Basic SGTC questions
Ash Small, Thu Jul 31 2014, 04:45PMOk, I've been doing a bit of reading up on TC's in general, and SGTC's in particular, and my first question is which of these basic primary tank circuits is better and why?
Most people, I think, say the second has advantages, although on Ritchie's site he tends to depict the first, which is also the layout I used when I built an HF TIG welder some years ago.
I'm after in-depth answers, please.
Also, are there any differences when driven from AC or DC supplies, or when using static or rotary gaps, etc?
(I've left out details of power supplies/diodes/chokes/resistors, etc. for now, although would either of the circuits above have advantages/disadvantages as far as the various alternatives regarding supply circuitry is concerned?)
Re: Basic SGTC questions
Dr. Dark Current, Thu Jul 31 2014, 06:13PM
The first one stresses the transformer with RF AC voltage much more.
Dr. Dark Current, Thu Jul 31 2014, 06:13PM
The first one stresses the transformer with RF AC voltage much more.
Re: Basic SGTC questions
Ash Small, Thu Jul 31 2014, 06:44PM
Ok, thanks. Presumably a choke, or even a resistor, in the supply line would eradicate, or certainly reduce, this?
Presumably the second one also stresses the supply with RF AC, but to a lesser extent?
Ash Small, Thu Jul 31 2014, 06:44PM
Dr. Dark Current wrote ...
The first one stresses the transformer with RF AC voltage much more.
The first one stresses the transformer with RF AC voltage much more.
Ok, thanks. Presumably a choke, or even a resistor, in the supply line would eradicate, or certainly reduce, this?
Presumably the second one also stresses the supply with RF AC, but to a lesser extent?
Re: Basic SGTC questions
dexter, Thu Jul 31 2014, 07:06PM
from what i read and from my attempts to build a SGTC: DC has the advantage of allowing voltage doublers but it will not work with a static gap or you can use a static gap and blow alot air through it (unreliable) or better use a rotary gap. Another advantage of rotary gap in DC operation is that it doesn't need any synchronization (DC motors can be used) and you can adjust the speed of the motor aka the BPS to whatever value you want
dexter, Thu Jul 31 2014, 07:06PM
Ash Small wrote ...
Also, are there any differences when driven from AC or DC supplies, or when using static or rotary gaps, etc?
Also, are there any differences when driven from AC or DC supplies, or when using static or rotary gaps, etc?
from what i read and from my attempts to build a SGTC: DC has the advantage of allowing voltage doublers but it will not work with a static gap or you can use a static gap and blow alot air through it (unreliable) or better use a rotary gap. Another advantage of rotary gap in DC operation is that it doesn't need any synchronization (DC motors can be used) and you can adjust the speed of the motor aka the BPS to whatever value you want
Re: Basic SGTC questions
Graham Armitage, Thu Jul 31 2014, 09:09PM
Agreed. I have fried several NSTs with this configuration too. Would use option 2 as a preference. Even still I prefer to have some RF chokes and safety gap across the transformer.
Graham Armitage, Thu Jul 31 2014, 09:09PM
Dr. Dark Current wrote ...
The first one stresses the transformer with RF AC voltage much more.
The first one stresses the transformer with RF AC voltage much more.
Agreed. I have fried several NSTs with this configuration too. Would use option 2 as a preference. Even still I prefer to have some RF chokes and safety gap across the transformer.
Re: Basic SGTC questions
Steve Conner, Thu Jul 31 2014, 09:16PM
The first topology sends sustained RF oscillations from the tank circuit back into the power supply. The second one sends a single abrupt voltage step as the gap fires and the voltage across it falls to zero.
You could argue for a while over which of these is more likely to damage the power supply. I think in practice the RF oscillations work out worse. In particular, the oscillations will quickly destroy a DC supply by reverse biasing the diodes and sending part of the tank current through them.
A good filter like the "Terry Filter" would probably protect the power supply in either case, but I think the RF oscillations would waste more energy in the filter than the spark gap firing transient.
Steve Conner, Thu Jul 31 2014, 09:16PM
The first topology sends sustained RF oscillations from the tank circuit back into the power supply. The second one sends a single abrupt voltage step as the gap fires and the voltage across it falls to zero.
You could argue for a while over which of these is more likely to damage the power supply. I think in practice the RF oscillations work out worse. In particular, the oscillations will quickly destroy a DC supply by reverse biasing the diodes and sending part of the tank current through them.
A good filter like the "Terry Filter" would probably protect the power supply in either case, but I think the RF oscillations would waste more energy in the filter than the spark gap firing transient.
Re: Basic SGTC questions
Ash Small, Fri Aug 01 2014, 10:11AM
Thanks for the detailed reply, Steve.
Of the various power supply circuits I've read about, DC resonant charging seems the most appealing (link to Ritchie's page on DC resonant charging here:
).
I assume the inductor in the DC resonant charging circuit will choke the RF (and spark gap spike, for that matter) and give adequate protection for diodes, etc?
(In my HF TIG welder I just used a high(ish) value resistor to reduce any RF that tried to get back to the DC LOPT (flyback) that powered the primary circuit)
I'm currently considering using an MOT with doubler lifted straight from a microwave oven, charging a series string of electrolytics, as the supply for the DC resonant charging circuit, maybe even with more multiplier stages, if practicable, as this seems a cheap, efficient and simple solution (I have several old MOT's, etc. in the shed).
Does this sound like a practical and viable solution?
Ash Small, Fri Aug 01 2014, 10:11AM
Steve Conner wrote ...
The first topology sends sustained RF oscillations from the tank circuit back into the power supply. The second one sends a single abrupt voltage step as the gap fires and the voltage across it falls to zero.
You could argue for a while over which of these is more likely to damage the power supply. I think in practice the RF oscillations work out worse. In particular, the oscillations will quickly destroy a DC supply by reverse biasing the diodes and sending part of the tank current through them.
A good filter like the "Terry Filter" would probably protect the power supply in either case, but I think the RF oscillations would waste more energy in the filter than the spark gap firing transient.
The first topology sends sustained RF oscillations from the tank circuit back into the power supply. The second one sends a single abrupt voltage step as the gap fires and the voltage across it falls to zero.
You could argue for a while over which of these is more likely to damage the power supply. I think in practice the RF oscillations work out worse. In particular, the oscillations will quickly destroy a DC supply by reverse biasing the diodes and sending part of the tank current through them.
A good filter like the "Terry Filter" would probably protect the power supply in either case, but I think the RF oscillations would waste more energy in the filter than the spark gap firing transient.
Thanks for the detailed reply, Steve.
Of the various power supply circuits I've read about, DC resonant charging seems the most appealing (link to Ritchie's page on DC resonant charging here:
).I assume the inductor in the DC resonant charging circuit will choke the RF (and spark gap spike, for that matter) and give adequate protection for diodes, etc?
(In my HF TIG welder I just used a high(ish) value resistor to reduce any RF that tried to get back to the DC LOPT (flyback) that powered the primary circuit)
I'm currently considering using an MOT with doubler lifted straight from a microwave oven, charging a series string of electrolytics, as the supply for the DC resonant charging circuit, maybe even with more multiplier stages, if practicable, as this seems a cheap, efficient and simple solution (I have several old MOT's, etc. in the shed).
Does this sound like a practical and viable solution?
Re: Basic SGTC questions
Steve Conner, Fri Aug 01 2014, 10:33AM
DC resonant charging worked for me!
Steve Conner, Fri Aug 01 2014, 10:33AM
DC resonant charging worked for me!
Re: Basic SGTC questions
Ash Small, Fri Aug 01 2014, 08:17PM
That's pretty much what I had in mind, Steve.
I assume the reason you used four chokes was to reduce the voltage across each one to ~2.5kV, but is an inductance of that order really necessary?
(By the way, Maplin no longer list that part number, but I'm considering winding some myself, maybe with a lower Ohmic resistance. I'll do a bit more research first, though. I've wound a number of inductors over the past year or so although I've been a bit too preoccupied to do much for the past few months)
Ash Small, Fri Aug 01 2014, 08:17PM
Steve Conner wrote ...
DC resonant charging worked for me!
DC resonant charging worked for me!
That's pretty much what I had in mind, Steve.
I assume the reason you used four chokes was to reduce the voltage across each one to ~2.5kV, but is an inductance of that order really necessary?
(By the way, Maplin no longer list that part number, but I'm considering winding some myself, maybe with a lower Ohmic resistance. I'll do a bit more research first, though. I've wound a number of inductors over the past year or so although I've been a bit too preoccupied to do much for the past few months)
Re: Basic SGTC questions
Hydron, Sat Aug 02 2014, 12:12AM
If you need more voltage, you can put two MOTs in series without any extra insulation stress, just connect both transfomer cores to ground, and with the parallel primaries phased correctly you'll get ~4kV between the output QC tabs, balanced around ground.
This arrangement (with a DC resonant charging circuit) was the basic plan for the coil I never got around to building about a decade ago, so it sounds good to me.
Should be reasonably easy to get a quick simulation of the TC (or at least the power supply/tank circuit) going in LT Spice to sanity check everything and get a handle on how much inductance is required. Modelling the spark gap and streamer load will involve a large amount of guessing but you'll at least be able to get a good idea of how everything should work.
Hydron, Sat Aug 02 2014, 12:12AM
If you need more voltage, you can put two MOTs in series without any extra insulation stress, just connect both transfomer cores to ground, and with the parallel primaries phased correctly you'll get ~4kV between the output QC tabs, balanced around ground.
This arrangement (with a DC resonant charging circuit) was the basic plan for the coil I never got around to building about a decade ago, so it sounds good to me.
Should be reasonably easy to get a quick simulation of the TC (or at least the power supply/tank circuit) going in LT Spice to sanity check everything and get a handle on how much inductance is required. Modelling the spark gap and streamer load will involve a large amount of guessing but you'll at least be able to get a good idea of how everything should work.
Re: Basic SGTC questions
Steve Conner, Sat Aug 02 2014, 09:27AM
Yes, I used several chokes in series just to reduce the voltage stress on each one. The chokes all survived the ordeal, 2 of them ended up in a valve hi-fi amp.
The impedance of the charging circuit is rather high due to the high voltage and low current, so you need more inductance for the charging choke than you think, but it's not really too critical. IIRC the resonant frequency between the charging choke and tank capacitor should be about one-half of the break rate you intend to run at.
If the inductance is too small, you'll get power arcing in the rotary gap because the capacitor will charge up before the electrodes have separated enough to hold off the voltage. I had some trouble with this, but only at lower gap RPM than I normally used. The effect was simply to blow the mains fuse.
As Hydron points out, it is probably more economical to use 2 MOTs driving a voltage doubler, than a single MOT driving a quadrupler as I did, but this was really more of a fun exercise to see how much power could be got from a single MOT.
The Tesla-2 was a great coil, it made some pretty big sparks and sounded like an air-raid siren. I only gave up on it because the rotary gap self-destructed (the motor bearings were shot and the electrodes melted) and the secondary got badly burnt.
Steve Conner, Sat Aug 02 2014, 09:27AM
Yes, I used several chokes in series just to reduce the voltage stress on each one. The chokes all survived the ordeal, 2 of them ended up in a valve hi-fi amp.
The impedance of the charging circuit is rather high due to the high voltage and low current, so you need more inductance for the charging choke than you think, but it's not really too critical. IIRC the resonant frequency between the charging choke and tank capacitor should be about one-half of the break rate you intend to run at.
If the inductance is too small, you'll get power arcing in the rotary gap because the capacitor will charge up before the electrodes have separated enough to hold off the voltage. I had some trouble with this, but only at lower gap RPM than I normally used. The effect was simply to blow the mains fuse.
As Hydron points out, it is probably more economical to use 2 MOTs driving a voltage doubler, than a single MOT driving a quadrupler as I did, but this was really more of a fun exercise to see how much power could be got from a single MOT.
The Tesla-2 was a great coil, it made some pretty big sparks and sounded like an air-raid siren. I only gave up on it because the rotary gap self-destructed (the motor bearings were shot and the electrodes melted) and the secondary got badly burnt.
Re: Basic SGTC questions
Ash Small, Sat Aug 02 2014, 01:58PM
Ok, I was going to just use one MOT for the same reason as you, Steve, but I think I'll go with the concensus and use two.
Presumably this means I can use the original microwave oven doubler circuits? Presumably I can wire them up for +/-5kV approx.?
I assume max break rate is a function of current/capacitance of tank capacitor?
(I think resonant frequency/2 < max break rate, because you only need one half cycle to charge the tank capacitor, although I did read something on Ritchie's site that implied that it's possible to run a faster break rate than this and still obtain full operating voltage, but I'd need to re-read it to fully understand all the implications)
I'm wondering if the reason the bearings failed in your RSG was due to EDM (electro-discharge machining), possibly indirectly due to induced currents?...but this is speculation. I'm considering maybe driving the RSG via a belt drive (I'd still need to mount it on a bearing), and using a variac to control RPM.
I only have one 2.5A variac, so this would mean using another method to reduce input power if I want to test the TC at reduced power.
Ash Small, Sat Aug 02 2014, 01:58PM
Ok, I was going to just use one MOT for the same reason as you, Steve, but I think I'll go with the concensus and use two.
Presumably this means I can use the original microwave oven doubler circuits? Presumably I can wire them up for +/-5kV approx.?
I assume max break rate is a function of current/capacitance of tank capacitor?
(I think resonant frequency/2 < max break rate, because you only need one half cycle to charge the tank capacitor, although I did read something on Ritchie's site that implied that it's possible to run a faster break rate than this and still obtain full operating voltage, but I'd need to re-read it to fully understand all the implications)
I'm wondering if the reason the bearings failed in your RSG was due to EDM (electro-discharge machining), possibly indirectly due to induced currents?...but this is speculation. I'm considering maybe driving the RSG via a belt drive (I'd still need to mount it on a bearing), and using a variac to control RPM.
I only have one 2.5A variac, so this would mean using another method to reduce input power if I want to test the TC at reduced power.
Re: Basic SGTC questions
Hydron, Sun Aug 03 2014, 06:54AM
The standard microwave doubler circuit is no good for tesla coils anyway (has extreme ripple on the output). Check out this wikipedia page: , there are a few methods there (including the circuit Steve used for his quadrupler).
The "bridge" circuit is the standard (and most economical) option for voltage doubling without large ripple, but has the disadvantage in this situation of not having the output nicely centered around ground (it will float up and down by ~2kV AC). This is not ideal but may be acceptable given the simplicity and lowest parts count.
Your two options if you want a supply centered around ground are to use one MOT in the same configuration as Steve, or two MOTs, each with a "Greinacher" circuit (one of each polarity). This would look rather similar to Steve's circuit and require the same components, with the advantage being the greater potential power from two transformers rather than one.
If you had to go with Steve's circuit or the Greinacher circuit then you probably could re-use the microwave oven cap for the first stage (doing exactly the same job as it did in the oven), but the microwave oven diodes may not be able to handle the peak currents, and second stage "peak detect" portion would have to use larger capacitors (many series electrolytics with conservative rating and careful balancing, or a number of series/parallel matched microwave caps) and diodes (same considerations as with the electrolytic caps). A "bridge" circuit would require the same parts as the "peak detect" portion of the other options. Those with greater experience may be able to weigh in on the question of whether the microwave oven diodes are re-usable.
Hydron, Sun Aug 03 2014, 06:54AM
The standard microwave doubler circuit is no good for tesla coils anyway (has extreme ripple on the output). Check out this wikipedia page:
, there are a few methods there (including the circuit Steve used for his quadrupler).The "bridge" circuit is the standard (and most economical) option for voltage doubling without large ripple, but has the disadvantage in this situation of not having the output nicely centered around ground (it will float up and down by ~2kV AC). This is not ideal but may be acceptable given the simplicity and lowest parts count.
Your two options if you want a supply centered around ground are to use one MOT in the same configuration as Steve, or two MOTs, each with a "Greinacher" circuit (one of each polarity). This would look rather similar to Steve's circuit and require the same components, with the advantage being the greater potential power from two transformers rather than one.
If you had to go with Steve's circuit or the Greinacher circuit then you probably could re-use the microwave oven cap for the first stage (doing exactly the same job as it did in the oven), but the microwave oven diodes may not be able to handle the peak currents, and second stage "peak detect" portion would have to use larger capacitors (many series electrolytics with conservative rating and careful balancing, or a number of series/parallel matched microwave caps) and diodes (same considerations as with the electrolytic caps). A "bridge" circuit would require the same parts as the "peak detect" portion of the other options. Those with greater experience may be able to weigh in on the question of whether the microwave oven diodes are re-usable.
Re: Basic SGTC questions
Steve Conner, Sun Aug 03 2014, 09:08AM
Re the lack of a variac: I used a universal motor for my RSG, and a dimmer switch rated for inductive loads (low voltage halogen lights etc) to control the speed of it. This freed up the variac to control the input to the MOT.
Once I was confident with it, I ditched the variac and used a relay from a microwave oven, controlled by a big red button, to switch the MOT directly onto the mains.
Ash, I'm pretty sure I still have the voltage multiplier and de-Qing diode stack from Tesla-2 kicking around somewhere if you would like them. It should be easy to modify the multiplier to take input from two MOTs.
Steve Conner, Sun Aug 03 2014, 09:08AM
Re the lack of a variac: I used a universal motor for my RSG, and a dimmer switch rated for inductive loads (low voltage halogen lights etc) to control the speed of it. This freed up the variac to control the input to the MOT.
Once I was confident with it, I ditched the variac and used a relay from a microwave oven, controlled by a big red button, to switch the MOT directly onto the mains.
Ash, I'm pretty sure I still have the voltage multiplier and de-Qing diode stack from Tesla-2 kicking around somewhere if you would like them. It should be easy to modify the multiplier to take input from two MOTs.
Re: Basic SGTC questions
Hydron, Sun Aug 03 2014, 11:03AM
I'd take Steve up on his kind offer!
Should be extremely easy to use the remains of Tesla-2 - just split the connection between the first two capacitors, feed one from each MOT, the center connection remains as ground (connect to both cores). Or you can just use one MOT as Steve did, and graduate to two if you need more power.
The beauty of a DC coil is that theoretically a ballast and variac are unnecessary - the choke stops the gap from power arcing and the power level can be set by the break rate rather than the input voltage.
All that said, do what Steve suggested - you want a variac for initial powerup, and even your 2.5A one should be enough for limited runs at lower break rates, with the option to ditch it and connect straight to the mains once it's running well.
Hydron, Sun Aug 03 2014, 11:03AM
I'd take Steve up on his kind offer!
Should be extremely easy to use the remains of Tesla-2 - just split the connection between the first two capacitors, feed one from each MOT, the center connection remains as ground (connect to both cores). Or you can just use one MOT as Steve did, and graduate to two if you need more power.
The beauty of a DC coil is that theoretically a ballast and variac are unnecessary - the choke stops the gap from power arcing and the power level can be set by the break rate rather than the input voltage.
All that said, do what Steve suggested - you want a variac for initial powerup, and even your 2.5A one should be enough for limited runs at lower break rates, with the option to ditch it and connect straight to the mains once it's running well.
Re: Basic SGTC questions
Ash Small, Sun Aug 03 2014, 02:18PM
Yep, I was studying that page again last night, before I came up with this:
Which I posted in a thread in the HV section (thread here: ). I posted it in a separate thread in case it was a non-starter, but the consensus is that it should work pretty much 'as is'. (I am planning to use a string of electrolytics for the output stage, if I try this circuit)
I'll PM you Steve, thanks. I want to try using the microwave oven parts, and see what happens, but I imagine modifying your old quadrupler will be an improvement. Will the de-Qing diode be up to the job, though? I think I'll have (potentially) ~twice the current your Tesla 2 had, obviously depending on tank capacitor and break rate, etc.
The next job will be tackling the maths regarding tank capacitor, Fres, break rate, secondary and topload, etc., and working out how much inductance I need for DC resonant charging. All this will require a bit more research......
Ash Small, Sun Aug 03 2014, 02:18PM
Hydron wrote ...
The standard microwave doubler circuit is no good for tesla coils anyway (has extreme ripple on the output). Check out this wikipedia page: , there are a few methods there (including the circuit Steve used for his quadrupler).
The standard microwave doubler circuit is no good for tesla coils anyway (has extreme ripple on the output). Check out this wikipedia page:
, there are a few methods there (including the circuit Steve used for his quadrupler).Yep, I was studying that page again last night, before I came up with this:
Which I posted in a thread in the HV section (thread here:
). I posted it in a separate thread in case it was a non-starter, but the consensus is that it should work pretty much 'as is'. (I am planning to use a string of electrolytics for the output stage, if I try this circuit)Steve Conner wrote ...
Ash, I'm pretty sure I still have the voltage multiplier and de-Qing diode stack from Tesla-2 kicking around somewhere if you would like them. It should be easy to modify the multiplier to take input from two MOTs.
Ash, I'm pretty sure I still have the voltage multiplier and de-Qing diode stack from Tesla-2 kicking around somewhere if you would like them. It should be easy to modify the multiplier to take input from two MOTs.
I'll PM you Steve, thanks. I want to try using the microwave oven parts, and see what happens, but I imagine modifying your old quadrupler will be an improvement. Will the de-Qing diode be up to the job, though? I think I'll have (potentially) ~twice the current your Tesla 2 had, obviously depending on tank capacitor and break rate, etc.
The next job will be tackling the maths regarding tank capacitor, Fres, break rate, secondary and topload, etc., and working out how much inductance I need for DC resonant charging. All this will require a bit more research......
Re: Basic SGTC questions
Ash Small, Sun Aug 03 2014, 11:44PM
Continuing on from my last post, Intuition suggests I go for a big toroid, maybe one metre diameter by 30cm high. If this turns out to be too big for breakout to occur, I then have two choices, either introduce a 'breakout point', or increase the capacitance of the primary tank capacitor, as, from what I've read, it is the ratio of these two capacitances that determines the voltage in the secondary tank circuit.
From what I've read, secondary coil height should equal toriod diameter, so secondary height then becomes one metre.
I want to keep copper losses to a minimum, so go for 1mm diameter wire, enamelled, I'll do some maths later on regarding skin depth, once frequency has been established.
I then need to establish diameter of secondary. This will determines Fres.
(These are just some thoughts I've had sitting here enjoying a bottle of wine on a Sunday evening. Any comments on the above 'thoughts' will be welcomed. I'll post again in the morning.)
EDIT: Maybe 30cm is a bit big, 20cm (8") or so may be a bit more aesthetic, although I seem to keep reading that people eiter try a bigger toroid and get better results, or add a second toroid and get better results. I assume most people use one of the simulations or whatever to decide toroid size, but then get better results when they add more capacitance. I've also not found much in the way of deciding toroid size on any of the sites I've tried. Ritchie's doesn't say much. Steve Connor's pretty much just says 'it worked better when we tried a bigger toroid', Terry's site is down, etc., although I have read that if the toroid is too big and no breakout occurs, 'add a breakout point, or add more primary tank capacitance.
I'll probably try one of the simulations, though, once I've got some initial numbers to input.
Ash Small, Sun Aug 03 2014, 11:44PM
Continuing on from my last post, Intuition suggests I go for a big toroid, maybe one metre diameter by 30cm high. If this turns out to be too big for breakout to occur, I then have two choices, either introduce a 'breakout point', or increase the capacitance of the primary tank capacitor, as, from what I've read, it is the ratio of these two capacitances that determines the voltage in the secondary tank circuit.
From what I've read, secondary coil height should equal toriod diameter, so secondary height then becomes one metre.
I want to keep copper losses to a minimum, so go for 1mm diameter wire, enamelled, I'll do some maths later on regarding skin depth, once frequency has been established.
I then need to establish diameter of secondary. This will determines Fres.
(These are just some thoughts I've had sitting here enjoying a bottle of wine on a Sunday evening. Any comments on the above 'thoughts' will be welcomed. I'll post again in the morning.)
EDIT: Maybe 30cm is a bit big, 20cm (8") or so may be a bit more aesthetic, although I seem to keep reading that people eiter try a bigger toroid and get better results, or add a second toroid and get better results. I assume most people use one of the simulations or whatever to decide toroid size, but then get better results when they add more capacitance. I've also not found much in the way of deciding toroid size on any of the sites I've tried. Ritchie's doesn't say much. Steve Connor's pretty much just says 'it worked better when we tried a bigger toroid', Terry's site is down, etc., although I have read that if the toroid is too big and no breakout occurs, 'add a breakout point, or add more primary tank capacitance.
I'll probably try one of the simulations, though, once I've got some initial numbers to input.
Re: Basic SGTC questions
Hydron, Mon Aug 04 2014, 10:43AM
For reference, my DRSSTC spits out ~2m streamers/arcs with a 160x700mm (dia.xheight) secondary and a 160x750mm (heightxdia.) toroid, and could probably be pushed a bit further if I play with the current limit.
You should be able to get at least close to that streamer length with a spark gap and a similar sized coil, but I suspect it would be pushing a two MOT supply to try a larger coil.
If anyone else has comments on this, please reply, I've never built a spark gap coil, though I did do all the research into it about a decade ago.
Hydron, Mon Aug 04 2014, 10:43AM
For reference, my DRSSTC spits out ~2m streamers/arcs with a 160x700mm (dia.xheight) secondary and a 160x750mm (heightxdia.) toroid, and could probably be pushed a bit further if I play with the current limit.
You should be able to get at least close to that streamer length with a spark gap and a similar sized coil, but I suspect it would be pushing a two MOT supply to try a larger coil.
If anyone else has comments on this, please reply, I've never built a spark gap coil, though I did do all the research into it about a decade ago.
Re: Basic SGTC questions
Steve Conner, Mon Aug 04 2014, 01:43PM
This is the biggest spark I got out of Tesla-2. For reference the secondary is 20" tall. I used almost 3x the bang energy the coil was designed for, and ended up burning out both the spark gap electrodes and the secondary. The single MOT power supply was probably delivering something like 2.5kW.
Re toroid size: the classical guideline is that the minor diameter of the toroid should be equal to the diameter of the secondary, and the major diameter equal to the height of the secondary. You won't go far wrong following this. Most people (myself included ) allow themselves to be influenced by whatever parts they can find to make a topload.
Steve Conner, Mon Aug 04 2014, 01:43PM
This is the biggest spark I got out of Tesla-2.
For reference the secondary is 20" tall. I used almost 3x the bang energy the coil was designed for, and ended up burning out both the spark gap electrodes and the secondary. The single MOT power supply was probably delivering something like 2.5kW.Re toroid size: the classical guideline is that the minor diameter of the toroid should be equal to the diameter of the secondary, and the major diameter equal to the height of the secondary. You won't go far wrong following this. Most people (myself included
) allow themselves to be influenced by whatever parts they can find to make a topload.Re: Basic SGTC questions
Ash Small, Mon Aug 04 2014, 01:50PM
I'd also be interested in comments from others regarding this subject.
The dimensions of your coil/topload sound about the minimum I'd want to try. (Intuition tells me I want a big toroid and a secondary with low copper losses)
I assume decreasing the break rate would compensate to some extent the effects of a large primary capacitor, although another option would be to just add more MOT's and doublers. Increasing primary tank cap also means I can use a smaller choke in the resonant charging circuit..
Any further comments will be appreciated.
EDIT: Missed your post as I was writing my reply, Steve. Thanks for the input. That photo is what I was referring to when I said you got better results by increasing the size of the topload
Ash Small, Mon Aug 04 2014, 01:50PM
Hydron wrote ...
For reference, my DRSSTC spits out ~2m streamers/arcs with a 160x700mm (dia.xheight) secondary and a 160x750mm (heightxdia.) toroid, and could probably be pushed a bit further if I play with the current limit.
You should be able to get at least close to that streamer length with a spark gap and a similar sized coil, but I suspect it would be pushing a two MOT supply to try a larger coil.
If anyone else has comments on this, please reply, I've never built a spark gap coil, though I did do all the research into it about a decade ago.
For reference, my DRSSTC spits out ~2m streamers/arcs with a 160x700mm (dia.xheight) secondary and a 160x750mm (heightxdia.) toroid, and could probably be pushed a bit further if I play with the current limit.
You should be able to get at least close to that streamer length with a spark gap and a similar sized coil, but I suspect it would be pushing a two MOT supply to try a larger coil.
If anyone else has comments on this, please reply, I've never built a spark gap coil, though I did do all the research into it about a decade ago.
I'd also be interested in comments from others regarding this subject.
The dimensions of your coil/topload sound about the minimum I'd want to try. (Intuition tells me I want a big toroid and a secondary with low copper losses)
I assume decreasing the break rate would compensate to some extent the effects of a large primary capacitor, although another option would be to just add more MOT's and doublers. Increasing primary tank cap also means I can use a smaller choke in the resonant charging circuit..
Any further comments will be appreciated.
EDIT: Missed your post as I was writing my reply, Steve. Thanks for the input. That photo is what I was referring to when I said you got better results by increasing the size of the topload
Re: Basic SGTC questions
Ash Small, Mon Aug 04 2014, 07:25PM
Well, I found this supplier of flexible aluminium ducting. They stock several sizes, including 200mm and 300mm. I've not looked around for cheaper suppliers.
I won't be ordering any just yet, but this does seem to be the stuff to use to get a basic toroid shape.
EDIT: While the price, ex vat, of 3m x 200mm is just over £20, with delivery and vat this increases to just under £40, so maybe I'll keep looking.....
EDIT: Found a supplier on Ebay, almost half the price including postage.
Ash Small, Mon Aug 04 2014, 07:25PM
Well, I found this supplier of flexible aluminium ducting. They stock several sizes, including 200mm and 300mm. I've not looked around for cheaper suppliers.
I won't be ordering any just yet, but this does seem to be the stuff to use to get a basic toroid shape.
EDIT: While the price, ex vat, of 3m x 200mm is just over £20, with delivery and vat this increases to just under £40, so maybe I'll keep looking.....
EDIT: Found a supplier on Ebay, almost half the price including postage.
Re: Basic SGTC questions
loneoceans, Mon Aug 04 2014, 08:59PM
I don't have very much to add seeing how you're going for a resonant charging system. My old Tesla Coil 2 which I built 10 years ago (has it been that long?) simply used 4 MOTs in series submerged in oil (8.5kVAC out) and a rotary gap. I recently fired it up earlier this year and it's still happily making 1.8m of spark or so, mostly limited by ceiling and old flakey ducted toroid at the moment.
and
Total apparent power from the wall as just about 4kVA. It used a 55x11cm secondary with a 51x12.7cm PVC ducted toroid wrapped with aluminium tape, with a flat primary and runs at just around 200kHz. Primary cap is 53nF. Right now my RSG is running at about 300bps so I'm sure there is a little bit more play and tuning I could do to optimize performance (e.g. maybe a sync rsg). Perhaps these dimensions may be useful for your project.
loneoceans, Mon Aug 04 2014, 08:59PM
I don't have very much to add seeing how you're going for a resonant charging system. My old Tesla Coil 2 which I built 10 years ago (has it been that long?) simply used 4 MOTs in series submerged in oil (8.5kVAC out) and a rotary gap. I recently fired it up earlier this year and it's still happily making 1.8m of spark or so, mostly limited by ceiling and old flakey ducted toroid at the moment.
and 
Total apparent power from the wall as just about 4kVA. It used a 55x11cm secondary with a 51x12.7cm PVC ducted toroid wrapped with aluminium tape, with a flat primary and runs at just around 200kHz. Primary cap is 53nF. Right now my RSG is running at about 300bps so I'm sure there is a little bit more play and tuning I could do to optimize performance (e.g. maybe a sync rsg). Perhaps these dimensions may be useful for your project.
Print this page