Spark Gap Tesla Coil
Spark Gap Tesla Coil
Embedded Files
While constructing my DRSSTC, I got sidetracked and built a spark gap Tesla coil. It was a fun side project and turned out to be pretty impressive.
While constructing my DRSSTC, I got sidetracked and built a spark gap Tesla coil. It was a fun side project and turned out to be pretty impressive.
An idea came to me about Tesla coils. I knew that for ones utilizing spark gaps, a large step-up transformer would be needed to generate a spark. Most people go with neon sign transformers, but they are expensive and fragile. I had an idea about using a microwave oven transformer (MOT) instead, and upon doing a bit of research, I found this schematic designed by Greg Hunter. These transformers are easy to find; most scrapped microwaves have functional high voltage capacitors and transformers.
An idea came to me about Tesla coils. I knew that for ones utilizing spark gaps, a large step-up transformer would be needed to generate a spark. Most people go with neon sign transformers, but they are expensive and fragile. I had an idea about using a microwave oven transformer (MOT) instead, and upon doing a bit of research, I found this schematic designed by Greg Hunter. These transformers are easy to find; most scrapped microwaves have functional high voltage capacitors and transformers.
I started scavenging for used microwaves. I bought two or three for around 10 bucks each and found a few which were going to be thrown out. As soon as I got my hands on my first microwave, I couldn't help but play around with the transformer. This is a Jacob's Ladder I put together. I triggered each "ascension" with a "chicken stick," aka a very long piece of PVC with something metal on the end. MOTs can provide hundreds of milliamps at more than 2,000 volts, so I was sure to stay well away.
I started scavenging for used microwaves. I bought two or three for around 10 bucks each and found a few which were going to be thrown out. As soon as I got my hands on my first microwave, I couldn't help but play around with the transformer. This is a Jacob's Ladder I put together. I triggered each "ascension" with a "chicken stick," aka a very long piece of PVC with something metal on the end. MOTs can provide hundreds of milliamps at more than 2,000 volts, so I was sure to stay well away.
One scary thing I found out is that concrete is conductive! One end of the secondary winding on the MOT is connected to the ferrite core. Placing the transformer on concrete and touching the other end of the secondary somewhere else on the ground produced an arc! These are scarily powerful.
One scary thing I found out is that concrete is conductive! One end of the secondary winding on the MOT is connected to the ferrite core. Placing the transformer on concrete and touching the other end of the secondary somewhere else on the ground produced an arc! These are scarily powerful.
Jacob's Ladder.mp4Like my previous coil, the first step was to wind a secondary. I planned on using the same CDE 942c series capacitors in the tank circuit. Having 2 strings of 7 in series gave a rating of ~0.045uf at 14kV. That number is very similar and yet at the same time very different to the 0.45uf's in my DRSSTC. I planned to make an identical secondary to the one I previously built, assuming the resonant frequency of the primary coil would be the same as in the DRSSTC. When I finished my second (beautiful) secondary, I got to constructing the primary circuit. Once it was done, I did a quick resonance check using a signal generator and oscilloscope, and my palm instantly landed on my face. Obviously, by reducing the capacitance in the primary, the resonant frequency would be way higher at the same inductance. Using an identical coil matched to a primary circuit with an order of magnitude greater capacitance simply won't work. The secondary I had built was way off target. I used what I had leftover from the previous coil, however, so I didn't really waste anything.
Like my previous coil, the first step was to wind a secondary. I planned on using the same CDE 942c series capacitors in the tank circuit. Having 2 strings of 7 in series gave a rating of ~0.045uf at 14kV. That number is very similar and yet at the same time very different to the 0.45uf's in my DRSSTC. I planned to make an identical secondary to the one I previously built, assuming the resonant frequency of the primary coil would be the same as in the DRSSTC. When I finished my second (beautiful) secondary, I got to constructing the primary circuit. Once it was done, I did a quick resonance check using a signal generator and oscilloscope, and my palm instantly landed on my face. Obviously, by reducing the capacitance in the primary, the resonant frequency would be way higher at the same inductance. Using an identical coil matched to a primary circuit with an order of magnitude greater capacitance simply won't work. The secondary I had built was way off target. I used what I had leftover from the previous coil, however, so I didn't really waste anything.
For my next shot at it, I used JavaTC, a useful online calculator which computes different aspects of Tesla coils. I tried to design a secondary near the resonance I had measured from the primary. I settled on a 4.5" OD (4" PVC tube) x 22" pipe wound with 26 AWG wire.
For my next shot at it, I used JavaTC, a useful online calculator which computes different aspects of Tesla coils. I tried to design a secondary near the resonance I had measured from the primary. I settled on a 4.5" OD (4" PVC tube) x 22" pipe wound with 26 AWG wire.
Here, I'm winding the actual coil. The 26 AWG wire was much easier to work with than 30 AWG. This was still version 1 of the winder rig. It broke down when I was 3/4 of the way through, which is when I decided to upgrade it. Please excuse the mop of hair on my head, this video was taken multiple months into the COVID-19 shutdown and I couldn't get a haircut.
Here, I'm winding the actual coil. The 26 AWG wire was much easier to work with than 30 AWG. This was still version 1 of the winder rig. It broke down when I was 3/4 of the way through, which is when I decided to upgrade it. Please excuse the mop of hair on my head, this video was taken multiple months into the COVID-19 shutdown and I couldn't get a haircut.
Once I fixed the rig and finished the coil, I decided to coat the coil in epoxy instead of varnish. I got some from Amazon and got to work.
Once I fixed the rig and finished the coil, I decided to coat the coil in epoxy instead of varnish. I got some from Amazon and got to work.
Here it is with three individual coats. I think using epoxy was a great choice, protecting it from scratches and adding some extra insulation. The mirror-like finish is a nice touch.
Here it is with three individual coats. I think using epoxy was a great choice, protecting it from scratches and adding some extra insulation. The mirror-like finish is a nice touch.
For the toroid on my first tests, I used the spun aluminum one from my oneTesla. I later changed to one made from aluminum duct.
For the toroid on my first tests, I used the spun aluminum one from my oneTesla. I later changed to one made from aluminum duct.
This is the first test of the completed coil. While it did not work, I thought of it as a success. Arcing from the primary to secondary indicated that this coil was powerful, and I needed to design around that.
This is the first test of the completed coil. While it did not work, I thought of it as a success. Arcing from the primary to secondary indicated that this coil was powerful, and I needed to design around that.
In my next version, I had a flat primary, putting as much distance between it and the secondary as possible.
In my next version, I had a flat primary, putting as much distance between it and the secondary as possible.
SGTC first test.movSGTC flat primary testThis is the flat primary in action. It worked, but the sparks produced were pretty small and diminished after a short time.
This is the flat primary in action. It worked, but the sparks produced were pretty small and diminished after a short time.
The primary circuit I first built was a near copy of Greg's schematic above. The only thing I changed was adding a few more microwave oven capacitors (MOC) in series with the output of the MOTs. Capacitive reactance is given by 1/(2*pi*f*C). Using just 2 MOCs gives 5300 ohms at 4.2kV. Substituting I=V/R means 792mA on the output of the MOTs, requiring 27.7 amps on the input. My measly 15 amp garage circuit breaker would not stand a chance. Bumping up to 4 MOCs requires 13.9, which is better, but still too close to the breaker's limit for me. I ended up going with 5 MOCs pulling 11 amps.
The primary circuit I first built was a near copy of Greg's schematic above. The only thing I changed was adding a few more microwave oven capacitors (MOC) in series with the output of the MOTs. Capacitive reactance is given by 1/(2*pi*f*C). Using just 2 MOCs gives 5300 ohms at 4.2kV. Substituting I=V/R means 792mA on the output of the MOTs, requiring 27.7 amps on the input. My measly 15 amp garage circuit breaker would not stand a chance. Bumping up to 4 MOCs requires 13.9, which is better, but still too close to the breaker's limit for me. I ended up going with 5 MOCs pulling 11 amps.
Back to the rest of the coil. I used a stationary spark gap, and it proved to be extremely finicky. It's supposed to work by arcing at every peak of an AC cycle from the wall, so 60 times a second. It just so happens that when it does arc, the spark gap gets VERY hot. The heat from a previous spark is absorbed into the metal electrodes of the gap, making the air in-between them break down at a lower voltage on the next cycle. This continues with each spark, making each successive one pour less and less energy into the coil, as shown above.
Back to the rest of the coil. I used a stationary spark gap, and it proved to be extremely finicky. It's supposed to work by arcing at every peak of an AC cycle from the wall, so 60 times a second. It just so happens that when it does arc, the spark gap gets VERY hot. The heat from a previous spark is absorbed into the metal electrodes of the gap, making the air in-between them break down at a lower voltage on the next cycle. This continues with each spark, making each successive one pour less and less energy into the coil, as shown above.
To fix this, I decided to build an asynchronous rotary spark gap. The asynchronous part means each spark isn't synchronized with the peak of a cycle, and the rotary part means there are multiple flying electrodes.
To fix this, I decided to build an asynchronous rotary spark gap. The asynchronous part means each spark isn't synchronized with the peak of a cycle, and the rotary part means there are multiple flying electrodes.
My first design was two stationary electrodes on opposite sides of a fan from a microwave. I had one flying electrode, consisting of two brackets joined by wire on a 3D printed holder. When it fired, the arc produced lasted for much too long, so I scrapped that idea.
My first design was two stationary electrodes on opposite sides of a fan from a microwave. I had one flying electrode, consisting of two brackets joined by wire on a 3D printed holder. When it fired, the arc produced lasted for much too long, so I scrapped that idea.
For my second iteration, I 3D printed a casing shown on the right. There are two parts; the housing for the fan motor and the flywheel holding the flying electrodes. I used #6 nuts/bolts for the job.
For my second iteration, I 3D printed a casing shown on the right. There are two parts; the housing for the fan motor and the flywheel holding the flying electrodes. I used #6 nuts/bolts for the job.
This is it printed and assembled:
This is it printed and assembled:
This puppy made some pretty drastic improvements in spark length.
This puppy made some pretty drastic improvements in spark length.
I transitioned from the small spun aluminum toroid to a slightly larger aluminum duct toroid in the hopes that an increased capacitance would allow for more energy to be stored before being released in the secondary coil. It worked pretty well, and I am happy with this project so far.
I transitioned from the small spun aluminum toroid to a slightly larger aluminum duct toroid in the hopes that an increased capacitance would allow for more energy to be stored before being released in the secondary coil. It worked pretty well, and I am happy with this project so far.
Not shown: hours of tuning the primary coil.
Not shown: hours of tuning the primary coil.
This is a scope screenshot I captured while running the coil. It's pretty cool to see the physics in action.
This is a scope screenshot I captured while running the coil. It's pretty cool to see the physics in action.
The frequency at which it operates is 1/(4.80*10^-6)=208kHz, which was pretty much spot on to what JavaTC predicted.
The frequency at which it operates is 1/(4.80*10^-6)=208kHz, which was pretty much spot on to what JavaTC predicted.
Page updated
Google Sites
Report abuse