caps arent suitable for a 4kw monster transformer. but if i were in your shoes i would buy a few maxwell pulse caps, a sync motor a spun toroid and some other things and would build a very big sgtc with sync rotary spark gap etc, a big scary monster
Re: I'm back, but now with government funding :D
Marko, Mon Sept 25 2006, 08:31PM
200nF at 40kV would be 320J of bang energy wich ''looks'' a bit too muh for 4kW coil, so he obivously doesn't need 200 of those caps.
Ignoring PF you can calculate that transformer gives roughly 110 miliamps.
For 60Hz he needs 8,1nF MMC for resonance, and I guess he will want to run at some 10nF wich is still quite a lot of caps if you want them to hold the voltage (2 in series for 80kV = 20caps).
Bigger problem is that ceramic knobs tend to change capacitance with heating (even if they are HF power rated, they do for a degree) and lots of people claim that they fail after long duties. Don't know for your murata caps though..
Marko, Mon Sept 25 2006, 08:31PM
200nF at 40kV would be 320J of bang energy wich ''looks'' a bit too muh for 4kW coil, so he obivously doesn't need 200 of those caps.
Ignoring PF you can calculate that transformer gives roughly 110 miliamps.
For 60Hz he needs 8,1nF MMC for resonance, and I guess he will want to run at some 10nF wich is still quite a lot of caps if you want them to hold the voltage (2 in series for 80kV = 20caps).
Bigger problem is that ceramic knobs tend to change capacitance with heating (even if they are HF power rated, they do for a degree) and lots of people claim that they fail after long duties. Don't know for your murata caps though..
Re: I'm back, but now with government funding :D
Reaching, Mon Sept 25 2006, 09:21PM
the only thing i wanted to say is that he would need a lot of this caps and that there are better uses for those expensive doorknobs as to destroy them in a tc. they would crack after a few seconds ,expensive or not they are ceramik caps. and 8nf for a big tc
how many primary turns do you want to use 60? he would need 60nF or so to match a secondary fres below 100khz (i think it will be a huge secondary) with a decent amount of primary turns 12 or 15 or whatever
Reaching, Mon Sept 25 2006, 09:21PM
the only thing i wanted to say is that he would need a lot of this caps and that there are better uses for those expensive doorknobs as to destroy them in a tc. they would crack after a few seconds ,expensive or not they are ceramik caps. and 8nf for a big tc
how many primary turns do you want to use 60? he would need 60nF or so to match a secondary fres below 100khz (i think it will be a huge secondary) with a decent amount of primary turns 12 or 15 or whateverRe: I'm back, but now with government funding :D
Marko, Mon Sept 25 2006, 10:25PM
Yep, just a bit more than 8 nF for a big TC because it operates at unusually high voltages.
Cap much bigger than that would be a big waste of resources and is a wrong way to get a coil into resonance. Your 60nF cap would load the transformer way too much and would never get charged to it's full voltage. And even if you build DC charging you wouldn't be able to mantain breakrate with that amount of power (60nF cap would store 150 joules at 36*1.414 ~ 50kV and would require 15 kilowatts for just 100bps.)
And ofcourse more obivous thing, even a huge secondary like TDU's biggest coil probably would probably blow with that amount of bang energy, and Koene's coil is actually somewhere in class of TDU's 6 inch coil wich uses 6J bang and works beautifully
I think that it would be a much better idea to wind the secondary with thicker wire for higher frequency, 200..400kHz, keeping Q-factor high, and using somewhat bigger, 15...25uH primary keeping surge impedance of primary tank low, what yelds lower peak currents and lower spark gap losses.
Experienced coilers like terry, TDU, Conner and etc. could probably help you much more with that thing, transforme voltage is a bit high but this coil could turn into something very nice when done.
Marko, Mon Sept 25 2006, 10:25PM
Yep, just a bit more than 8 nF for a big TC because it operates at unusually high voltages.
Cap much bigger than that would be a big waste of resources and is a wrong way to get a coil into resonance. Your 60nF cap would load the transformer way too much and would never get charged to it's full voltage. And even if you build DC charging you wouldn't be able to mantain breakrate with that amount of power (60nF cap would store 150 joules at 36*1.414 ~ 50kV and would require 15 kilowatts for just 100bps.)
And ofcourse more obivous thing, even a huge secondary like TDU's biggest coil probably would probably blow with that amount of bang energy, and Koene's coil is actually somewhere in class of TDU's 6 inch coil wich uses 6J bang and works beautifully
I think that it would be a much better idea to wind the secondary with thicker wire for higher frequency, 200..400kHz, keeping Q-factor high, and using somewhat bigger, 15...25uH primary keeping surge impedance of primary tank low, what yelds lower peak currents and lower spark gap losses.
Experienced coilers like terry, TDU, Conner and etc. could probably help you much more with that thing, transforme voltage is a bit high but this coil could turn into something very nice when done.
Re: I'm back, but now with government funding :D
Tim Koene, Mon Sept 25 2006, 10:39PM
Thanks for the replies so far.
For the time being let's just assume that the coil will be about 10" diam and about 60"high, wound with whatever number of turns is optimal. I have about 10 of those caps and I know those will do fine in SGTC service. Each doorknob is the size of a fist, and doorknobs work great in TC service. Good ones that is. These are ;)
The point is, for some demos we want this thing to be indoor and not a huge monster, just some nice sparks but once or twice a year we have a huge outdoor/evening event where we could ride it into one of our public squares and give a show.
A big pulse cap might be arranged, but I would like to use the doorknobs we have.
I will talk to the people involved to get you more information :) Thanks for the replies! keep em coming!
Tim Koene, Mon Sept 25 2006, 10:39PM
Thanks for the replies so far.
For the time being let's just assume that the coil will be about 10" diam and about 60"high, wound with whatever number of turns is optimal. I have about 10 of those caps and I know those will do fine in SGTC service. Each doorknob is the size of a fist, and doorknobs work great in TC service. Good ones that is. These are ;)
The point is, for some demos we want this thing to be indoor and not a huge monster, just some nice sparks but once or twice a year we have a huge outdoor/evening event where we could ride it into one of our public squares and give a show.
A big pulse cap might be arranged, but I would like to use the doorknobs we have.
I will talk to the people involved to get you more information :) Thanks for the replies! keep em coming!
Re: I'm back, but now with government funding :D
Sulaiman, Mon Sept 25 2006, 10:43PM
I guess first I'd check the temperature coeficient of the capacitors;
If NP0 there should be little problem with resonant frequency,
and GOOD ceramic capacitors have surprisingly low dissipation factor,
which is the next thing to check.
Any specs. on the capacitors?
Sulaiman, Mon Sept 25 2006, 10:43PM
I guess first I'd check the temperature coeficient of the capacitors;
If NP0 there should be little problem with resonant frequency,
and GOOD ceramic capacitors have surprisingly low dissipation factor,
which is the next thing to check.
Any specs. on the capacitors?
Re: I'm back, but now with government funding :D
Marko, Mon Sept 25 2006, 11:00PM
that looks like a real beast over there, and an overkill for 4kW of power input (wich isn't all that much actually, close to 2 big MOT's).
Your maximum predicted spark length for 4kW would be around 100 inches and sparks would be dwarfed by secondary.
Diameter is probably OK but I would cut the height to about a metre for closely 1:4 ratio.
Since you are allowed high resonant frequency you could probably get on with even smaller secondary (TDU help there ^ ^ )
Marko, Mon Sept 25 2006, 11:00PM
For the time being let's just assume that the coil will be about 10" diam and about 60"high
that looks like a real beast over there, and an overkill for 4kW of power input (wich isn't all that much actually, close to 2 big MOT's). Your maximum predicted spark length for 4kW would be around 100 inches and sparks would be dwarfed by secondary.
Diameter is probably OK but I would cut the height to about a metre for closely 1:4 ratio.
Since you are allowed high resonant frequency you could probably get on with even smaller secondary (TDU help there ^ ^ )
Re: I'm back, but now with government funding :D
Steve Conner, Tue Sept 26 2006, 12:12AM
Hi Tim! Why not use your time cone to ask Tesla what to build.
It sounds like you have all the parts except for the spark gap and ballast. A sync rotary running at 200Hz would probably be nice, and a ballast like Richie Burnett's: 
To get impressive sparks you ought to aim for a bang energy high enough that you can draw the full 4kW at 200Hz. So that would be 20J I guess. If you wanted the maximum spark length per watt used, 100Hz seems to be better, but I think 200Hz arcs look brighter and more scary.
I believe potential transformers can be overloaded quite a lot in intermittent duty, something to bear in mind if you're in a power crazed mood. Pretty much all transformers can deliver twice their rated power for a few minutes.
I used to do Tesla coil demos for the university EE department, even though I graduated a long time ago, they still asked me to do it because my coils pwned theirs. I used the OLTC2 a couple of times, but the ceiling of the demo room was too low to let it run at full power, so I started bringing my DRSSTC instead. At the last demo, I overheated the tank cap bad enough that it leaked wax, and burnt a hole in the secondary, but the kids loved it
Steve Conner, Tue Sept 26 2006, 12:12AM
Hi Tim! Why not use your time cone to ask Tesla what to build.
It sounds like you have all the parts except for the spark gap and ballast. A sync rotary running at 200Hz would probably be nice, and a ballast like Richie Burnett's: To get impressive sparks you ought to aim for a bang energy high enough that you can draw the full 4kW at 200Hz. So that would be 20J I guess. If you wanted the maximum spark length per watt used, 100Hz seems to be better, but I think 200Hz arcs look brighter and more scary.
I believe potential transformers can be overloaded quite a lot in intermittent duty, something to bear in mind if you're in a power crazed mood. Pretty much all transformers can deliver twice their rated power for a few minutes.
I used to do Tesla coil demos for the university EE department, even though I graduated a long time ago, they still asked me to do it because my coils pwned theirs. I used the OLTC2 a couple of times, but the ceiling of the demo room was too low to let it run at full power, so I started bringing my DRSSTC
instead. At the last demo, I overheated the tank cap bad enough that it leaked wax, and burnt a hole in the secondary, but the kids loved it Re: I'm back, but now with government funding :D
Marko, Tue Sept 26 2006, 12:19AM
Hm, ballasting the thing properly seems to be one of problems.
guessing he can't get such a ballast, would something like capacitive primary-side ballast work?
Marko, Tue Sept 26 2006, 12:19AM
Hm, ballasting the thing properly seems to be one of problems.
guessing he can't get such a ballast, would something like capacitive primary-side ballast work?
Re: I'm back, but now with government funding :D
Tim Koene, Tue Sept 26 2006, 08:01AM
I have a few 8a and 16a variable inductor style variacs, not the new galvanically separated two coil thingamajigs. Those could be used as a variable inductor. Oh and 4KW is a rough estimate. The PT is like This One only about 3 times as large. And maybe 60"might be a bit large, think i'll make it 40-50"then.
I will get on those capacitor readings today. Thanks for all the replies.
(Hiya Steve Conner, my time cone is in storage for the time being ;) )
Tim Koene, Tue Sept 26 2006, 08:01AM
I have a few 8a and 16a variable inductor style variacs, not the new galvanically separated two coil thingamajigs. Those could be used as a variable inductor. Oh and 4KW is a rough estimate. The PT is like This One only about 3 times as large. And maybe 60"might be a bit large, think i'll make it 40-50"then.
I will get on those capacitor readings today. Thanks for all the replies.
(Hiya Steve Conner, my time cone is in storage for the time being ;) )
Re: I'm back, but now with government funding :D
Hazmatt_(The Underdog), Tue Sept 26 2006, 11:49PM
I'm finishing my simulations for my 1KW Tesla Coil for CSULB right now, and I must say that its going to give us over 6 foot sparks, which is plenty for demonstrations.
Here is the major question I would ask myself: How much power do I really need?
High power comes at a large price because of mobility, breakers, huge sparks, power controllers, safety, and of course grounding.
* A 4kw sgtc can't easily be demonstrated in a lecture room or hall because of breakers, grounding, etc. and the shear weight of the beast!! Must be close to 400 lbs of equipment ...maybe more if you have multiple controllers for the input.
* Ballasting is a big problem. Large inductors with large sliding armatures increase weight and size of the control pannel and its components. You may consider using an SCR on-time phase controller. This is my plan for my 5kva step down for my PT.
*Grounding, do I need to say more? AT 1KVA I'm probably looking at 30 Amps RF into the ground. 4KVA well....that's enough to drive a pretty nice size motor.
*36KV input! Man....that cap will have to withstand huge voltage reversals. You may need 100KV+ ratings on your capacitor. Also, my Cx bank composed of doorknobs had LARGE flashovers from plate-plate at a pewny 12KVAC input.
*The ringdown voltage and reversals are going to kick your input voltage WAY up there. Your oscillating voltage could be up there at 160,000V or more!! This is a big dielectric problem for all components, not just primary or Cx!!
I would get my hands on PSpice or something and start a simulation right away and see what it gives you.
Just some thoughts.
Matt
Hazmatt_(The Underdog), Tue Sept 26 2006, 11:49PM
I'm finishing my simulations for my 1KW Tesla Coil for CSULB right now, and I must say that its going to give us over 6 foot sparks, which is plenty for demonstrations.
Here is the major question I would ask myself: How much power do I really need?
High power comes at a large price because of mobility, breakers, huge sparks, power controllers, safety, and of course grounding.
* A 4kw sgtc can't easily be demonstrated in a lecture room or hall because of breakers, grounding, etc. and the shear weight of the beast!! Must be close to 400 lbs of equipment ...maybe more if you have multiple controllers for the input.
* Ballasting is a big problem. Large inductors with large sliding armatures increase weight and size of the control pannel and its components. You may consider using an SCR on-time phase controller. This is my plan for my 5kva step down for my PT.
*Grounding, do I need to say more? AT 1KVA I'm probably looking at 30 Amps RF into the ground. 4KVA well....that's enough to drive a pretty nice size motor.
*36KV input! Man....that cap will have to withstand huge voltage reversals. You may need 100KV+ ratings on your capacitor. Also, my Cx bank composed of doorknobs had LARGE flashovers from plate-plate at a pewny 12KVAC input.
*The ringdown voltage and reversals are going to kick your input voltage WAY up there. Your oscillating voltage could be up there at 160,000V or more!! This is a big dielectric problem for all components, not just primary or Cx!!
I would get my hands on PSpice or something and start a simulation right away and see what it gives you.
Just some thoughts.
Matt
Re: I'm back, but now with government funding :D
Steve Ward, Wed Sept 27 2006, 02:21AM
Just as much reversal as any other TC... I dont see your point.
I dont know where you are pulling these figures from, and why you seem to think that 36kVAC input is any more "troublesome" with respect to the possibility of the capacitor voltage increasing beyond the supply rating. This TC will play by the same rules of electrical circuits as any other TC.
The tank cap should be rated for 2*Vpk (due to a near 100% voltage reversal), which in this case is about 100kV. MMCs (using film/foil PP caps) can usually really push the rating, but, if this is for a professional system, its best to use a proper rating.
I suggest using inductive ballasting to the effect of making your tank capacitor "LTR" (larger than resonant). You can figure out the resonant cap value by (mathematically) reflecting the primary inductive ballast to the secondary side (multiply by the turns ratio) and solving the equation for resonance at 50hz. Its been found that its pretty safe to then to use 1.5X this value to avoid the possibility of the capacitor voltage ringing up over many mains cycles.
The 50kV charging voltage presents some interesting issues with spark gap design. You would likely want to run a rotary gap at this power level. I would definately suggest using a series static gap as well (with forced air across it). Perhaps use large rounded electrodes for the rotary gap to prevent "long" sparks from forming between the electrodes. The series gap should help with this too.
Have you considered the amount of "noise" that will put back into the line?
Steve Ward, Wed Sept 27 2006, 02:21AM
*36KV input! Man....that cap will have to withstand huge voltage reversals.
Just as much reversal as any other TC... I dont see your point.
*The ringdown voltage and reversals are going to kick your input voltage WAY up there. Your oscillating voltage could be up there at 160,000V or more!! This is a big dielectric problem for all components, not just primary or Cx!!
I dont know where you are pulling these figures from, and why you seem to think that 36kVAC input is any more "troublesome" with respect to the possibility of the capacitor voltage increasing beyond the supply rating. This TC will play by the same rules of electrical circuits as any other TC.
The tank cap should be rated for 2*Vpk (due to a near 100% voltage reversal), which in this case is about 100kV. MMCs (using film/foil PP caps) can usually really push the rating, but, if this is for a professional system, its best to use a proper rating.
I suggest using inductive ballasting to the effect of making your tank capacitor "LTR" (larger than resonant). You can figure out the resonant cap value by (mathematically) reflecting the primary inductive ballast to the secondary side (multiply by the turns ratio) and solving the equation for resonance at 50hz. Its been found that its pretty safe to then to use 1.5X this value to avoid the possibility of the capacitor voltage ringing up over many mains cycles.
The 50kV charging voltage presents some interesting issues with spark gap design. You would likely want to run a rotary gap at this power level. I would definately suggest using a series static gap as well (with forced air across it). Perhaps use large rounded electrodes for the rotary gap to prevent "long" sparks from forming between the electrodes. The series gap should help with this too.
You may consider using an SCR on-time phase controller. This is my plan for my 5kva step down for my PT.
Have you considered the amount of "noise" that will put back into the line?
Re: I'm back, but now with government funding :D
HV Enthusiast, Wed Sept 27 2006, 03:11AM
I believe he is stating that the voltage reversals in this system which utilizes a 36kVrms transformer are considerably higher than that of the typical 14.4kVrms powered Tesla coil.
Personally, i would do everything i could to get rid of the 36kVrms transformer. That higher voltage just adds lots of complexities to the system and its components. I remember in several of Richard Hull's Tesla coil tapes that he was using 36kVrms potential transformers for some coils and if i remember correctly, stated that the higher voltage posed many unique problems which required specialized solutions which were not straightforward or obvious.
Tim already stated he had full funding for the project, so i'm sure he would be able to procure an inexpensive pole transformer. Dr. Resonance would sure be able to ship him one if one couldn't be found anywhere.
Believe me, the change from 35kV to 14.4kV will make MANY things much simpler.
HV Enthusiast, Wed Sept 27 2006, 03:11AM
I believe he is stating that the voltage reversals in this system which utilizes a 36kVrms transformer are considerably higher than that of the typical 14.4kVrms powered Tesla coil.
Personally, i would do everything i could to get rid of the 36kVrms transformer. That higher voltage just adds lots of complexities to the system and its components. I remember in several of Richard Hull's Tesla coil tapes that he was using 36kVrms potential transformers for some coils and if i remember correctly, stated that the higher voltage posed many unique problems which required specialized solutions which were not straightforward or obvious.
Tim already stated he had full funding for the project, so i'm sure he would be able to procure an inexpensive pole transformer. Dr. Resonance would sure be able to ship him one if one couldn't be found anywhere.
Believe me, the change from 35kV to 14.4kV will make MANY things much simpler.
Re: I'm back, but now with government funding :D
Hazmatt_(The Underdog), Wed Sept 27 2006, 03:26AM
I just took a look at my simulation profile and my ringdown voltage is about 20kv pk-pk from a 9kvac transformer.
36KVAC would then suggest that the peak ringdown voltages would be in excess of 72kvac, which is not as bad as I had thought, but its still a big insulation problem for convienent materials like wood. The primary might arc to itself or corona badly at these voltages.
I did mean the magnitude of ringdown voltage is very large in comparison to my system.
I'm going to stop being helpful because it just ain't worth it.
Matt
Hazmatt_(The Underdog), Wed Sept 27 2006, 03:26AM
I just took a look at my simulation profile and my ringdown voltage is about 20kv pk-pk from a 9kvac transformer.
36KVAC would then suggest that the peak ringdown voltages would be in excess of 72kvac, which is not as bad as I had thought, but its still a big insulation problem for convienent materials like wood. The primary might arc to itself or corona badly at these voltages.
I did mean the magnitude of ringdown voltage is very large in comparison to my system.
I'm going to stop being helpful because it just ain't worth it.
Matt
Re: I'm back, but now with government funding :D
Tim Koene, Wed Sept 27 2006, 08:52AM
Hi everyone,
I have been doing some research on the potential transformer in question, the exact ratings were missing because the faceplatge was screwed off. It turns out that it is a 3 phase line potential transformer and it has an output rated of 36kv/sqrt(3) which is closer to 20KV, which should pose considerably less problems.
Furthermore I would like to stress the point that I am aiming for about 1-1.5 metres of streamer, so a pole pig would be a bit overkill. Maybe when this medium sized coil is done and approved they want a large >>10KW setup.
I have some nicely designed forced air single electrode sparkgaps, (Which I could put in series) or I could make a nice 6-10 piece series gap. I would like to avoid rotary design unless really necessary. It adds complexity in the building, possible service when it breaks down and does not add to the understanding of students of the inherent simplicity of the tesla coil.
Next week I will probably start building a test setup. Thanks for all your input!
Tim Koene, Wed Sept 27 2006, 08:52AM
Hi everyone,
I have been doing some research on the potential transformer in question, the exact ratings were missing because the faceplatge was screwed off. It turns out that it is a 3 phase line potential transformer and it has an output rated of 36kv/sqrt(3) which is closer to 20KV, which should pose considerably less problems.
Furthermore I would like to stress the point that I am aiming for about 1-1.5 metres of streamer, so a pole pig would be a bit overkill. Maybe when this medium sized coil is done and approved they want a large >>10KW setup.
I have some nicely designed forced air single electrode sparkgaps, (Which I could put in series) or I could make a nice 6-10 piece series gap. I would like to avoid rotary design unless really necessary. It adds complexity in the building, possible service when it breaks down and does not add to the understanding of students of the inherent simplicity of the tesla coil.
Next week I will probably start building a test setup. Thanks for all your input!
Print this page