Donald Smith Devices too good to be true

[tesluh]

Thank you Clarence! Read through hundreds of pages of this thread and yours and mr cleans results were very interesting and I learned quite a bit from them. Were you able to get the problems with your device solved? I was reading you were having problems with "too much output" which is a problem I hope to have eventually

the video only shows me a cpi brand nst, not sure if #30 was the video to watch? looks like this one: CPI Advanced 12,000 Volt 42 mA Neon Sign Transformer - Power Supply - CPI-EZ12

Can you tell me from the above schematics which is the better place to put the primary spark gap, the primary is different in each drawing, I was going to use the top schematic but need a third connection on my nst that wont fry it when i connect it? Also if the nst doesn't have a ground lug, can it still be used?

[clarence]

Maybe someone here can help? Looking for power supply that will fit the needs of this device. I have 6 different ones here and all seem to have something wrong. I believe what I need is to have either 12vdc or 120vac input and 4-12kv 20hz plus output, no gfci and the center tap of the secondary separate from the chassis ground of the device for use with the "return" spark gap that also gets earth grounded. the old 60hz steel box nsts don't seem to fit that description nor do the newer high frequency solid state devices as they have gfci or no separate center tap.

The Bertonee Don used is not in production and I have searched and cant find a used one anywhere. I saw one in the video above that is pn hm106nd but I also cant find that exact one. the similar ones to hm106nd that I ordered both have ground lead terminal connected to ground pin on input cord and likely also a gfci inside potted resin. I fried one by touching the ground lead to spark gap.

I keep going back to the idea that nsts are used on tesla coils and this is basically building a version of a tesla coil. What high frequency nsts are used to make tesla coils and what does the primary wiring diagram look like?

I am not opposed at all to using something like that pvm12 plasma driver, a function generator, plasma globe power supply, television flyback transformer, handmade transformer etc, but just need to know details so that i have the desired outputs.

I thought I had the primary side all figured out and now I am back to basics and trying to find appropriate parts again! arrrgh!!!!

the schematic in this post is what I am trying to duplicate if you are wondering why i need 3 output leads on the nst. not saying this is right, if someone has a better schematic i am open to trying it but it's similar to Don's board. I tried it tswift way and my nst doesn't have the same outputs as his plasma driver.

Hello Sir,

Here is the rest of the schematic that you are looking for.

Also for the type of NST and how to apply it got to YouTube for the KdKinen
file and I believe its the one that says "Capacitors Charging".
KdKined (MR Clean) is also a member of the energetic Forum. (Hope So).

Hope this helps Sir!

Regards,

Clarence

[tesluh]

trouble finding nst

Maybe someone here can help? Looking for power supply that will fit the needs of this device. I have 6 different ones here and all seem to have something wrong. I believe what I need is to have either 12vdc or 120vac input and 4-12kv 20hz plus output, no gfci and the center tap of the secondary separate from the chassis ground of the device for use with the "return" spark gap that also gets earth grounded. the old 60hz steel box nsts don't seem to fit that description nor do the newer high frequency solid state devices as they have gfci or no separate center tap.

The Bertonee Don used is not in production and I have searched and cant find a used one anywhere. I saw one in the video above that is pn hm106nd but I also cant find that exact one. the similar ones to hm106nd that I ordered both have ground lead terminal connected to ground pin on input cord and likely also a gfci inside potted resin. I fried one by touching the ground lead to spark gap.

I keep going back to the idea that nsts are used on tesla coils and this is basically building a version of a tesla coil. What high frequency nsts are used to make tesla coils and what does the primary wiring diagram look like?

I am not opposed at all to using something like that pvm12 plasma driver, a function generator, plasma globe power supply, television flyback transformer, handmade transformer etc, but just need to know details so that i have the desired outputs.

I thought I had the primary side all figured out and now I am back to basics and trying to find appropriate parts again! arrrgh!!!!

the schematic in this post is what I am trying to duplicate if you are wondering why i need 3 output leads on the nst. not saying this is right, if someone has a better schematic i am open to trying it but it's similar to Don's board. I tried it tswift way and my nst doesn't have the same outputs as his plasma driver.

[tesluh]

i will add this to my experiments. I have a new nst to try out tonight. seems wired the same as the other one but I haven't cooked this one yet! are the grounding rods separate?

[tswift]

After further testing, I'm still calling this one a qualified success. Starting with an almost completely discharged battery reading 10.1 volts, the voltage on the battery climbed to 10.42 over some time. However, the battery gets weird so it's hard to be sure. At first the voltage droops for a few minutes. Then if you stop the circuit, let it rest for a few minutes to get a good battery reading, and restart it, the voltage starts gaining. It's impossible to get a good reading with a digital voltmeter while the circuit is running. But as it runs some more, over time the battery readings get strange. After several hours, it seemed like the spark gap was running less strong, so I stopped it and the battery read 8.00V initially... but then very slowly climbed up to 10.42 over a period of 15-20 minutes. My voltmeter also likes to go crazy and give erratic readings on the battery, you can watch it start climbing slowly and then peak out over the 20V range (not an autoranging meter), then go back down to a reasonable reading and repeat this over and over. If I put the same meter on a conventionally charged battery it doesn't do this. Sometimes it calms down and gives believable readings. If I check the battery voltage with the oscilloscope it doesn't show this erratic behavior, but the scope is much less precise.

The conclusion I am reaching is that the battery is getting charged with at least some radiant energy, as Don described. Since the test configuration is so simple I recommend everyone to test it for yourself and see what results you get. It appears to be self-running some of the time, but not all of the time. Why, I don't know yet. Test it for yourself and see what it does for you.

[tesluh]

That's awesome! I have all but the diodes to put this version together, found both of my inverters were square wave, finally found one that is closer to sine wave. still not the same smooth sine wave as mains power but should be close enough.

got a new nst in that gave me rising trumpet waveform right away the first time I hooked it up. I think I partially fried it though by touching the ground lead to the spark gap. It still works but the wave form changed permanently after I touched it (only $37) so not a huge deal.
no diodes, connected one lead to the end of the coil with capacitors across it and spark gap in series on the return line.

once diodes come in I will wire it up right and see if I get any output.

[tswift]

After some more testing, this configuration APPEARS to be self-running. Upon starting the PVM12, the battery voltage droops a bit for the first 5-10 minutes but then starts gaining very slowly afterward. I need to run it continuously for many hours to be sure, but if true it confirms what Don said about making the battery leads 1/4 wavelength long keeping the battery charged. The battery leads shown are 68 inches long, the same as the wire length in the coil to within an inch or two.

The scope trace shows what's happening, the yellow trace is from the probe attached to the battery positive terminal. At each spark it bounces quite a bit and the "kickback" diode should be pushing in current on each positive peak. The blue trace is from the scope probe sitting next to the coil on the bench separated by an air gap with only capacitive coupling, so it's measuring the "disturbance to ambient" as Don would call it. The two waveforms are quite evidently in phase so the electrons being pushed into the battery are picking up the scalar characteristic of the ambient voltage at that point in the waveform, which then causes reduced Lenz's law in the high-voltage transformer within the PVM12. The spark gap is adjusted very narrow, almost touching, and the PVM12 is set to maximum frequency (about 70 KHz). I have run this configuration for a number of hours off and on now, but not continuously. The battery is very nearly dead, it was reading 10.1 volts yesterday before I ran the test. After about half an hour it was at 10.3, when I shut it down. This morning it's still at 10.3 after sitting overnight so the gain appears real. As I said, more testing is needed but this looks like a preliminary positive result.

[tswift]

I took a few scope shots to better illustrate the ringing you're looking for. This is essentially the same bench setup shown before, the PVM12 into a high speed high voltage diode, then through a very narrow spark gap, then into the coil which has the other end grounded. The spark frequency is 69 KHz. The ringing frequency of the coil (with no cap across it) is about 12-13 MHz, as can be seen in the second scope shot. I am using the cursor feature of the scope to measure three waveform peaks, so multiply the computed frequency by 3. This is the same waveform as the first picture, just zoomed in greatly on the timebase axis to show the first microsecond after the spark happens, when the coil is shock-excited. The waveform is complicated and messy at first as the spark excites not just the fundamental frequency but all kinds of overtones. The overtones damp out more quickly than the fundamental as it rings down. Since this is a system with one end grounded, the overtones will be primarily odd order, like 3f, 5f, 7f, etc. However, the grounding is not zero impedance and the ground wire has its own length so in the real world it's quite complicated as the scope shows.

The length of this coil is about 68 inches. Using the 247/L formula it should have a frequency of around 43 MHz, not 13. The additional inductance and distributed capacitance caused by winding the wire into a round form causes this. I assume a straight conductor of the same length would test reasonably close to the computed number. This is one of the great difficulties with the Don Smith device, you can't just do a little quick arithmetic, slap some parts together, and have it work. You have to be able to measure and tune. Even Don said to take a neon bulb and find the point of maximum output on the secondary coil because it's hard to compute exactly.

[tesluh]

[VIDEO]https://www.youtube.com/watch?v=43hted5YTCw[/VIDEO] Need This!

[tesluh]

Yes! I just realized my diodes might also be backward to the way I wanted them to go, which adds to my problems. going to buy some faster diodes like you said and make sure I am putting them in the right direction. Agree that having built a tesla coil would probably answer many of my questions...if I stall on this design I will go back to basics and build one which should help for sure. been trying all different configurations and seeing what happens. but most of them were with the diodes backward from the way it figure they should go.

***just figured out my inverter is square wave. clearly labeled sine wave inverter, definitely Chinese.***

[tswift]

Looking good, you're off to a good start. The real problem we face as experimenters trying to recreate the Don Smith device is that the available information is incomplete and we know it. So since I don't have a fully working device either, all I can do is (A) share my own theories and test setups and (B) give you advice on equipment and technique. What you will need to do is accumulate a selection of different parts that you can fit together in numerous different ways, and get good at building configurations, testing and observing the results, and then rebuilding to examine different aspects yourself. For example, the spark gap position. What happens when it's in parallel vs. in series? What do you observe?

A couple other things... you have a high voltage diode but it's not a high frequency ultrafast diode. When working with one of those NST's that isn't 60 Hz but something higher you'll need a faster diode. That one will (A) heat up and (B) not block well if at all in the reverse direction so it won't really be rectifying. The ones I have I got from Ebay and they are marked UX-FOB, rated 16 KV and sold for use with inverter type microwaves. Also, it looks like you have a digital storage scope. There is probably a way to capture a single waveform trace and store it so you can measure it more easily and it isn't jumping around all over the screen. Try to find this and get a clearer picture of what is happening immediately after the big vertical jump in the waveform, this is where the spark is happening and you want to zoom in on the first part of the ringdown.

In a regular classical Tesla coil with a primary and secondary both ringing, the ringing waveform seen by a scope probe some distance away is mostly due to the secondary and the modulation envelope will first rise within a few cycles as the primary transfers energy to the secondary and then slowly ring down over some dozens of cycles. Experience with building and tuning a regular Tesla coil is highly relevant to building the Don Smith device, and you even need most of the same parts, so I would recommend doing it as a small side project. If you have a 60 Hz NST without a GFI and a variac you're mostly in business already. You'll need to wind a primary and secondary, but this doesn't have to be complicated or expensive. I scavenged a microwave oven transformer for the wire, and used the 16 gauge primary wire for the Tesla primary (shown in the picture) wound around an oatmeal can. The secondary is about 14" long wound around a 3" cardboard mailing tube using the 26 gauge wire from the MOT secondary and the two match quite well with 10 nF of capacitance across the primary. I can drive it with a transistor as a CW slayer exciter, or with a spark gap as a real Tesla coil. For just a few watts of input power I can get 2" sparks out of it but of course our primary purpose here isn't to make pretty sparks. For the Don Smith device we need to make this oscillating electric field around the ringing coils (which Don called "disturbing ambient") and see if we can can collect power from it. It's easy enough to transmit power wirelessly between two Tesla coils with one as transmitter and one as receiver, but this by itself is not overunity. You can test this yourself by winding a second identical Tesla primary and secondary, one MOT is enough wire to wind two this size. As with all science, the way to progress forward is to TEST THINGS YOURSELF instead of just accepting textbook explanations or other people's opinions. Your goal should be to acquire the necessary test equipment, supplies, and skills to be able to do this.

Also, you can begin looking for signs of overunity now with this configuration you have going. Do what I did and connect your battery with a wire the same length as the coil you're zapping. Another coil of identical dimensions would be perfect, but start with just the wire length and find out. Actually you would need two coils, one for positive and one for negative, or else make a bifilar coil out of something like speaker wire that's two strands. Put a high-speed or ultrafast diode in the battery positive lead. Test the battery voltage, run the circuit for a few hours, then turn it off and wait for the battery voltage to settle (perhaps 5-10 minutes, watch it on a voltmeter and observe how it recovers after the load is removed), then find out whether it's self-charging or not as Don said. You will also need to make sure there is an earth ground on the NST that's properly grounded (into the third prong of the house wiring works OK) in order to have any hope of seeing the Don Smith effect manifest. Then coil up the battery cables into a coil with similar geometry as the one you're zapping and test again. Does it really make any difference? If you can observe self-charging happening, then you already have a self-running overunity device. Then it's a matter of figuring out the L2 coil and output circuit to get larger amounts of power.

[tesluh]

Very nice, thank you for the example, I set mine up just like that and took some scope readings, I used my other nst as well and tried two different coils, also played with capacitors across the coil, they were all similar to this and not much I did changed anything. I believe I am looking to have a trumpet shaped trace or "whale shape" instead of what I have now. I had a ground wire connected to the sparkgap.

Any advice on this? it's hard to know what to do/change when I am not exactly sure what I am looking for in the first place.

[tswift]

Here's an example of this applied to a Don Smith style circuit. The coil is the primary from my small slayer exciter / benchtop Tesla coil (depending on input). It's 3.75 turns around a 5" coil form (an oatmeal can). The capacitor is 10 nF in a series/parallel array so it can take up to 6 KV. This coil and cap in parallel resonate at about 675 KHz, which is a perfect match for my 3" mailing tube Tesla resonator (not shown in this picture). Here I am zapping this coil/cap combo through a spark gap placed between ground and one side of the coil/cap combo. The NST in this case is the PWM12 I use for many of my experiments, through a high voltage diode. The whole cap/coil combo is hot (electrically with HV) at the NST frequency when operating, and each spark of the gap also triggers the coil quarter-wave resonance, as can be seen with the scope probe just barely visible in the picture. As per Don's instructions, the battery leads are also tuned to quarter-wave resonance at the same frequency, because they are the same length (measured and cut to about 68 inches in this case) and I have coiled them similiarly to try and get a more precise resonance match. With the kickback diode in the positive lead (it's small but can be seen at the positive battery terminal), this arrangement is supposed to be self-charging according to Don. Indeed, in the hour or so I have had this configuration running on the workbench this evening, the battery voltage has not only not dropped, it has gained slightly. Not that this is really conclusive: a battery this big can run this circuit for many hours and lead-acid batteries exhibit numerous behaviors that can fool you. As you observed, trying to use a digital meter anywhere near an operating circuit is useless, I have to turn it off to take measurements.

Essentially this is the Don Smith configuration up through the L1 coil. Don's "tabletop" device is really two separate overunity stages in series, and this is the first. The battery acts as radiant receiver and stays self-charged, with the overunity happening in the NST transformer. The amount of overunity power involved is quite small, only a couple of watts probably. The second overunity stage uses the big cap bank as the radiant receiver and the overunity happens in the output transformer stage, and this is where the big power gain happens. Or at least I'm pretty sure will happen, once I have it working.

[tesluh]

Thank you!!! That was a bunch of what I needed to know, like I said if I can get one up and running I will be glad to share as much as I can back to help. I think my actown nst is a 60hz and my 12v lowglow automotive nst will put out the ~37khz. it made a hiss sound as opposed to the pop sound across the spark gap. I have tungsten tig welding electrodes for spark gap, it stays really cool. I have high voltage gloves and wear safety glasses when I turn the thing on.

been reading this DON SMITH BOOK and he says that once radio freq is achieved the electrons begin spinning free outside of the inductor. Definitely sounds like what I would think of as an open system.

[tswift]

First of all, let me welcome you to the rabbit hole. I have gone quite a ways down it myself and I am happy to give you some pointers.

First, capacitive coupling is your friend. The high input impedance of an oscilloscope means that anywhere within a few inches (to even a few feet) of an oscillating circuit will show you a scope trace. Don't try to attach the probe directly to anything with high voltage or you will fry your oscilloscope. Scopes and probes differ in ratings, mine is supposed to handle 300V at the probe inputs without damage. The probes are rated for 600V. You will get better bandwidth and transient recording, and possibly better protection of your scope from high voltage, by using the x10 switch setting on your probes rather than the x1 setting. High-voltage probes exist that can handle several KV input range, and these operate on an x100 or x1000 multiplier setting. I don't have one, but have considered getting one to help in my experiments. The good news is that the low voltage probes work fine and all you have to do is get them anywhere near an oscillating circuit to at least see the frequency of oscillation. If you have a digital scope with cursor features it's simple to get frequency readings from the scope trace, otherwise with an analog scope you have to count divisions on the screen and so some arithmetic.

One of the things I have learned about the Don Smith circuits so far (even though I don't have a measured overunity gain-producing arrangement) is that there are definitely TWO different resonances going on. The first resonance is the quarter-wave resonance with the wire length, and is generally in the megahertz range (as in 247/feet=MHZ). If you take your NST and zap one end of a coil through a spark gap (ground the other end of the coil), you will see this quite clearly. Don't put a capacitor across the coil, just the coil itself. Put a scope probe a few inches away from the coil and turn down the volts/div setting on the input channel until you see a trace with every spark. Adjust the timebase setting until you can see a decreasing sine-wave envelope which may take many cycles to ring down. Now you can use the scope trace to get a frequency by calculating the period. You will see it's not exactly 247 divided by the wire length, but somewhere in that ballpark. The coil geometry matters for the exact frequency.

Now put a high-voltage capacitor (something like a polyester film snubber cap) across the coil terminals and zap it again. Adjust the spark gap so you don't exceed the voltage capability of your capacitor, you probably want the gap as narrow as it can be and still have a gap. Now look at the scope trace and you will still see the same ringing at some megahertz frequency, but now increase the sec/dev knob several clicks and you will see that there is also a much lower frequency ringing happening. This is the frequency determined by the LC resonance of the coil and capacitor. If you have an LC or LCR meter, measure the coil and capacitor separately and do the arithmetic for the combined resonant frequency. It should be quite close to what you measure with the scope. Again, it will be a ringdown with every spark. If you zoom out farther with the timebase it will be obvious what the spark repetition frequency of the NST is. Unless you have a variac to turn down the voltage of the NST so it just barely fires the spark gap at the peaks of the AC waveform, the gap will fire many times each cycle (and probably heat up quite quickly, beware). If you put a high-voltage diode on the NST lead before the spark gap, and a variac on the AC input to the NST, you can start with the variac at zero and turn it up just until the gap begins firing and it will fire 60 times a second. If you put separate diodes on both hot leads and use both of them going into the spark gap it will fire 120 times a second. If you use a high-frequency NST (such as a 12V unit with a built-in high frequency inverter), the gap will fire with every AC cycle, so it will fire at 35 KHz or 50 KHz or whatever the frequency of the NST is, usually somewhere in this range. This is much quieter than a low-frequency spark and basically just sounds like a hiss.

And obviously, using a hot lead sparking to ground will trip the GFI of a GFI-equipped NST unless you defeat it.

Now as to which combination of gap position works in the Don Smith device to make the magic happen, I'm still working on it too. Hopefully what I've written makes sense and you can start testing configurations yourself. Please do be careful and observe all necessary precautions for working around high voltage, I assume if you've acquired the equipment and gotten this far then you're probably experienced.