Post by: Quentief on February 03, 2020, 06:57:51 PM
I would like to know if it is possible to design an arc oscillator (the spark gap circuit, or the primary circuit on a Tesla Coil) which would be able to use a voltage as low as 230 volts, instead of something about some kilovolts. According to my own experience, we cannot make electric arcs from 230 volts, the voltage is too low to make stable and sustainable electric arcs.
But I would like to have your mind-view on this question.
And I apologize in advance if my English is not perfect, actually I am French :) But I will try to do my best to be understandable.
Post by: klugesmith on February 03, 2020, 07:06:50 PM
Those are enclosed spark gaps with relatively low, well specified breakdown voltages. Easy to get from electronic component distributors or ebay, etc.
https://www.bourns.com/products/circuit-protection/gas-discharge-tube-(gdt)-surge-arrestors/2-electrode-gdts
They might not be a good fit for Tesla coil circuits where you want a spark on every cycle of AC power.
Maybe worth a try. You could make a mechanical or solid state interrupter to make TC primary circuit fire less frequently. Once fired the GDT will have a very low impedance in both directions, just like a spark.
Another idea is a regular spark gap in partial vacuum.
A look at Paschen's Law suggests that that can't get you down to 230 volts, no matter how close the electrodes are set, at any pressure. The peak voltage at 230 volts RMS might be reachable. In argon?
Post by: Quentief on February 03, 2020, 07:21:03 PM
Post by: Quentief on February 03, 2020, 08:16:29 PM
Furthermore, I have seen people on YouTube which were able to draw electric arc from 230 volts.
/> /> /> /> /> /> />
Apparently they were using coils or capacitors. Can I use this in a Tesla spark gap ? And how can I design capacitors or coils for this function ? I mean, I know the phenomenon (I=C*dU/dt and U=L*dI*dt) but how can I find the capacity or the inductance required for this purpose ?
Post by: MRMILSTAR on February 03, 2020, 08:55:11 PM
Even if you could bridge a spark gap with such low voltage, you probably would not want to build such a SGTC. This is because the energy in the capacitor is related in a non-linear fashion to the voltage but only in a linear manner to the capacitance as can seen in the capacitor energy equation as:
E = 0.5*C*V*V
where E is joules, C is farads, and V is voltage. You would need a very large capacitance from a pulse-rated capacitor to obtain a decent amount of energy per bang from such a low voltage. Its easier to get energy from voltage than capacitance.
Post by: Quentief on February 03, 2020, 10:05:19 PM
I know there is maybe nothing to do with my previous question but if you know how to calculate the capacity or the inductance required to stabilize the electric arc, I am more than interested.
Post by: davekni on February 04, 2020, 04:54:31 AM
I haven't personally played with 230V arcs, but inductor filtering is likely best. A capacitor filter would require current limiting after the capacitor to keep the arc from discharging the capacitor too far. For inductance calculation, a SPICE simulation would be ideal. For rough estimation, the AC component of rectified 50Hz line is 100Hz. Find an inductor with at least the same impedance as your electric kettle (series current-limiting impedance) at 100Hz. Higher is even better.
As Steve pointed out, 230V for a Tesla primary has many issues. Another is that your primary coil would need to be ~0.2 turns :)
If you really want to experiment with SGTC circuits "directly" from 230V line (no line-frequency transformer), a diode/capacitor voltage multiplier could get the voltage high enough for a single-turn primary. Perhaps two voltage-doublers, one + and one - , to get +- ~600V peak. The first cap in each doubler could be a larger electrolytic. The second could be each side of a 2S MMC. This would provide only 50 sparks/second, requiring a full cycle to charge both halves of the MMC. For the spark gap, either a rotary synchronous design with electrodes that touch slightly, or a triggered spark gap (short gap with small trigger electrode connected to a small HV pulse circuit). Even at 1200V total, your primary will be one turn, so stray inductance in the spark-gap and wiring will make the coupling factor rather low. (My SGTC is fairly low voltage, 9kV into a 3-turn primary. Even there, stray inductance reduction is important.)
If you want a 230V spark gap for some other use, triggered gaps can run that low, at least for argon. Higher-end TIG welders use a high-voltage trigger pulse to start the arc, avoiding the need to touch the electrode to the work piece, and thus keeping the electrode cleaner. The same thing should work for a separate trigger electrode. (Of course, this is for high arc currents. May not work at lower current. May not work as well in air either.)
Post by: Quentief on February 06, 2020, 02:49:43 PM
It is not as powerful as what I was expecting. I don't have the instruments to measure the current and the voltage, to calculate the power of the electric arc, but I was expecting something able to melt the iron from the electrodes. But again, thanks it was really interesting to experiment it.
Now concerning my first question, well actually I do not want to build a Tesla coil, I just want to increase the frequency of the main 230 volts 50Hz. Indeed, I am not looking for very high voltages. I found the primary circuit of the Tesla coil very interesting and I would like to know if it is possible to apply its working principle to a lower voltage.
So, according to you, a classical spark gap cannot operate with 230 volt because of the air breakdown voltage. I suppose the rotary spark gap solution could work : if the contacts are touching them together at each cycle, the voltage does not need to be very high.
The solution proposed by klugesmith is also interesting : a spark gap in a different gas. Can I use a Fluorescent lamp as a spark gap ? If I have understood their functioning properly, the gas inside the lamp is designed to let the current passes threw with a lower voltage than if it was in air.
Post by: ElectroXa on February 06, 2020, 06:31:54 PM
1) they trigger at around 500V, so 325V DC won't pass through the lamp unless if it was triggered with a high voltage pulse ;
2)the DC voltage will tear one electrode faster than the other, leading to failure,
3) and finally, fluo lamp won't withstand such a high (pulse) current.
Post by: davekni on February 07, 2020, 05:47:25 AM
What is the resistance (or power rating) of your electric kettle ballast?
Arc welding is typically done at 50+ amps and <20V. A 1000 watt kettle would limit to ~4.4A, much less than an arc welder. Voltage across the arc is still dropping to ~20V, so arc power may be too low to melt the electrodes. Even so, at 4.4A, I'd recommend some eye protection for the relatively-intense UV.
As already mentioned, a fluorescent tube won't work for multiple reasons. A low-pressure argon gap might work. A spark gap triggered by a small HV trigger pulse is more likely to work.
Depending on what you are trying to accomplish, electronic switching is probably much better than any spark-gap or mechanical (rotary) switching. Are you after short high-power bursts of higher frequency? Would longer bursts of lower peak power (same energy) work? If you share more about your goals, it would be easier to suggest solutions.
Post by: Quentief on February 07, 2020, 11:04:18 AM
Concerning the electric kettle, it has a 22 Ohms resistance, so the current is limited at 10 amps, but it still stay very lower than the current from arc welders.
However, some argues that it is possible to weld by using the main voltage, is it a fake ?
https://www.instructables.com/id/Water-and-salt-Welder/
Concerning what I am looking for, actually I would like to try to induce current in metal, to heat them. In that respect, I would like to know if the primary circuit from Tesla coil can be adapted for that purpose.
Post by: Twospoons on February 07, 2020, 11:00:54 PM
So theres your clue - you can build a 230V spark gap, but you'd want some form of pre-ionisation to lower the trigger voltage.
Electrode shape will also be important - you want sharp points to increase the field gradient.
Post by: davekni on February 08, 2020, 05:04:13 AM
There are many induction heating threads on this forum. If your goal is useful induction heating, I'd suggest reading those. If you are more interested in the fascinating challenge of trying something unique, have fun with spark gap induction-heater drive!
Concerning saturation current of your inductor, you can measure that with a scope and capacitor. Charge the capacitor, then connect it across the inductor, measuring the voltage waveform across the inductor. The capacitor's initial energy needs to be high enough to saturate the inductor, which will show up as flat-topped half-cycles in the ring-down waveform. Once the amplitude rings down enough to not saturate, the waveform will be a better sine-wave approximation. Pick a half-cycle that is still slightly flat-topped, then zoom in on the zero-crossing at the end of that half-cycle. The voltage slope times capacitance gives the current.
Welding with line voltage: Most breakers can handle short-term overload. The 120V circuit in the instructable was probably being ran at ~30A for welding. There are comments from people being unsuccessful with 230V circuits (probably lower current available) and of breakers tripping.
For a low-voltage spark-gap, yes, an ionization source helps. Radioactive isotopes are hard to come by and generally dangerous. Hard UV light is another option, but also hazardous. Given the high currents needed for your low-voltage spark gap, it will need aggressive cooling. That makes contained gasses difficult. The spark (arc really) needs to be extinguished quickly after each discharge, implying either electrode motion (rotary gap) or high gas velocity through the gap. My off-hand guess is that a rotary "gap" with tungsten electrodes may be the only reasonable option. Electrodes will need to touch to start the arc, and likely air flow directed at the gap to avoid too much molten metal.
At the high currents you need, any sharp point will melt/sputter away quickly. That's why I'm thinking a rotary gap with contact is the only option.
I'd also suggest a resonant charging circuit (inductor in the 230V supply tuned to one half-cycle per spark event when coupled with the resonant capacitors). There's more about resonant charging for SSTCs on the forum.
Good luck!
Post by: Quentief on February 09, 2020, 01:12:41 AM
I would like to ask you some questions about the inductance that I was trying to make : I think I did not calculate its inductance propely. Like I previously said, I use an iron core from an old transformer to make this inductance. A photo should be available in attachement.
To calculate its inductance, I use this formula :
L=μ0*μr*N²*S/l= 69 mH
with:
μ0 = 4π × 10−7 T m/A
μr = 2000
N = 35 turns
S = 9 cm²
l = 4 cm
Actually, I used the same formula for a solenoid. The parameters S and l are the dimensions of the core's part which is encircled by the coil. I do not know if I am wrong or not by considering the problem in this way. Furthermore, the core does not have a space to limit its saturation. I am sorry, I cannot find the English translation for that, in French we call that "l'entrefer" https://tse2.mm.bing.net/th?id=OIP.srfRsINmFNKAi-aLxlUFLAAAAA&pid=Api&P=0&w=300&h=300 , so maybe you are right, the saturation makes my inductance unable to have the desired impedance for the creation of 230 volts electric arcs. (the current used is about 10 Amps)
Post by: johnf on February 09, 2020, 03:52:08 AM
All your linked videos had either stepup transformers AKA neon transformer in one vid or high frequency home made transformewr pair in another vid, or used DC with an inductor to raise the voltage substantially above mains voltage. Yes all were powered from mains but not direct connect to mains
Go find a nice big OLD neon transformer without the GFC
Post by: klugesmith on February 09, 2020, 05:33:29 AM
The formula you used seems appropriate, except the term for length is for a complete magnetic flux loop (through center bar, end bars, outside bars, then returning to center bar. length of 9-square-cm steel path that completely encircles the coil.
WHen you understand why the formula works, then it's easy to determine the saturation current, and to see what happens when you add an air gap (reduced inductance and increased saturation current). There are many excellent introductions to transformer design.
Post by: davekni on February 09, 2020, 06:23:41 AM
Adding a small gap also makes inductance vs. current much flatter. The permeability of silicon-iron (transformer steel) varies significantly as flux density increases. The gap is linear (air), making the resulting inductor more linear.
To calculate saturation current, you'll need the saturation field of silicon-steel, around 1.6 to 1.7T if I recall correctly.
Post by: Quentief on February 10, 2020, 01:19:11 AM
Post by: klugesmith on February 10, 2020, 02:38:00 AM
Just as the wire length per turn is driven by need to go around the center leg of core.
There's not much room for improvement, over conventional E-I core shape. The magnetic path length in a toroid still needs to reach around the winding window area, and the wire length per turn still needs to reach around the cross-sectional area of the core.
If your coil of wire doesn't fill up the winding window, performance can be improved by using thicker wire, or find a smaller core.
When you make a choke with intentional air gap, the maximum stored energy (0.5 * L * I^2, at saturation) is proportional to air gap volume, and to the square of B_max. Steel beats ferrite by a large margin in that last factor.
Post by: davekni on February 10, 2020, 04:56:27 AM
If you can get a couple microwave oven transformers and access to a steel-capable band-saw or large chop-saw, two primary halves wired in parallel and physically stacked with a small gap works well for line-frequency inductance.
Post by: Quentief on February 10, 2020, 11:47:58 PM
To do this, the each coil has the same number of spire, but the middle coil is winded in a different direction than the two others. Moreover, the iron section of the middle coil equals the sum of the two others sections coils. So in theory, the magnetic flux of each coil will cancel each other, and the core cannot be saturated, according to what I read on the article.
Do you think it is possible ?
I calculate the inductance of the coil, I find 60 mH, but I am not sure. The details are available in attachement.
Then, I connect the coil with my setup to try to make electric arcs from the main 220 V. It does not work better as my previous attempt.
Post by: klugesmith on February 11, 2020, 01:21:11 AM
If you really have 60 mH, the impedance is 19 ohms at 50 Hz.
With the coil in series with 19 ohm resistor, both elements will have the same AC voltage (each about 71% of the applied voltage).
If voltage across coil is lower than voltage across resistor, then inductance is less than 60 mH, and you can get it from the voltage ratio and known value of resistor.
You might substitute a 230 V heating appliance for the resistor, and skip the low voltage transformer. 1 AC ampere should produce 19 AC volts across a 60 mH inductor. Incandescent light bulbs are not good to use as reference resistors.
Connecting winding sections to oppose each other increases the saturation current. But it reduces the inductance by the square of that factor. You are wasting wire.
If the coil winding directions are gauche, droite, gauche when viewed from top, then they are not going to cancel.
When middle coil is forcing flux up in middle leg of core, the outside coils are forcing flux down in the outer legs.
Heads I win, tails you lose. :)
Even if you want three coils to aid each other, using the outer legs can't increase the effective number of turns. All turns have to fit through the same winding window! Turns around an outer leg are less well coupled to the middle coil, and are coupled in opposite direction to the other outside coil.
You can change the relative winding directions without uncoiling the wire.
Bring out two wires for center coil, and two other wires for the outside coils.
Then you can externally connect them in series, or in parallel, with any polarity.
Learn something by measuring the inductance of series connection both ways.
Post by: Quentief on February 17, 2020, 12:10:28 AM
I did not check the inductance of my homemade coil. However, I replace it by a big transformer from my homemade spot welder (which is built from a microwave oven transformer). Here is the result : https://photos.app.goo.gl/z85tXhyhWVKtNm5K7
(sorry for the bad quality of the video, I was scared to damaged my phone because of the UV)
It works pretty good but I am wondering if the 10 amps are not dangerous for the MOT, the current may be too high for the primary coil and could overheat the transformer. I am wondering if I could use an electric motor instead, like a hair dryer motor : its universal motor should be able to run on the rectified 230 volt and its rotation should cool down its coil.
I have also another question, not linked with the topic : I cannot identify what is this component (it cames from an old PC power supply). Please, could you tell me what is it ?
Many thanks for your help
Post by: Quentief on February 17, 2020, 12:14:07 AM
Post by: johnf on February 17, 2020, 03:19:32 AM
That is an X2 rated AC capacitor ie AC275 volts used in common mode filter on mains input
Post by: davekni on February 17, 2020, 03:59:59 AM
Concerning microwave transformer current, look for the power rating of the microwave it came from (or a typical microwave if you don't know). Use that power and the line voltage it was designed for (ie. 230V) , and calculate a current. The actual will be a bit higher due to power factor and losses in the magnetron, but that provides a lower limit to the likely value.
Post by: Quentief on February 18, 2020, 01:25:06 PM
The power of the microwave oven was about 1300W, so about 6 Amps under 220 volt. In my setup, the current is limited with a 2200W kettle, so the current is about 10 Amps. I do not know if a such current can be tolerate for long time by the transformer.
Post by: davekni on February 19, 2020, 03:53:13 AM
Post by: Quentief on February 21, 2020, 12:28:20 AM
I made some tests with different inductances :
This is a motor stator from an old vacuum cleaner. I cannot calculate its inductance but by considering the cable resistance, its section and the copper electrical resistivity, I calculated the lenght of the coil wire : 35 meters. Then, I was able to estimate the number of wire turns ; about 441 for each coil.
https://photos.app.goo.gl/jwejCwZyEb1WGdBH6
I was able to make some little arcs through this setup.
https://photos.app.goo.gl/JMVTAXJ9KT9xAno96
Then I tried again with the MOT with its secondary short circuited. The arcs was longer. So Davekni seems to be right : it is question of current smoothing, and the bigger is the inductance, the better are the result.
https://photos.app.goo.gl/dogYxUDi1m6iG6oGA
Then I made a last test with the MOT but this time, the secondary was not short circuited. I had supposed the inductance should be higher and I think I was right, the electric arcs were much more longer than before.
https://photos.app.goo.gl/68FDk6ufyj1T9ny2A
However, I would like to know if it is safe for my diode bridge. I mean, at each current variation, the coil produce a voltage and if it exceeds 600V, my diode bridge could be damaged. For the same reasons, I prefer to not use my multimeter to measure the voltage of the coils used in the setup, to estimate with precision their inductance.
Post by: davekni on February 21, 2020, 04:37:53 AM
If I understand your circuit correctly, 220VAC line goes through a 22-ohm kettle to a 600V bridge rectifier. The rectifier output goes through an inductor to your two nails that form the arc. In that case, the high-voltage spike across the inductor when the arc extinguishes will show up across the nails and inductor, but not the bridge. The bridge sees only 220V (~320V peak), plus a small bit more due to stray inductance in power wiring to the 220VAC plug. If you want more diode protection, add TVS diodes or MOVs across the bridge (input or output), or use a surge-protection outlet strip for power (which has MOVs internally).
Your DMM should be safe measuring voltage across the kettle (to determine current) or across the diode bridge. Measuring across the inductor or arc electrodes might require some protection circuit (resistor divider and clamping diodes) before measuring.
Be very careful not to touch the MOT secondary when not shorted!
Post by: klugesmith on February 21, 2020, 07:21:13 AM
Post by: Quentief on February 22, 2020, 01:27:41 AM
Davekni you are right about my circuit, here is a diagram : https://photos.app.goo.gl/2MwyUU4DuAxzMFF79
However, I do not understand why you claim the high voltage spike produced by the coil cannot reach the diode bridge : according to the Kirchhoff's circuit laws the voltage across the diode bridge output equals the coil voltage + the arc voltage. So if the coil makes a voltage which exceeds 600V, it will damage the diode bridge, won't it ? Or have I missed something ?
To answer you Klugesmith, I knew the all motor would have a higher inductance, however I chose to remove the rotor to have a motionless inductance, safer. But do you think I could add a big iron part at the center of the stator, to improve its inductance and have something closer to the MOT ? I still have difficults to calculate the inductance of coils when they have a magnetic circuit.
Post by: klugesmith on February 22, 2020, 02:10:34 AM
If you don't have brushes connected, the armature won't try to turn when you put it inside energized stator.
A solid steel cylinder in its place would introduce large eddy current losses.
When are you going to measure the inductance of your ballasts? We've shown you how to do that with AC voltmeter and known resistors.
Post by: Quentief on February 22, 2020, 02:32:55 AM
Concerning the inductance, indeed I would like to know how to calculate it to understand how I can increase it. But you right, I need to do some measures.
Post by: klugesmith on February 23, 2020, 04:07:51 AM
Remember when you figured the inductance of coil on laminated E-I cores? One factor was the length of magnetic flux path in the steel. Not just the part inside the coil, but the return path that includes outside legs of E-I core.
Same formula applies to your universal motor stator, with an important difference. The flux path includes a large air gap between the poles, where the armature belongs.
[ Invalid Attachment ]
The inductance formula has a path length term, which goes into the core's magnetic reluctance (resistance to being magnetized). When computing reluctance of core with an air gap, each millimeter of air counts as much as at least 500 millimeters of steel. In the case of your motor stator, the reluctance from air gap is much larger than that of the steel path, which makes the inductance relatively small.
When you put the armature back in place, its copper windings don't matter. You are putting laminated steel where there used to be air. Remaining air gap will be less than 1 mm "long" in the direction of magnetic flux, so the stator coil inductance will be more than 10 times larger than with no armature present. Measure it!