I've built several classic Tesla coils and smaller units, and even a "pulse coil" which ran on about 30kv. The photo on the top of my blog was taken from the pulse coil system. I decided to try a TV flyback before I got an NST. The NST version is on YouTube, I may also have the pulse coil on there too) channel="mostlymacros". It worked, but the output is intermittent and like all coils that use a spark gap-- really messy. My first SSTC design which I will publish here later was a success and stable, but not very powerful and I had problems keeping the single mosfet cool. I built it about 2 years ago. I have the components to do another SSTC and probably will eventually do something using a PWM and some kind of push pull mosfet drive. There are advantages that tubes have for some circuits over solid state however. So then again--if I can find some of the types I need--I might just stick with tubes. Tubes have properties that are very resilient, not to mention really good sound so they are making a comeback! Tubes or thermionic "valves" have some pretty amazing properties. I chose to use a tube because of the overload capacity and flexibility. I have made it a point to study tube design.
I started doing this when I was about 11. The first real assignment with tubes I gave myself was to transpose simple circuits like a Hartley oscillators into tube forms from solid state circuits. In that case, a center tapped coil and variable capacitor for tuning. With a little math, the tank circuit resonant frequency can be figured out. Thus I was able to begin to build radios that listened or transmitted where I wanted them too! Then I moved into experimenting with other devices. Tubes are very forgiving within reason, and if something goes wrong you far less likely to lose the tube then an expensive mosfet. You can see what's going on inside, and keep an eye on things, they also are really tough--I've seen tubes get so hot they glow entirely across the plate but when cooled back down--they still ran fine again. I even pulled a few tubes out of an old TV that had been laying in the bushes for decades in Guam, fired it up and it worked until a typhoon wreaked my place and I lost most of my stuff. Years of harsh rain and sun did nothing to damage the glass sealed device. As fragile as they may be in shipping--they are really tough compared to the mosfets and transistors that replaced them. Tolerances can be looser, substitutions simpler, and some kinds of circuits just work better and can have a lot less components if built with tubes. Tubes have a lot of gain and bandwidth--the only thing that comes close to them in solid state are mosfets and a few other emerging technologies. The military is interested in tubes since they are virtually EMP proof. Russia kept making all there radios for fighter jets out of tubes at least into the 1990s for this reason. And for sound quality there's no comparison according to many, myself included, who say tubes have the best sound in the world for audio gear.
In 2010, I shattered my myth about tube plate voltage since every book I read talks about how important B+ needs to be pretty high voltage. Never under 45volts. I was able to power a 12AX7 tube oscillator with just 15volts! That's right, it was actually possible to get a sine wave and tune a tube Hartley tank circuit that had a plate voltage of only 15 volts. The circuit was built in a cascode fashion, which means that one triode was driving modulation into the 2nd triode which was an oscillator. Plate into cathode from triode 1 to triode 2. The 1st tube lost it's bias with audio and sound quality around 30volts or so, but the CW signal and sine wave remained strong and could be turned on or off at will with a plate voltage not much more then 12 volts! I could have used it as a local oscillator for a receiver. I never thought a tube circuit could run at such low voltage. The filaments were actually running a bit lower then specs, but electron emission can happen to the plate for an oscillator at very low voltages. I plan to do some more tests with this, this is especially true with smaller high gain signal tubes. In most literature about tube design--45volts is as low as you can go. I wonder what else you can do with them--cold cathode? Some tubes do operate if you use enough voltage--such as magic eye tubes and even TV CRTs-- but space charges and other issues get in the way of grid function so they are not good for much without a hot cathode unless they were designed for a particular function like a plate relay control. You can create wideband noise out of a gas discharge tube, or use one to turn a plate relay on and off. But if you want audio frequency amplification, gain and any radio projects--you pretty much do have to keep your plate voltage (or B+ battery supply) above 45 volts and towards the tube's intended voltage and filaments/heaters must be hot enough for the electrons to begin to "boil off" so that they can be grabbed by the plate. But they are very flexible as far as design goes and just how far you can push them. I have not posted my SSTC design yet, but I will say it is the simplest one I have ever seen. I was able to build a pretty good SSTC from a 555 timer chip by feeding just enough power back so that it became a resonant circuit through the timer. A strange and unusual way to use the 555 chip that I just decided to try, and it worked well with large mosfets like the IRFP260/460. The timer is basically a PWM, and when it has almost no timing capacitance on it--it will still oscillate! I found that they can go into several hundred Kc, so getting up to the resonant frequency of a air core primary was possible. I protected the mosfet as I always do with several parts, and the timer need only 50ohm 1/4W resistor and 0.47mfd capacitor or so to drive the gate and insure that the chip never gets totally shorted if something happens to the mosfet. I did not however get nearly as much power out of it as I do here with the tube system. And the mosfet was in constant danger of overheating or being damaged. Tubes are almost immune to this kind of transient high voltage and they can overheat, but you can monitor this if your using glass tubes. This makes them ideal for power RF amplifiers or Tesla coils that would have to be far more complex and/or expensive if done solid state. The corona on my SSTC was only about 1-2inches long and the mosfet was under a lot of pressure for many reasons so I could not run it for very long without letting it cool. I really had to only run it for a few seconds at a time but it was a stable viable circuit that I built twice. The corona on my VTTC here is about 4-5 inches with sprites up to 7+ There are many great SSTC schematics online, the concept is pretty simple but coil building and winding is difficult work sometimes since there are so many possible problems with arcing and induction that must be addressed. I did not find such an abundance of good VTTC schematics, but then again I did not look for very long. I designed this circuit myself. When I did finally check out other schematics online, I found many VTTC schematics to be pretty low-powered. This may be because the volts needed at currents this high are VERY dangerous. At 1000volts 2amps or so, you really don't get 2nd chances. And to run a larger VTTC voltages quickly get into the 2000+ range at 1 amp or more. The means of doing this is so dangerous that I hesitate to post it. You don't get burned, you don't get shocked--you get electrocuted! Lights out. The general voltage for the electric chair in many states was 2000volts at 2-4amps AC. I don't think I need to say more about why this is so dangerous! MOTs are very dangerous and often underestimated, this is why microwave ovens are setup so you cannot use them with the case off. The transformer core is one side of the 2000volt secondary and they are capable of producing 2kv (2000volts) at as many as 2-5amps if the electrical outlet powering them can handle it. I may need to use a 2 phase (240volt) outlet for my next VTTC as I will need a really big voltage. So safety is very important with VTTCs. There are tube transformers out there that can produce about 1200volts or more and can be found online in different places. Lots of old radio and other gear use them. The other way is to build an inverter supply that will convert 48Vdc volts into 1200volts at 1 amp or so at higher frequency--which makes it safer if it's up in the 10kc or above range--but still very dangerous!
I have also seen people using a MOTs to ballast another MOT. A bad idea. A shorted MOT pulls so much current (more then 15amps at 120vac) that using one as a ballast will do almost nothing for keeping current down or reducing voltage and danger. It's pretty much equivalent to using a wound up 30ft extension cord as a ballast. You can still get almost as many amps through it. MOTs are high-current specialized transformers that generate a great deal of magnetic flux, so much you can feel it with any metal near by. The primary is only about 130 turns or so if I recall, #10 or 12 AWG solid--this is really big wire for a primary and produces a very large magnetic field. The secondary is #28 or larger AWG so the current you can actually get out of a MOT far exceeds the 15amp limit on most home outlets. In a microwave they push it as far as about 1300watts. That's almost 11amps. These are so powerful that if you shorted it you'll be drawing something like 30amps from your outlet and most certainly kick a breaker off or blow a fuse. If not, the arc becomes very dangerous because it's energetic and large--not something to try, seriously. To ballast a MOT for less power, use an auto-transformer or a series of 100W incandescent light bulbs in parallel and then in series with the primary. This reduces current to reasonable levels that will not blow breakers or cause dangerous flash-arcs. It's still very dangerous, but if anything happens the excess current will light up the bulbs instead of running the MOT. The same can be done with other types of large tube transformers that may also produce high currents in similar ranges. You can see the idle current in how bright they get when it's on, as well as how much your using. It also helps to put about a 6-10uf AC capacitor on the primary, this reduces the impedance of it and lowers the idle current. If a MOT is left on without an AC cap, it will get quite hot eventually even if not pulling any load. Not to mention that left idle, a MOT can pull as much as 1.5amps or so at 120vac. That is enough to run a big TV.
I recommend an insulated remote momentary only switch as I did here that cuts all power to such a HV transformer virtually instantly. With the big tube transformers--you don't get a 2nd chance. These shocks are almost as serious as touching an industrial power line. There are some really great VTTCs shown on YouTube, some very powerful transmitting tubes and coils. A lot of watts are needed no matter how you do a VTTC/SSTC. My SSTC used 90volts out of several power supplies in series. I only built a prototype and never did build a permanent version of the circuit, but for a beginner coil builder who wants something simple--it would be a great start. So I will post it here soon once I draw it up in digital form, it's only in my notes now. It is easy to power with a 5A auto-transformer. A better source then a series of transformers. Under it's requirements, the voltage dropped to only about 50volts while it was running. The load on my 4amp laptop DC supplies and others together really was hard on them as well. I just bought an auto-transformer and put a bridge on it--creating a DC power supply. One can filter this and make it really nice, you can select your voltage and use it as a kind of "brute force" supply. There will be a drop in the voltage but compensation is as simple as turning it up at bit more. Auto-transformers make great power supplies since they give out even current from 2volts or so all the way to 120 or in some models, 140volts. They also are built in a way that accidental overload or over-volt is very unlikely since the dial must be turned for the transformer to use more power, and it's a pretty big dial. So you can trust that it won't short and suddenly dump 120volts into your circuit that can only handle 40V! As is the trouble with using a load and light dimmer for such a supply. The SSTC I first built did however let me show some of the fractal effects of plasma and arcs. How one AC arc can form a sprite on it's own. I dubbed this effect "arc flares". The coil here produces a fractal of plasma reflecting complex components of the AC/RF and the DC used to power it.
A sealed jar becomes a pressurized chamber and then a partial vacuum when plasma is released into it. This is really amazing to watch, I can create all kinds of effects in the jar. Smoke gets obliterated, and gasses in the air from the smallest things cause changes to the corona.
RADIO RFI -- and PICTURES of wave function:
So the wave we see most often in the coil is the 120hz and then sweeping at 300kc (300,000hz). This is the fundamental frequency of this coil. It also has many harmonics as does any oscillator so for more powerful coils, especially if your going to use it for long periods of time you should be careful to filter the 120AC line. Especially if your going to hook it to anything like a wire. This is a very powerful radio transmitter but lacking a matched antenna, unless you live near an airport you should be fine. I often operate it late at night and check how much I am creating harmonic RFI in broadcast band areas. Also, keep all electronics away from such coils! This thing can scramble the circuits on just about any digital device from several feet away! Filters on the 120V AC side can reduce the amount of RF that gets back into the power grid without diluting the DC ripple effect. A bifiar wound toroid would also work to modulate the grid if enough power was used, thus it could be modulated either via the grid or even possibly the cathode with a good enough driver. Filters are a good idea for large coils to reduce harmonics from being transmitted as radio interference. Since we are at 300kc, we are also harmonically putting out quite a powerful signal in the AM broadcast band (530-1700kc) and down into the long wave spectrum where a lot of Navigational beckons are. So keeping leads short and not using any long wires is a good idea. Don't send any Morse code! So once again, basically I deliberately allowed 120hz/60hz into the coils plate supply. The first design used a bank of capacitors and a 60hz filter choke, I quickly found that adding a bit of line frequency 60hz was good for creating large sprites and integral frequency effects. This is complex AC theory. Components of 120hz/60hz and 300hz+ are allowed into the oscillator and thus show up in the plasma sprites (the largest arcs visible) 'painting' an image of an actual waveform. This allows you to see the wave function exactly how it really is. In 3d. I've never seen this done before. I have seen other great coils, but seeing waveforms in the arcs I never noticed. You pretty much need to take a picture of it at normal speed to be able to see it long enough. Many of these shots were done at 1/120th a sec. Most of these pics are not long exposures. Above you are seeing what it actually looks like in real life. You can see the same thing by freeze framing an HD video of it. Most cameras shoot at 24-30fps. More professional cameras can go beyond that into 60 or more FPS so that slow playback is possible. This is very useful since we can see one event in 1/60th of a second which is exactly 1 cycle of 60hz. I was amazed when I captured it for the first time. This is NOT my best picture example--more are coming of this effect including the fractal pics that show some incredible images in them with even more detail then these. I have at least 5 more to post. It can take dozens of photos to get a good wave picture.
different tubes will operate it. Some more common then others. You need a pretty big pentode power beam tube for this. The 6KD6 tube, a large TV horz. tube, was able to run this coil but got a bit too hot. Only for short times if you use that type of tube since the plate may become too hot. I'm using a larger type, the 27LE6. For looks, I moved the meter to a better spot on the board on the panel's right side. It does not directly reflect plate or grid current, measurements are really difficult since there is so much of an EM field. It is an overall circuit current meter. I know what the reading should be depending on how far up I dial the power. So if anything is wrong I can tell if it is drawing too much power or not enough. I can also get some measurement of the oscillator's current efficiency and how well matched it is. If too un-matched, the oscillator will lock up or go into a really high frequency mode that is very hard on the tube and produces no arcs. The circuit includes a tuner that matches the primary so that loads on the secondary will not cause problems with the oscillation process. This is basically the same thing as SWR with a radio, there must be a good load on this--or it will pull more power and heat up the tube's plate. If this got too out of hand the plate would literally start to glow red, but that does not happen even under a lot of load. A circuit breaker prevents more then about a 1.5amp load. As long as I match it with the delta IE control--I can connect it to anything I want such as the experiment with the jar where the plasma flows down into it. And as I said already, I importantly added a remote control switch so that I can turn the coil on/off very quickly and without fear of an RF burn or shock. Touching ground or grounded metal switches when this is on can cause RF burns if your close to the coil. So an insulated switch is important. The tuner does not effect frequency much, the air variable instead controls coil load impedance matching. This makes it possible for me to connect the coil to items and experiments without oscillator problems. An important idea for my design since I wanted to do many experiments with this. I'm sure other people have come up with the same concepts, but this schematic and coil design is my work. The schematic is available on request only, since this is a very dangerous project and I want more people who view my site and write comments! Please subscribe and more projects will be posted. Questions and such are always welcome, give me time to respond since I don't check my email all the time.
60hz/300khz or 120hz/300Khz? --both--More about the ripple effect
As I explained above already, since I used a diode bridge in the plate B+ DC power supply design, we are getting a 120hz ripple in the DC output. Basically, when you power a diode bridge with 60hz AC, you will end up getting AC components out of it if you have little or no capacitance on it or you draw so much current that the capacitance you do have cannot compensate. Since a diode bridge will allow something DC powered to operate with current in either direction, when you change directions 60 times a second, it becomes a pulsed source of DC that is 2xF(ac) 120vac 60hz. I can't write much of the math here, I will just explain as best I can. So in addition to RF AC harmonics from the oscillator, we have a 120hz signal--2x60 here. This is because each cycle either way has a phase change interval. It is difficult for me to explain in detail here, but this translates to a 120hz signal in the DC if your power supply is full wave. Most are, two diodes or a diode bridge will produce this effect. I am just learning the complexities of AC theory and AC mathematics here myself. So I hope I am correct in this theory. It is also based on multiple scope tests. If you are powering something that uses lots of current but as I did here with only a small capacitor (in this case a total of 4.7uf) you will end up getting components of both 120hz and 60hz in the output. Mainly the diode bridge produces 120hz if little or no capacitor is used of current drain is so high the capacitor can't keep up. Thus, the frequency is multiplied once the diode bridge is under load. We can even see this with a frequency counter if placed near the bridge or connected to it. Since we are using really high voltage here--you want to be sure your meter can handle it. I have a CatIII which means it's rated for 1000 volts. Any more could damage the meter, but since the voltage is so high--we can keep the coil and tube off with just the transformer and supply idling and pickup the signal of both 60hz and 120hz from the supply. 120hz will be prominent on the + side of the bridge. Basically, we have a couple of sign waves that overlap to produce the waveform pics. There is also a question of timing. How fast does the effect get created? The plasma stays in the air after the coil is off but only for about 1/120th of a second or so.. I deliberately kept capacitors low and used no DC filtration for these frequencies so that the supply would make a 'messy' output. This produces really big sprites when ever the two fundamental sine waves converge with 120hz. I would say it's 120/300000 or basically, a 2.5kc (2500hz) difference exists between the coil's fundamental operating frequency and 120Hz. The AC main 60hz frequency will also have something to do with these incredible visible wave functions. There are many factors that come into play here, it's a complex situation but what results is a fractal series of sign waves and plasma effects from the output tip, which is an AC load via capacitance. When photographed at high speed with my full frame camera, it is possible to see just how amazing these effects are and even measure them since I can set my exposure time exactly. Many more and better photos will be posted soon.