Romanian ZPM (Zero Point Module) - Enhancements Stage

  • Last Post 01 November 2019
Fighter posted this 27 September 2019

As the ZPM thread has so much content that it started to load slowly especially on mobile devices, I'm creating a new thread continuing the initial thread.

If you want to search for specific information posted in the initial thread you can find it here:


Romanian ZPM (Zero Point Module)


In this initial post I'm also adding the donation section as I did in the initial thread:

If you find ZPM intriguing and the experiments and data I share here with you useful for your own projects and research, you may help the further research by donating:



Your support will be very much appreciated and it will be used ONLY for equipment and building customized electronics for further research and experiments.

If you're working with a ZPM replication please share your findings here like I do so we can enhance this device.

Thank you,

Order By: Standard | Newest | Votes
Fighter posted this 29 September 2019

Another UA7812 voltage regulator from the MOSFET driver smoked:

So I decided to get rid of this approach with voltage regulator for powering the switch LED, now that LED is powered by the separate power used for cooling fans and switched on or off by the main switch. Those switches have two set of contacts so when switch is turned on it can close two separate circuits.

Another thing I did was to put high-voltage super-fast SF28G diodes on the MOSFETs but I had some problems with this, when exploring with different frequencies few times I had sudden surges of the power drained from the DC source, the source entered in auto-protection and the 12V/55W light-bulbs from ZPM's output became extremely bright and were destroyed. So far I lost 3 light-bulbs of that type, I have only one left which is still functioning (link to larger image here):

For some reason seems ZPM doesn't like those protection diodes and at certain frequencies it suddenly starts draining full power from the DC source and smoking the 12V/55W light-bulbs.

For now I have set optimum frequency and the 12V/55W light-bulb seems to work fine but probably as soon as I will encounter one of those frequencies the same thing will happen again. If it happens probably I'll remove these protection diodes, I don't like this behavior and it's provoking damages, I don't know what's the reason but it's a fact.

cd_sharp posted this 29 September 2019

You need to limit the current using a resistor to 10-20 mA, whatever your LED needs. 12v / R = 0.01A.

Fighter posted this 29 September 2019

It's okay, now I removed that voltage regulator and both LED's are powered by the separate circuit powering the cooling fans. I suppose that voltage regulator didn't liked the pulses present on DC source's input, it expected clear DC voltage there.

  • Liked by
  • Chris
  • Atti
Atti posted this 30 September 2019



Because the bulb is constantly getting damaged. Perhaps it would be better to use a wire resistor in this form. Or it would be better to use a heating element. This way, the energy of the heat could already be measured. This measurement method is also recommended by Robert Adams. Temperature of water heated per unit time.
Disassembled old appliance heater
(if someone is trashed):
sandwich oven has two 110V heaters
water heater 220V

In my opinion, FET overheating is caused by the load current (12V55W bulb).
You can see it here.

ambient temperature is 20.6 ° C
The drive is FET 22.5 ° C
Bulb switch FET 44 ° C

Fighter posted this 30 September 2019

Sorry but that's an abnormal situation, not only the light-bulbs are destroyed but the DC source enters in auto-protection mode.

Is nothing to measure in that situation, I intend to avoid it as it could destroy my equipment...

  • Liked by
  • Chris
Fighter posted this 01 October 2019

Finally found some time to check the PowerSource, I verified IRFZ46N's drain-source continuity and found they were shorted so the MOSFET was damaged by that malfunctional DC adapter.

After I removed the MOSFET I checked that the circuit where its gate pin was is still receiving square wave signal:


As you can see the square wave was present so I mounted a new IRFZ46N. I must admit I had some big concerns as my soldering tool is not that delicate for this kind of fine/small circuits so few times I was thinking the copper areas where the MOSFET was soldered will be destroyed by the heat of my tool. There was a high risk of damaging the circuits from the MOSFET area...

After I finished I've made a test with the PowerSource module powering up a PowerSwitch module driven by my signal generator. Wanted to make sure the repair succeeded:


So now the PowerSource module is functional again but I don't intend to use again a cheap DC adapter, I bought a 12V/2A AC transformer and a big 4700uF electrolytic capacitor and using one of the bridge rectifiers I already have I'll build a dedicated DC adapter:

Of course I'll build it when I'll find some time again...

Fighter posted this 10 October 2019

Just a short post, a few days ago somebody asked me through email what's under that red tape covering the middle of the Metglas, if there is a gap and if it's something in it which I don't show:

There is no gap or something secret there, I did put that red tape when I first built the ZPM in order to to keep as close as possible the both pieces of the Metglas core together. To have everything crystal-clear I removed that red tape and used a strong plastic zip tie to continue keeping the pieces of the Meglas core together (link to larger image here):

All the information about ZPM is public, complete and published in this forum. And will always be this way.

It will be updated as I continue the research depending on the free time I can find.



Zanzal posted this 10 October 2019

Hey Fighter, while it is very nice of you to show that there is no secret sauce hiding under the tape I would urge not to get distracted with changes to convince people. I take you at your word that the device is exactly as you describe it. Do what you think is good, but you don't need to appease or convince the naysayers IMO.

Chris posted this 10 October 2019

My Friends,

Hahahaha Yes, I agree with Zanzal!

"Whats under the tape", hahaha a 26 Plate Lead Acid Battery?

Apologies, they say: "Sarcasm is a metric for potential - Howard Stark"?

I agree, sweep the silly aside!


Fighter posted this 10 October 2019

Hahaha, the joke about the battery made me laugh after a hard day at work, thank you Chris !

I don't know who the guy is but most probably he was thinking about some permanent magnets or God knows what kind of "secret" thing hiding under that red tape.... Well, I think he didn't expected me to reveal the "secret", I may disappointed him, I apologize for this...

patrick1 posted this 11 October 2019

lols yes, although it must be said. i dont see any nay-sayers.  more like kids in a candy store

  • Liked by
  • Chris
  • Vidura
Fighter posted this 14 October 2019

I've build the customized DC source for the PowerSource from Vidura, I tried to make it as compact as possible (link to larger image here):

I put the bridge rectifier directly on the transformer's core so the core can be used as a radiator if the bridge rectifier becomes warm.

But there are two problems with it:

  • the 4700uF/100V electrolytic capacitor seems it's not doing its job very well, I have a 620mV Vpp, the capacitor is brand-new;
  • the output is not 12V as I expected, it's about 16V.

(link to larger image here)

The transformer is 230V/12V and also brand-new (link to larger image here):

@Vidura, I suppose this is not acceptable for powering up the PowerSource, isn't ?..

About voltage I suppose I could put a resistor on output for lowering those 16V to 12-15V but I'm not sure what electrolytic capacitor I should use in order to have a linear DC.

I need to check on the scope the output of the 12V secondary, maybe the factory put more turns on it and it's actually 16V there...

Vidura posted this 14 October 2019

Hi Fighter, This voltage is normal when it is rectified and a capacitor connected. The rated transformer voltage is RMS , and so the output rectified and filtered will unloaded output the peak voltage. Anyway 16 volt should be ok, the driver IC is rated up to 18v if the resistors of the snubber circuit becomes hot you can lower the duty cycle on the preset of the module. Regards Vidura.

Vidura posted this 14 October 2019

I suppose that the scope trace is the ripple, 620mv is no problem at all , it is filtered again in the module , and the final stage, the power switch modules have a filtered voltage regulator, there will be pure DC always.

Fighter posted this 14 October 2019

So this output is acceptable for PowerSource. Then next when I find some time I will start tests with ZPM.

Thanks for clarifications Vidura, I really appreciate your help...

patrick1 posted this 14 October 2019

Hello Fighter. im liking your work. very much similar too mine, although your desk is in much nicer state ;=). 

i wonder how your machine will respond,  if you try too add a bridge rectifier across the output in addition too the light globe,  and put a second light globe at the bridge output,   ....

i am also curious if you were too use ONLY an output bridge , with a series globe on the bridge input, - and perhaps an additional globe on the output again,  i dont remember this being mentioned by our members as yet.

  • Liked by
  • Fighter
  • Chris
Fighter posted this 14 October 2019

Hi Patrick, thanks but you don't want to see the mess on my desk when I'm soldering something or repairing or building something new

About bridge rectifier on output, I had an attempt at the beginning, with a bridge-rectifier I've made with two channels and Schottky diodes (you can see images of it here:

It didn't worked, ZPM destroyed one of those electrolytic capacitors, the capacitor became that hot that I wasn't able to touch it for few minutes after I stopped the experiment. I suppose normal electrolytic capacitors are not made to withstand thousands of KHz frequency and considering these spikes of 100-200V I see on scope on output sometimes that poor capacitor had no chance because it was 10000uF at just 50V. Right now I don't intend to try that but I may try again in the future if I find electrolytic capacitors capable to work at ZPM's frequency. That was a failed attempt and I will need to find the cause of the failure before trying again.

Fighter posted this 14 October 2019

Today I found about one free hour so I've experimented with ZPM and Vidura's PowerTool.

I recorded this video:


Sorry I'm not that good at presenting, while i was talking I tried to not miss anything relevant therefore the pauses/delays - while talking I was always thinking what to talk about next.

Here are some images taken with this configuration.

In these images I have the oscilloscope's probe connected directly on ZPM's output (link to larger image here):


And in these images I have the oscilloscope's probe connected on the light-bulb's pins (link to larger image here):

The ZPM's specific waveform pattern is present on output, maybe a bit different than what I saw with my MOSFET driver but the cause could be that my driver is using one of those Cree MOSFETs, I intend to use one of those on the PowerSwitch in the near future.

A difference I noticed is the light emitted by the light-bulb on output seems to be pulsing at high frequency but it's noticeable, this is not something bad it's just something different.

As it happened with my MOSFET driver too, the MOSFET's radiator on the PowerSwitch is becoming very hot in 20-30 seconds so in order to be able to have longer experiments I'll need to build a active cooling system for the PowerSwitch like I did for my MOSFET driver.

Fighter posted this 14 October 2019

Another thing I did was to put high-voltage super-fast SF28G diodes on the MOSFETs but I had some problems with this, when exploring with different frequencies few times I had sudden surges of the power drained from the DC source, the source entered in auto-protection and the 12V/55W light-bulbs from ZPM's output became extremely bright and were destroyed. So far I lost 3 light-bulbs of that type, I have only one left which is still functioning (link to larger image here):

For some reason seems ZPM doesn't like those protection diodes and at certain frequencies it suddenly starts draining full power from the DC source and smoking the 12V/55W light-bulbs.

For now I have set optimum frequency and the 12V/55W light-bulb seems to work fine but probably as soon as I will encounter one of those frequencies the same thing will happen again. If it happens probably I'll remove these protection diodes, I don't like this behavior and it's provoking damages, I don't know what's the reason but it's a fact.

And another (short) update: I definitely need to remove these diodes because while setting back my MOSFET driver and finding optimal frequency I encountered again one of those frequencies, this time the light-bulb survived, the DC source entered in auto-protection mode but one of the Cree MOSFETs from the driver (from the channel which was in use) is dead. Definitely ZPM don't like those diodes there and they are provoking too much damage, I will remove them as soon as I find some time to take care of this...

patrick1 posted this 15 October 2019

Thiis strange flashing effect, is also present on my ZPM build. - however i think its due too the output on the signal generator. - which i do not believe is damaged.  it only happens as specific frequencies, - that i have noticed. - around 100-200khz. but critially it seems tied too the output voltage of the signal generator. -  

i am using IRFP N250's ,  which taper off between about 100khz and 800khz., and they seem too appreciate about 8-11v from my identical signal generator.  - and will not operate directly from the signal generator without at least 5volts. -    but i seem too recall when you were using the identical transistors, you had your signal generator voltage turned right up.  - which is also what i have been doing....  - this prevents the flashing phenomenon so far as i can tell .yet.

Fighter posted this 15 October 2019

Yes that is correct, I'm using minimum AMP (5V),I'll try to go up with the AMP value of my signal generator. I don't know how much I should go, I don't want to damage the PowerSwitch but this could be the cause.

Also recently I removed that red tape keeping together both pieces of the Metglas and I'm using a zip tie now, that could be a cause too, maybe that zip tie is not keeping the Metglas pieces together tight enough and there is a small gap in the center of the Metglas.

I'll investigate, thanks Patrick.

Fighter posted this 15 October 2019

I also had some private messages exchange with Vidura so it seems for avoiding that light-bulb flickering I should go up with the AMP on my signal generator while monitoring the voltage on PowerSwitch's MOSFET gate. I will take care of that when I find some time, sorry about this but I still have things to learn and to accommodate myself with Vidura's PowerTool.

Edit: One of Vidura's PowerTool tutorials is about this calibration step for the signal generator, I saw this video but I didn't do this step:

I'll need to do this one-time calibration step for AMP on signal generator in order to have it working well with PowerTool and avoid that light-bulb flickering.

  • Liked by
  • Zanzal
  • Chris
Fighter posted this 17 October 2019

I replaced the dead Cree MOSFET from my MOSFET driver and removed the diodes which caused problems and provoked damages:

Unfortunately this new MOSFET had a short life, I had an experiment where I tried to use a choke-coil to smooth the voltage produced by a 10A/75mV (0.0075 Ohm) current-sensing shunt on the input.

I learned a lesson, choke-coils provoke damages if they are placed anywhere in the circuit. The choke-coil made the DC source enter in auto-protection, burned the new Cree MOSFET and also destroyed the last 12V/55W light-bulb I had:

I remember I had similar damages before when i tried to use choke-coil put on DC source's output trying to smooth the noise/spikes coming back from ZPM.

So right now I'll need to order some new 12V/55W light-bulbs and as the last Cree MOSFET is already placed in the MOSFET driver probably I should order some new Cree MOSFETs too.

Until then I still have some halogen light-bulbs, I'm gonna use 2 x 12V/35W on output.

The goal of the choke-coil experiment was a first step in the process of building a real-time current and voltage sensor on input and output without oscilloscope, using just some 10A/75mV current-sensing shunts.

As this one experiment failed I changed the approach: what about using (for the beginning) some LEDs as visual indicators of the current passing through the shunts ? Of course that's not a precise measurement but the LEDs could provide at least a clue of what's going on with the current on input and output in real-time while experimenting with different parameters on the signal generator. A next step would be, if I succeed in rectifying the voltage provided by those shunts, to place digital voltmeters measuring in real-time the voltage provided by the shunts so actually providing in real-time the accurate current measurements on input and output.

So for this first step (LEDs as visual indicators for current) I took two identical LEDs from two identical Chinese lighters where those LEDs are powered by 3 x SR626SW (1.55V) batteries (link to larger image here):

So this is the result (link to larger image here):

So for now I have some real-time visual indications on current values on input and output.

From now on I can experiment not only with frequency but also with AMP and duty-cycle on my signal generator and see what's going on with the current on input and output. In the following photos I tried changing these parameters. In reality the luminosity difference between the two LEDs is easy to observe but seems my phone is amplifying low luminosity sources and dime high luminosity sources when taking photos in dark. But still you can see these differences (link to larger image here):

Just as a note: the current-sensing shunts are identical and the LEDs are identical in luminosity, I verified that by switching the shunts and LEDs between input and output and having the same results (in terms of LEDs luminosity). So I checked that both shunts and LEDs have the same characteristics.

And I took one photo which I think is very significant, I played with parameters in order to see if I can have the LED on input off while still having the LED on output on. In this photo you can see the current on input is clearly not enough to light that LED while on output you see the LED still on and also on the left 2 x 12V/35W halogen light-bulbs still on (link to larger image here):

I am aware about the voltage differences between input and output but considering on input we have 25V and on output we have spikes between 100V and 200V in my opinion this image gives an additional clue about ZPM's over-unity.

The next goal in building this real-time measurement system is finding a way to smooth the voltage provided by the shunts so that voltage becomes DC (link to larger image here):

If I can have clean DC voltage provided by the shunts then I can use it it with some digital voltmeters I just bought so I can have accurate real-time digital display of currents on input and output.

So for now the problem is how I find a way to get DC from this but without choking-coils (link to larger image here):

I'm thinking of building bridge-rectifiers using these SF28G super-fast diodes then maybe some capacitor(s) for smoothing the voltage. I'll see how it works.

I'll post updates as I find time to continue the experiments...

Vidura posted this 17 October 2019

Hey Fighter,
Sorry about the losses of the Mosfets, they are kind expensive. If I may make a couple suggestions to avoid this damages on your custom made switch:
use MOV as surge protectors on the drain - source terminals, the max. value K420 which clamps around 1100V maybe better a K400 for more safety margin. Do not attach other components as the Mosfet on the heat dissipator, although you use insulation pads, the high frequency currents will pass thru. Be careful when placing a grounded scope probe on the center tap of the coils, as this will shift the hot end (longitudinal high potencial) to the other terminals, where the switch and Power supply are placed , and may cause interferences and failure.

I wanted to say that I love the idea with the LEDs on the current shunts, this visual indicators really make our work easier, maybe you could add a second led with oposite polarity to the shunt, which will give indication on the direction of the currents also. keep up your good work, VIDURA. 

Fighter posted this 17 October 2019

Hi Vidura,

Yes they are really expensive and unfortunately they die so fast on every failed experiment. What can I say, I expected them to be more resilient but this is the cost of researching this kind of devices we're working with...

Right now I don't know what exactly changed in device's behavior when I placed those SF28G diodes between drain and source but after that I saw appearing these dangerous frequencies which killed MOSFETs and light-bulbs, that's why I removed the diodes. They were supposed to do exactly this - protect the MOSFETs, but it was exactly the opposite.

Right now I kind of hesitate to try with MOVs between source and drain of the MOSFET as this is the last Cree MOSFET I have until I can order another batch, I'll try when I'll have some backup MOSFETs.

About attaching other components on the radiator inside the MOSFET driver, I will not, I had that voltage regulator attached before but I found a solution to power that LED on the front panel without the voltage regulator. So right now I have only one MOSFET attached to the radiator, for the other channel of the MOSFET driver the MOSFET is missing, didn't mounted one there anymore.

About the center tap of the coil, I've made measurements there with oscilloscope probes some time ago, didn't had any problems or failures, are you saying that could destroy my equipment ?..

Thanks, necessity force us to build the tools we need for this kind of devices... These LED indicators are just the beginning, I will try to get pure DC voltage from those pulses you see on the scope in my previous post, if I succeed then I'll be able to use that DC voltage with digital (and maybe analog) voltmeters which will actually show accurate real-time current measurements on input and output, I hope I'll find a way to accomplish that.

Yes, having a second LED with opposite polarity is also a nice idea and very useful, that way we can "see" the currents going in both directions through the shunts, I will try that, let's see how it works

Fighter posted this 17 October 2019

Vidura, I tried putting a LED with inverse polarity on both shunts (from input and output), the LED doesn't light up at all.

But I found something on output, when connecting LED's positive (green connector) to the negative of the other LED while the negative of the LED (white connector) is in air we have some kind of one-wire transmission, the LED light up just as powerful as the one from the shunt (link to larger image here):

This makes me think of something, should I be concerned about the electromagnetic radiation around this thing ?...

Maybe I should make a box covered with aluminum foil like a full Faraday cage to cover ZPM while it's active ? I was thinking about this sometimes before but I didn't noticed any negative effects while or after making experiments. Maybe I should reconsider this.

Zanzal posted this 17 October 2019

When searching for a mosfet - I recommend this approach - Sort by rdson - look for as low as possible - find mosfet with appropriate voltage rating and lowest gate charge (Qg), pref around 50 nC or less. Check the price per unit and if it is reasonable then it may be worth using. You may need to compromise a bit, for this reason you may find that adapting your circuit to a lower voltage mosfet may offer an advantage.

A quick explanation of my perspective:

Thermal failure is a universal failure so low rdson is typically the most important characteristic, especially if the power is greater than a couple watts.

Voltage rating at first glance may seem important, but I've found that some circuits can easily exceed the voltage rating for a mosfet and the mosfet may clamp voltage at its avalanche voltage. So a mosfet rated at 60V may not fail immediately but could just avalanche (causing losses and extra heating). You can try to design a snubber circuit to clamp the voltage spikes to avoid needing to use a high voltage mosfet. So while operating voltage is an important consideration, there might be better ways to handle spikes than just using a HV mosfet. You can experiment with adding snubbing circuits and see if it negatively impacts AU device performance. I can't tell you if this is appropriate for your device.

Frequency range will determine if the gate charge along with the on/off and rise/fall characteristics are deal killers. For low frequencies, these details may be relatively unimportant, but once you start to go above a couple kHz these things may begin to make huge differences. Keep in mind, high rdson compounds any problems you will have here as the mosfet may spend less time fully on or fully off and rdson only tells you the lowest nominal resistance in the fully on state.

Anyway, these are just opinions from a non-expert. I am curious to know if you find mosfets that work well for your application. When you do, please share the part number.

Fighter posted this 17 October 2019

Hi Zanzal, I'm also a non-expert but I find it's a lot to learn from your post.

C2M0160120D made by Cree/WolfSpeed seems to be the fastest HV MOSFETs on the market for now and, even if they are expensive, during my research (after Vidura told me about them) I didn't found information/opinions about more performant ones.

But using your criteria of selecting who knows, maybe I'll find others better in terms of matching ZPM's circuit, I'll keep these criteria in mind and I'll search.

Of course, if I'm gonna change the MOSFETs model I'll post updates here.


Vidura posted this 17 October 2019

Hey Fighter, I am using the Cree MOSFET s in all my latest experiences, and have not lost one till the moment, the performance is really excellent and the RDS on the lowest in that voltage range. But I always use MOV , it's a must, also the zener diodes for gate protection are strongly recommended, and I didn't note any negative side effects for using this protections. Regarding the effect of single wire transfer it is obvious, you have high frequency longitudinal oscillations, this is why the current not always behave like expected, and the reversal LEDs on the shunts might be off most of the time, but they will indicate when a reversal current flows, which might occur only with very specific parameters. I hope this helps and explain some things. Vidura

Jagau posted this 18 October 2019

Yes Vidura 

 zener diodes for gate protection are strongly recommended and a totem pole activation for the gate is also a must to preserve the shape waveform that activates the gate.

Hi fighter
I just received my metglass and I notice they have a very high saturation chest 1.56 Tesla compared a ferrite it is well better.
Nice experiment to do.

It is a hitachi amcc320


Fighter posted this 18 October 2019

@Vidura I know about their performance, the only issue I had was about those dangerous frequencies after I put these protection diodes, I don't know why it happens but seems ZPM don't like having these diodes on MOSFETs.

For sure I intend to give MOVs a try when I'll have some backup MOSFETs for replacement just in case MOVs create the same behavior.

About the single-wire transmission that was fun to see but still after I saw that phenomenon I wonder how strong and how extended is the electromagnetic field around ZPM and if I should take measures to shield it during experiments. What would be your advice about this, what would you do ?

About the additional LED indicators, I understand they may become active in certain situations so I'll certainly add them to the shunts, even if they don't light up all the time as I kind of expected your idea is neat and very helpful, thanks !

@Jagau That's a very nice core ! I suppose you will try a ZPM replication ? If so welcome to the club, I'm glad and I'm curious to see how your experiments work, please share them...

Yes, saturating that core would be kind of difficult, that's one of the special characteristics about Metglas. Being made by Hitachi make me sure your core have exactly the same physical characteristics (related to alloy composition and production process) just like mine because those guys have a strict production and quality control process.

I gathered two quotes from the old ZPM thread which I think could be useful in understanding more about Metglas:

I'm putting links about AMCC-200 technical specs here:

Hitachi - AMCC Cores Technical

Hitachi - AMCC Cores Presentation.

I recommend checking how Metglas cores are produced and read about their special characteristics so you know what you're dealing with; remember Metglas was something produced in 70s for military and space technologies.

Actually this is Metglas - a iron core. Its composition is 85-95% iron, 1-5% boron and 5-10% silicon as specified here. But what's special about this core is it's made of foils created by melting the compositions under high heat then cooling it down so fast (almost instantly) that the Metglas alloy don't have time to form a crystalline atomic structure when it become solid again like any metal does when being melt then being cool down normally. That's why the metals obtained this way are also called amorphous (their atomic structure is not crystalline like the usual metals).

You'll notice the differences about AMCC compared to other cores, enjoy your experiments and can't wait to see it at work !

Vidura posted this 18 October 2019

Hey Fighter, Note that the SF28 diodes are not TVS devices, and thus can not be considered as protection, if connected in parallel they would only bypass the internal body diode of the MOSFET. But as the internal diode of the cree devices is already very fast, it don't make much sense to place it in that way. I think there was a suggestion to put a diode in SERIES with the MOSFET , to prevent the effect of bus pumping, or current being sent to the supply, which would make more sense. But maybe it is better to test this things with a cheaper MOSFET first, or an IGBT, for a short time would be okay. Regarding the radiation (longitudinal) , there are several reports of being beneficial for health, but this might depend on the frequency also, actually there exists a device based on Tesla oscillators used for cure several diseases, development of a Russian investigator. I have many years worked with Tesla coils and i think i didn't note any negative side effects. But if you like you can anyway use a shielding, or keep some distance for any case. The only thing you should take care is that at the hot end of the coils you might get rf burning if you touch with bare fingers, this effect is stronger with lower frequency, so the core devices can make smoke your fingers in an instant 😀. Vidura.

Fighter posted this 20 October 2019

@Vidura, you mean to put the SF28 between DC source and MOSFET driver on the positive wire ? I remember I tried that with another diode, I admit it was not a super-fast diode. I can try this scenario too, thanks.

About touching the ends of the coils, I was thinking to get some protection gloves for electricity, for safety I'll buy some. Considering there are spikes of 100-200Vpp, at what frequency could that become dangerous ? Right now I don't go below 80 KHz, usually I'm in 100-1770 KHz range.

Posting updates...

I've build a new version of current indicators, this time I added a second LED with inverted polarity so they will show currents going in both directions through the shunts (link to larger image here):

This is a test with normal polarity (link to larger image here):

And this is a test with inverted polarity (link to larger image here):

For some reason the LEDs for inverted polarity (positioned lower) don't have the same luminosity even if the current should be the same like with normal polarity. All circuit is made using normal 0.8mm wire (just like the one ZPM's coils are built with) enclosed in plastic. I verified the solderings and found no issue.

But for now it's not that important, I'm interested more in what the upper LEDs are showing.

As a note, the fact that a LED indicator is off doesn't mean there is no current, it means there is not enough current to produce enough voltage to power-up the LED.

When I find some time again I will try to rectify the pulsed voltage produced by the shunts, meaning this:

I intend to build a bridge-rectifier using some super-fast SF28 diodes and then some electrolytic capacitor maybe (choosing the right one to smooth that waveform will be difficult, I'll need to do a lot of tests).

If I'll succeed then having pure DC produced by the shunts would mean I can use digital voltmeters to get real-time accurate reads for currents passing through shunts.

Zanzal posted this 20 October 2019

About touching the ends of the coils, I was thinking to get some protection gloves for electricity, for safety I'll buy some. Considering there are spikes of 100-200Vpp, at what frequency could that become dangerous ? Right now I don't go below 80 KHz, usually I'm in 100-1770 KHz range.

There is a certain frequency threshold above which your heart muscles cannot respond quickly enough and so alternating currents above that frequency won't stop your heart. Not sure where this frequency is exactly, but I vaguely recall it is between 2k and 4k. Above that frequency you should be fine. Below that, and you are at risk any time you are grounded as your blood vessels may act like conductors and your finger tips would be capacitively connecting you to the circuit. High currents should cause burns at any frequency but these would likely be over your skin at very high frequencies. Not sure what the current threshold is for that.

I worry more about my family when I am doing these experiments. Little ones especially have a way of getting into trouble grabbing things they shouldn't. A family member checking on you after you make a mistake may suffer the same. Precautions are good. Take note if any family member or guest has a pacemaker. Stay safe.

Vidura posted this 20 October 2019

@ Fighter
The Fast diode would be placed between the ZPM and the mosfet better, I am not sure if the current capability of the SF 28 is enough.
About the led indicators of course there is a threshold voltage , it will only lite with considerable current spikes, I have thought about this.
@ Zanzal
regarding the current, generally it is dangerous at low frequency or DC specially, above 30Khz it is considered RF , and the skin effect begins to predominate, so the current will flow on the surface, the RF burning is superficial normally , anyway for safty keep the kids at distance when experimenting with this kind of devices.


Jagau posted this 24 October 2019

Hi fighter
My C core mestglass is progressing well.
I build the insulators in acetate and I prepare to mount the 2 coils.

I have a little technical question to ask you, so with about 150 laps for the primary you have an inductance of 101.4 mh and for the secondary 300 laps with 365 mh on an AMCC200.

What I do not understand if i check inductance factor and with the specifications of the manufacturer (photo below) we have 1.26 uh / N2 for ungap of 0.5 mm.


If I remove the gap we could say that it is about 2 uh / n2
so with 101 mh it would take 200 laps for the primary.
This is the question I am asking.

JL Naudin used an AMCC320 which is like mine and is a little bigger than the 220 and with 24mh he had 100 laps.
Could you tell me if I have the right number of turns?
It does not matter if you do not know I will find.
thank you


Fighter posted this 24 October 2019

Hi Jagau,

This is what my LCR meter shown after I built the coils:

It has two bucking coils (marked the with "L" and "R") winded in a way that when they're connected in series and powered on, their magnetic fields are opposing.

L coil has 101.4mH inductance and approx. 150 turns made of 0.8mm standard copper wire for coils. These are the complete characteristics I measured with my LCR meter (link to larger image here):

R coil has 365.5mH inductance and approx. 300 turns made of 0.8mm standard copper wire for coils. These are the complete characteristics I measured with my LCR meter (link to larger image here):

I ran auto-calibration before making measurements just to make sure. About the number of turns, the counting error is maximum +/- 10 turns (because I've made the coils manually), so yes it's 150/300 turns ratio with maximum +/-10 turns counting error. The gap is displayed in the latest photos, it should be 0mm because I stretched as much as possible that zip tie keeping the pieces together. I would say to follow one of the coil ratios reccomended by YoElMiCrO, ignoring the turns counting error mentioned I consider my ZPM configuration with 150/200 ratio is matching one of them. It's possible that the inductances shown by my LCR meter are because I've made the coils manually and the turns are not ordered as they would be if using a coiling machine, I'm sure the inductance is different between coils made manually and coils made with a mchine.

Jagau posted this 24 October 2019

Hi fighter
thank you for answering,

I will do my 2 coils and see how they behave


Chris posted this 24 October 2019

My Friends,

If I may interject here for a minute.

Fighters Coils are: 1 : 2 basically, as YoElMiCrO kindly pointed out:


N2 = sqrt(n) * N1 = sqrt(0.25) * 300 = 150 


Which is Fighters Coil Combination!


We have a situation here, where Fighter is using 25% Duty Cycle on the Input:


What occurs during all that Off Time?

No one is looking at this???

This 25% Duty Cycle is the same as the 0.25, the Quarter Wave. 25% Duty Cycle is a Quarter Wave, they are the same thing! This is important, please, something people need to start thinking about.


cd_sharp posted this 25 October 2019

The sqrt(0.25) is a measure of the time constants ( t ) difference between the coils. 5 time constants is a number which defines the time needed for the current to reach max value which is equivalent to max magnetic field.

Chris posted this 25 October 2019

Hey CD,

The Coils reach maximum Current ( I ) when five time constants have passed.

The equation is: T = L / R

  • T is five Time Constants and is measured in seconds
  • L is the Inductance and is measured in Henrys
  • R is the total circuit Resistance and is measured in Ohms.


So no, not really correct, as the coils become resonant, yes we do get maximum Amplitude, but this cant always be assumed this is maximum Current ( I ) for the applied Voltage ( V ).

Resonance, especially this type of Resonance, is quite different. Magnetic Resonance.


cd_sharp posted this 25 October 2019

Hey, man

During the off-time, assuming both coils are charged, they will decay. L2 will try to decay twice faster than L1. That is a changing magnetic field ahead of the changing magnetic field of L1. This results in delayed EM induction in L1, as we named it counter-reaction and it's adding up to the decaying field of L1.

Please correct me if I'm wrong.

Chris posted this 25 October 2019

Hey CD,

Ah ha, yes you're right!

The Interactions between the Coils:


In Fighters ZPM we see a very interesting Spike:


Marked in Red is the process I describe above in the video. The Coils are like a spring, and Act and React together in Magnetic Resonance. This is an Electromagnetic Induction dance, or disco. Electrons are getting some moves on and the Disco is going off...

This is excess Electromagnetic Induction after the fact! After the Input has excited the Coils into this Resonance.


cd_sharp posted this 25 October 2019

I wish to go further with my imagination. While L1 is discharging slowly, the voltage over L1 is not what it's supposed to be if L1 was discharging alone. The voltage induced by the L2 collapsing adds up to the voltage over the ends of L1 and it causes a much faster discharge (almost instant) than normally. We have increased V, we obtained increased I.

This almost instant collapsing of L1 causes EM induction in L2 and so they start to dance as you said. This device uses the Reduced Impedance Effect so visibly. I look forward to playing, I'm waiting for some hardware..

Fighter posted this 25 October 2019

Talking about hardware, some updates on my side, I just received the shipment with 6x12V/55W light-bulbs:

... and I have a shipment with 6 x Cree MOSFETs which is still on the way. This time I bought the MOSFETs from TME, they are more expensive comparing to Digi-Key but as TME is in EU I'll avoid the tax for shipments coming from outside of EU which is signifiant.

So my backup stock will be restored soon and I'll have replacements available again for experimenting.

Jagau posted this 31 October 2019

hey fighter

slowly but surely
My acrylic supports are made
and I found (maybe a way to make coils more symmetrical)
I tried to make this coil winder


Fighter posted this 01 November 2019

Hi guys, unfortunately I need to take a break from researching because I need to relocate. I will find another place to rent (the owner of this one wants to sell it), pack all my stuff and transport it there. As you can imagine this will take a while and during this process I will not be able to do experiments. I will resume the research as soon as it's possible. Sorry about this but this is how life is sometimes.

Please continue your research, I'll still be here but I will not be able to present anything new until I resolve this problem.

Chris posted this 01 November 2019


Good luck with your move, I hope it all goes smoothly!


Members Online:
Since Nov 27 2018
Your Support:

More than anything else, your contributions to this forum are most important! We are trying to actively get all visitors involved, but we do only have a few main contributors, which are very much appreciated! If you would like to see more pages with more detailed experiments and answers, perhaps a contribution of another type maybe possible:

Donate (PayPal)

The content I am sharing is not only unique, but is changing the world as we know it! Please Support Us!

Donate (Patreon)

Thank You So Much!

Weeks High Earners:
The great Nikola Tesla:

Ere many generations pass, our machinery will be driven by a power obtainable at any point of the universe. This idea is not novel. Men have been led to it long ago go by instinct or reason. It has been expressed in many ways, and in many places, in the history of old and new. We find it in the delightful myth of Antheus, who drives power from the earth; we find it among the subtle speculations of one of your splendid mathematicians, and in many hints and statements of thinkers of the present time. Throughout space there is energy. Is this energy static or kinetic? If static, our hopes are in vain; if kinetic - and this we know it is for certain - then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature.

Experiments With Alternate Currents Of High Potential And High Frequency (February 1892).