Видео

I couldn't tell you exactly how much NOx compounds were produced, but it certainly would produce some via the birkeland-eyde process. It wouldn't be in very significant amounts though, especially with such short run times, I wouldn't think. Then again, my opinions are based on how long it takes for an arc to make enough nitrogen compounds to acidify water, mostly from videos I have watched, so take that with a big grain of nitrate salts lol.

Correct me if I'm wrong, but I feel like I remember you, weren't you a member or 4hv? Either way, agreed, two peas in a pod!

That would be a cool project. I'm not sure how you could use multiple magnetrons though since they would inevitably be out of phase and slightly different frequencies from one another causing destructive interference. Maybe if the microwaves were absorbed quickly enough, and perhaps having 4 separate waveguides might work, honestly not sure about that. Since I have water cooled a magnetron I should see just how much power I can push through one. Maybe next I'll use 3 MOTs with full rectification next and continue to add transformers until the anode or cathode evaporates. I mean is it really producing enough microwaves if your magnets aren't reaching their curie temperature? :)

@@triggerhappy77707 hey, thanks for reply. I thought about that too. As well as destructive waves you could get constructive waves if you tune it properly. Hope someone chimes in on this. Actually, you do have it fully rectified would that mean all waves would be constructive?

@@gordon6029 That's true, but from everything I have seen you can't use multiple magnetrons, at least in a cavity. While the waves might be equally as constructive as they are destructive, I think, or rather know it it will overheat one or both magnetrons. It's not the input power's phase that you need to worry about, but rather the phase of the 2.45GHz microwaves. Not only the phase but the frequency will differ slightly, probably by many KHz considering the operating frequency is 2.45GHz There is, however, something known as a waveguide multiplexer that is specifically made to avoid this interference, which is more or less what I was trying to suggest in my previous response when I spoke about multiple waveguides, except I wasn't thinking about multiplexing but rather just keeping them separate, though that would create it's own challenges.

Thanks, beside the dangers it was lots of fun :p. And yes, it is an unmodified magnetron running the same way it would in a microwave (except fullwave rectified ~4kv with 2 transformers instead of a halfwave voltage doubler) with a breakout point touching the antenna, and to make the breakout start more reliably I put a tiny piece of paper towel wetted but not soaked in a saturated solution of salt water. It would eventually create plasma without it, but the magnetron would get very hot limiting the run time.

Yes, I haven't build one in a long time but I still have a gu-81m and some other tubes

@@triggerhappy77707 I was interested in your mini 811A VTTC

@@Headbutter-Lettuce90 It was the basic 811a vttc schematic from Steve Ward, I just changed some of the values of the capacitors and grid leak resistor to work well with the higher freq of the small coil. The concept is exactly the same, I just tuned it as well as possible

@@triggerhappy77707 but you made the grid leak just by trial and error?

@@Headbutter-Lettuce90 Yes, all of the components values were determined by trial and error. I tried different values of components, primary windings, feedback windings and topload until I achieved the longest sparks and the least plate dissipation possible.

That would be pretty cool, it's absolutely possible too! They are known as microwave electrothermal thrusters. If I knew more about them that would make an awesome project, if only I had the materials and a proper vacuum chamber... You're giving me ideas 🤔

@@triggerhappy77707 Wonder if you could make it work with a RC jet engine..? hmm

@@ggesdsdsdsd I think the main obstacle (other than not being in a vacuum) would be supplying 4kv at 2-3kw in something so small. Would be cool to see though

It is just an ir2153 halfbridge driver with a few resistors, capacitors, and 2 mosfets. It is practically exactly the same circuit as in the datasheet's application circuit, just with some zener diodes to protect the gates and some TVS's to clamp the voltage across the drain and source though they aren't strictly necessary.

@@triggerhappy77707 had you ever tried to make it a Tesla coil?

@@Headbutter-Lettuce90 As a matter of fact I did! I just used an antenna for feedback, cleaned up the signal with a 74hc14 hex inverter, and connected the output of the 74hc14 to the timing capacitor pin of the half bridge driver. Worked great, quite efficient thanks to the gate driving characteristics and proper dead time of the gate driver! I've also made a simple full bridge class D amplifier using 2 of them, which I have a video of that amplifier before I worked all the bugs out on my channel if interested.

I should mention that I had to use a fairly large coil with very fine wire and a topload for lower resonant frequency, the gate driver couldn't drive the gates at the ~400khz resonance of one of my smaller coils

@@triggerhappy77707 that is simply, smort

be nice

It's just gravity, the glass tube loosely holds the wire in place. You may be able to solder to the antenna directly if you remove the metal cap.

I think it is starting to get mad at these voltages and output power! The diodes are definitely dead :p

The smaller the diameter the better the coupling which usually translates to more spark length per watt of power. This circuit was actually meant to run continuous full wave for as long as I wanted, the hardware to put my proper heatsink on the FETs just hadn't come in so I figured I'd see how long they'd last on these tiny ones

Well considering I'm in the USA, I have 120v input, so 170v dc after rectification. I ended up using 4.3k ohms, mostly because that is what I had laying around but it is important to be sure the resistors won't have to dissipate more than they are rated for, in my case I used a 5w resistor. Using ohms law, I can calculate that 4.3k ohms will provide about 40milliamps and will dissipate about 7 watts. 40ma does work, but it would be best if I had 100ma to be sure the ir2153 had enough current to operate and to drive the MOSFET gates hard enough, but it works well as it is. I would suggest looking at the datasheet if you plan on making somwthing similar with a resistive dropper to provide the 12-15v to get an idea of how much current you need. Then just use an ohms law calculator to find what resistance is needed for your input voltage and current needs, as well as calculate how many watts the resistor will need to be rated for.

@@triggerhappy77707 what is the formula?

The reason they exploded is because I let the junction temperature get way too high 😅 my gate driver circuit implemented an adjustable dead time generator, and I set it to have just a little extra dead time in case the antenna feedback caused a weird parasitic oscillation (which has happened before I had the deadtime)

Oh and I did not use any snubbed across the drain/source as it wasn't needed. I did add extra decoupling capacitor on the buss to get rid of some slight ringing, but the waveforms all looked decent

Thank ya! This SSTC setup was just a test of my driver circuit, whenever I finalize it I expect at least twice the spark length. The secondary I am using for this test is so far from optimal, the aspect ratio is too high for sure. I did try a small topload just to see what it would do, and it actually decreased spark length likely due to the secondary being so wrong for this application. I am using IRFP460 MOSFETs (I found a place on eBay that sells 20 for about $15 usd shipped!) though I've also used some 300v 80 amp IGBTs which worked well of course. The main reason the fets died is because of a lack of heatsinking, but I think I need more deadtime, it seems I am getting a bit of shoot-through, causing more heat. But the driver works great, so good enough :p

@@triggerhappy77707 Hey I use IRFP460s in my SSTC too! Not an uncommon FET to use I guess. Wishing you good luck on getting double that spark length and making a new secondary coil, keep up the good work man!

I'm glad you enjoyed, too! The 'resonator' is indeed just a piece of wire with a certain length. The wavelength of these 2.45GHz magnetrons is 12.2cm/4.8in, though I'm not using the exact right size as it wouldn't matter. The metal melts so it is always changing length anyway! Close enough is good enough for this haha.

@@triggerhappy77707 i have also another question: you say you use two mot's for the full wave rectifier? Is that for the voltagedoubling? So did you use the fullwaverectifier with 4diodes and the caps only for smoothing the 50Hz(no,its of course100Hz then) ? Or did you made it with 2 diodes, so with each mot has one cap at the normal magnetronway-place(difficulties for dutch me, describing my question in english) as each cap direct in series with each mot his hv-out, and so the 2 mot's their cores and also of course the connection between the 2 diodes is grounded? Yes i think that's most logically, so the housing of the magnetron is at groundlevel what is needed even more with watercooling i think. But still you use some extra caps for smoothing the 4kV? Wow yes, that's an interesting set-up to say the least!!

Ow i can think of hundred things i want to ask you about this built of tours. But let me ask only this one, on what i myself will worry most before or when i try this myself, something i want to do (ánd to ask someone who already knows) about since blueprint and kreosan and keysightscience made magnetronguns with hornantenna's. I find that not especially interesting, but that changed with your experiment wich shows that plasma flame. Thát is indeed interresting to experiment with!) So my long lasting question: if you did nót used your Faraday cage, can you féél something from the microwaves?? Warming-up your hand of so? I know you can go blind when it hits your eyes too much but watching the microwaveguns they builded and their nonchalance while demonstrating, i always wanted to ask someone like you for your experiances with escaping&handhitting microwaves, if you can feel them warming a hand or finger (of course,or so i hope, you cán feel it before it damages unrecoverable. But only one way to know: ask it (ór try it myself..))

@@robson6285 Hey no worries, your English is good and I understand you perfectly! My circuit is different, i put the 2 MOTs in anti-series (so that the cores are ground level) and the primaries also in anti-series. So basically it is just 2 MOTs in series, so 4kv ac or so. Then I fullwave rectify that to get the dc supply. In most runs I don't even use capacitors! You bring up a really great point, since I am using this schematic, the magnetrons body is floating at +4kv! Very dangerous to water cool it like this, but with pure filtered water, I did not have an issue. I really should do it the way you said, so that I can have my magnetron grounded, but I don't want to use any voltage doublers, because the capacitors capacitive reactance will limit current if I do not use enough capacitance. I wanted absolute peak power output without relying on those capacitors. I don't know if what I just said makes sense 😅

@@robson6285 Feel free to ask as many questions as you want! Im glad you found this interesting as well. If i didnt use a faraday cage, i would feel the standing waves on my hand for example, and you can feel the heat for sure. I would very very strongly suggest that you don't try to find out for yourself, it is just so dangerous. It can kill cells in your testicles easily and it can very easily heat your eyeballs until it causes cataracts, and you likely wouldn't feel pain before irreversible damage was caused! I took a huge risk to do this, even though I am aware of the dangers and I understand the operation

Haha I can pretty well guarantee that the magnetron hates me, it only had passive cooling and it got *hot*. I now have implemented water cooling, which works quite well. These little magnetrons can withstand a ton of power, in order to break it, either the copper cavities would have to melt, the filament would have to break, or the beryllium insulators would have to break. Extremely dangerous indeed!

@@triggerhappy77707 Dear friend First of all, thank you for sharing I have two questions to ask Can this Faraday cage prevent microwave leakage from the magnetron? Can magnetrons generate such a long plasma flame and work like this for a long time?

Thanks! If I would have used a tank capacitor with more capacitance and retune the secondary for the new primary resonant freq I could have made much bigger sparks at a lower breakrate/spark gap speed. Oh well I rather liked the very thick sparks from the high breakrate

Hey thanks for the reply! I totally agree, I have seen others use magnetrons in these different ways but there never seems to be info. That's part of the reason I decided to do this and upload to RUclips, since there is little info out there on using magnetrons with a small resonator to act kind of like a hf vttc/plasma tweeter. Now I can hopefully help other enthusiasts like myself to understand how it works!

From what I understand, the magnetron itself is the resonator, not the breakout tungsten piece. But yes, it is a similar principle to a HFVTTC although at a significantly higher frequency. I've seen a couple videos of magnetron plasma flames floating around, but in general (and in my experience) they seem to be weaker than the flame from a proper HFVTTC.

Thanks! I imagine that once I employ the fullwave rectified 4kv dc power supply, the plasma will be much more stable. And then I audio modulate it 😁

@@triggerhappy77707 You may need to implement some type of way to induce an ac (audio in series with the primaries of the transformers).

Welcome are you by the way!

@@triggerhappy77707 WoW, i want to see that!

@@anthonyvolkman2338 I will be using a different means to facilitate the audio modulation. I will have a transistor between the cathode and negative supply to drop the voltage, which should give me decent amplitude modulation. I may just use my GU-81M pentode in series with the HV supply and modulate the entirety of the HV supply. Too bad these magnetrons don't have grids lol

I was using a driver that was very similar to Steve Ward's "new drsstc driver". I have no idea what the inductance was, I think the potentiometer was around 100 ohms. The inductor is something you can easily experiment with, just a few turns over a former like mine will work and you can add or subtract windings and see how it effects your IGBT switching times or how it effects the DC bus. Just be sure to keep the inductor away from the primary circuit, it could induce some nasty oscillations that could hurt the bridge!

@@triggerhappy77707 thank you so much, I know it’s been awhile but this circuit worked perfectly for my coil! I have a 49uh inductor and a 100ohm potentiometer and it almost completely gets rid of ringing!

@@blackcappedchickadee2581 Well that's awesome man! That should do a lot to help keep your IGBTs alive, considering your circuit has OCD, I think the most likely way to kill IGBTs would be that nasty ringing accompanied by voltage spikes on the DC buss. Once you have the feedback phase set just right, I figure you can probably minimize the ringing by using a really good low ESR low ESL decoupling capacitor directly across the leads of the IGBTS that go to the DC buss. Good luck man, if you have any other questions I'll try my best to answer

@@triggerhappy77707 thanks so much!

The input to the bridge was through an isolation transformer (which I was using one of it's taps for 80v instead of 120v in this run). Therefor I was able to connect my scope lead directly to the bridge output. The power was low enough that the voltage didn't ring above the voltage rating of my scope. The other channel was connected to a current transformer with a burden resistor to give me an accurate graph of the current, which is how I tuned how much phase lead I needed to switch the IGBTs properly without causing massive ringing issues as shown in the video. Hope that was somewhat helpful!

triggerhappy77707 it is very helpful! Thanks so much!

triggerhappy77707 wait so you got the square wave from the bridge output or the current transformer?

@@blackcappedchickadee2581 Glad I can help! The output of the bridge is nominally a square wave, and the current through the primary circuit are sinusoidal

Hello, the type of circuit you described may work, but I'm not sure what kind of circuit it is. Can you show me the motor driver so I can give you better advice?

@@triggerhappy77707 www.robotshop.com/nl/en/cytron-10a-5-30v-dual-channel-dc-motor-driver.html This is the motor driver I mentioned. Please have a look

Hi, and thanks. Both fets are the same, both N channel though I have also used N channel IGBTs. The IC I use for this circuit is made to drive 2 n channel mosfets in a halfbridge. The fets stay cool because I was able to make them switch efficiently. The ir2153 does most of it for me, such as adding some dead time to prevent shoot-through current. But I also used decoupling capacitors for the input right next to the fets which prevents unwanted oscillations and I used transient voltage suppressors to keep any voltage spikes from causing the fets to 'avalanche'. I used a proper sized resistor on the gate as well as a de-queing diode in parallel with the resistor. This helps turn the gates on a bit slower to prevent damage to the gate or shoot-through, and the diode help discharge the gate very fast for better switching. I hope this answers some questions, if you have any more I am more than happy to help!

The way that I fixed my issue was by increasing the inductance for the input choke and lowering the speed of the rotary spark gap.

its for the voltage multiplier. Electrolytics are no good fkr high frequency

One thing I would definitely suggest is to use a grounded metal enclosure to reject any noise if it is for a HV transformer. The potentiometer I used acts like an antenna if not grounded properly and causes severe freq fluctuations and possibly high shoot-through currents through the mosfets, killing them. (be careful grounding anything using mains! I made an earth ground exclusively for projects by jamming a conductive pole 6 foot/2 meter into the ground) . Another thing that definitely need consideration is the dropper resistor (if you power the IC off the DC bus instead of an external supply). I decided to use a much lower resistance than suggested for 120v, and I used an external 12.6v zener diode. This provides a higher current available for the IC, which is important to ensure proper switching of mosfets/IGBTs. In your case I would suggest around a 10K ohm resistor rated for at least 10 watts, which will provide an average of 34ma. I suggest a quality smoothing and decoupling capacitor, for example a 100uf electrolytic cap on the IC supply bus, with a .1uf film cap or even a tantalum cap connected directly across the gnd and Vss of the IC, as close as possible to those pins and as short of a connection as possible. If you are using a dropper resistor to supply Vss, the circuit will not operate correctly if your AC input is too low, with lower voltage the resistor will provide less current to the IC and at some point the IC may either go into low voltage shut down, switch the mosfets less effectively and cause higher losses, or may just act very erratic. You can see this happen in my video, when I am using a lower input voltage it starts arcing in "fits and starts". You can use an external supply to avoid this, and I would actually suggest that when you are first testing the circuit to use an external source for the IC so that you can test the switching part of the circuit at lower voltage to make sure everything operates as expected, just to try to prevent damage as much as possible if something were to fail. It is a good idea to start with extra primary turns when testing, maybe 30 or even more turns in your case. More turns will produce less output voltage as well as less output current, and will prevent something unexpected. I have experimented a lot with different primary turns and different frequency, and when I am using less primary turns for more power it is very easy to run a flyback at resonance or one of its harmonics, which can unwanted potentially VERY high voltages. For example, I have accidentally run a flyback at resonance and it produced such a high voltage that the HV output wire arced across the flyback to it's ground, causing carbon tracking that ruined the flyback. I have also had the same problem cause failure of the internal diodes which caused the flyback to produce AC which very quickly made it arc over internally. I've also experienced an accidental overvoltage that arced to the primary and killed the IC. I would HIGHLY recommend using a spark gap across the HV output to limit the voltage to prevent the failure I've had, at least while testing and experimenting with primary and frequency. I can give much more information if you happen to need it, and if you have any specific questions or problems I will likely be able to help!

Speaking of running flybacks at a resonant or harmonic frequency, in my video titled "Offline driver outputting ever higher voltage", I experiment with the frequency and run the flyback at resonance and you can see that the arc will start at a much longer distance, and at one point you can see a lot of corona around the top of the flyback. This is very bad as the corona very quickly eats away the plastic near it which eventually creates carbon tracking, effectively shorting the secondary.