Schematic Verification

edit: please see updated schematic in later post. I removed an Earth ground symbol and added a ground line from the Arduino to the mosfests source.

2nd edit: added 3rd version of schematic to include the switch I actually have on the mains.

A few months ago I knew nothing about circuits, but over the course of building a few different PEMF devices I think I now know just enough to be dangerous.

(A Pulsed Electromagnetic Field device is essentially a signal generator which will switch "high" currents. Mine switch in the 1hz-40hz range. A short video showing the usage of a previous 9v model can be seen here

My latest model is much more powerful that anything else I've done and I'm unsure if I designed the circuit correctly. I'd appreciate and suggestions or feedback. And if you think it's weird that I'm using so many mosfets, remember that recently I had no idea what I was doing so this kinda evolved naturally from parts I had and knew. Also, the faster switching the better, and I believe the smaller the fet the faster the switching(?).

Parts list:

  1. Nano
  2. 18v 15 amp power supply (rated 24v 15 amps)
  3. IRLB8721 Mosfets (x7)
  4. 24awg copper coils (x7)
  5. a heatsink

IRLB8721 Mosfet specs:

Some other relevant info:

  1. The mosfets will be the only thing that the Nano controls so each will share access to the 200mA total, making it 28.5mA a piece.

  2. The Nano pins will activate via port manipulation so all mosfet gates will open simultaneously.

  3. Each coil has a resistance of 10.3 ohms.

  4. 18v / 10.3R = 1.74 amps per coil or 31.4 watts. 1.74 amps * 7 coils = 12.18 amps total.

My main question is: do I need to ground the Nano to the mosfets' source even if they share the same original power supply, and if so do I need a capacitor? I've seen them in certain schematics.

Some more pics!

Manufacturer's page of IRLB8721:

Please don't use the "earth ground" symbol for a normal 0V or GND.

The mosfet has a gate capacitance of about 1nF. That means the Arduino Nano sends peaks of 40mA to charge and discharge the gate of the mosfet. You might want to limit that with a resistor, but that will make it slower. Well, that is why there are mosfet-gate-driver chips.

Which port do you use for the 7 signals ? PORTD ? Are you giving up the Serial communication to the computer ?

You need to connect the Arduino GND to the GND of the circuit with the mosfets. Watch out for ground loops.

How much radio signals will it generate ? Is that legal ?

You are showing the Gates jumpered to the Sources?
The wide prong of the 5V module goes to Neutral ("AC-"). Neutral is connected to Earth back at the service, but it's not the same thing.
[PE: O-I-C, that G-S jumper is your way of making a 10k resistor.]

@ Koepel

I've updated the schematic.

Slowing the mosfets is the last thing I want to do so unless absolutely necessary I won't do that. Its the magnetic flux which charges your body on a cellular level so the faster the better. I think going with a power mosfet or 2 and a gate driver would have been the better option, but whatever, as long as it works. Those mosfets have been fine so far without resistors.

Yes PORTD, I'm going to sacrifice TX on D0 I guess. Seems better than RX.

How do I watch out for ground loops? A diode? Capacitor?

In regards to radio signals... dude I have no idea. I hope its legal. These things sell for thousands of dollars and I haven't ever seen anything about radio concerns.

@ runaway_pancake

I tried making the 10k resistor part a little more clear. Might have made it worse

If your Arduino resets, then it could be because it has to charge all those gates with a current peak.

For high speed, you could add capacitors near the coils (one side to the 18V side of the coil, the other side to the Source of the mosfet).
The + and - wires of the 18V power supply should be thick wires. I suggest wires that are used for the mains and are for 16A or so.

Turning off the coils gives a peak of hundreds of volts. Adding a flyback diode will turn the coil off slower.
ElectroBOOM about relais (coils) and the influence of the flyback diode:

You watch out for ground loops by avoiding them.
Suppose there is a path from the mosfet to the Arduino and then to the 18V power supply. Then the ground current can decide to use the small wire to the Arduino board instead of the thick wire directly to the power supply.
Ground currents have a mind of their own, they are attracted to inductively coupled opposite currents.

Yes :face_with_raised_eyebrow:

If your Arduino resets, then it could be because it has to charge all those gates with a current peak.

Ok thanks I'll definitely be on the lookout for that. I have the resistors if need be.

Suppose there is a path from the mosfet to the Arduino and then to the 18V power supply. Then the ground current can decide to use the small wire to the Arduino board instead of the thick wire directly to the power supply.

That seems to be what I have now

Every country has different electromagnetic radiation standards, but in different regions they are "harmonized" meaning they are similar. They are largely based on common sense and engineering knowledge about how EM waves are generated and propagated, also on almost 100 years of field experience.

Your device has multiple non-shielded and open coils, through which you launch powerful, rapidly changing currents. It is almost certainly illegal, and will definitely disrupt communication devices, and possibly other appliances in the vicinity. It may even be a human health hazard, as it appears from your wording that you are using it as some kind of medical treatment device.

You are also involved with some touchy, possibly dangerous electronics without much real understanding of the risks. Your schematic shows a connection to a wall socket that would keep any electrician awake all night.

Probably, you don't really care about any of this. I post this for other people who might attempt to copy you.

It's surprising how little you need to know to be dangerous with electronics!

Do you mean to attach the system ground to one prong of what looks to be a regular USB cell phone charger wall wart?

I don't think that connection exists within the wall wart, that is neither prong is wired to ground on the USB connector.

Most of the ones I have (all?) don't have polarized plugs or ground third legs.


@ Koepel

ElectroBOOM about relais (coils) and the influence of the flyback diode:

That was a good video. 100v spike when turning off the coils seems to be too much to ignore... Are 2 zener diodes in series the best option? (open question to anyone reading)

@ alto777

Do you mean to attach the system ground to one prong of what looks to be a regular USB cell phone charger wall wart?

Heh... no. I'm omitting a little bit of the real schematic on the mains part. In reality I have a master switch to turn on both the PS and 5v converter. I feel like that part is pretty straight forward though so I didn't want to complicate my schematic.

Please overlook that detail :grimacing:

@ aarg

I appreciate your response. It makes me wonder about all the other devices that are sold because mine is probably mid range in terms of power. I do plan on making a faraday cage around the walls of my bedroom, eventually. Also, for what it's worth, the PS has it's own current limiter and will shut itself off. So should things go sideways I think my circuit will fail before I flip a breaker.

I can do that. Or you could fix your schematic. Glossing over or misrepresenting circuitry, especially around mains, is not a real good idea.

Just sayin'.


Just sayin'

I kinda knew someone would say it. It was my bad for not just doing it from the start

In regards to the lingering charge that would exist in my coils when I close the gates, what if I just put the mosfets in front of the coils?

edit: I guess to do that I’d need a voltage in the mid 20 range to overcome the 18v from the PS

Regarding the circuit:

You will need RC snubbers on each of your FETs (MosFet). The coils as you have shown them will act as a transformer and couple each coil to the next. When the FETs go off there will be an inductive spike that has the potential to destroy the FETs.

Your device is likely illegal and possibly annoying to your neighbors. The 1Hz to 40Hz is not the issue. The rise time of the on and off transitions create harmonics well into the Megahertz. A rough extimate of the frequency where the harmonics are down to 50% is. Freq = 0.35 / risetime (freq in Hz, risetime in seconds).

Will have no effect.

I know it’s not what is intended… but why are the two outer pins of each transistor sorted with that black line ?

You could have drawn the 10k resistors vertically between the G pins, and the horizontal S buss…

What about a zener diode to ground? like 80v?
The harmonics aspect is really interesting. I think I'll drop my PS voltage down a little.

I didn't do it how you suggest because I didn't think of it till too late. I stupidly drew them like that because it's how I solder them

Your IRLB8721 has a max voltage of 30V (kinda low being used with 24v) . You could probably get by with a 28 to 30V zener diode.

Folks will point out a 28V or 30V zener will not limit the voltage to 30V (see zener voltage vs current curves). However I know IR was very conservative in their ratings so I'm guessing this MosFet can go safely to 50V and likely survive 80 to 100v. Neither is recommended but you have the parts so give it a try.

1 Like

I was planning on running at 18v before all the "this things illegal" and "your neighbors will hate you" stuff so now I think I'll start at maybe 15v.

Thanks for the recommendation, and thanks to @Koepel for the video he posted.

I feel like I've probably arrived at a solution now, apart from the illegal aspect. But again, I'll be shielding the room it runs in sooner than later

Actually one more question. Will the diodes need to handle the original wattage of what was running through the coil, or is there anything special to know in this case? 30w zener diodes are expensive. 1w diodes are about $6 for 100, but putting 30 of them in parallel per mosfet would be ridiculous.

Why 7 coils not 1 ?

They're likely sold as is because the sellers & manufacturers either don't know or don't care (I suspect the latter) if what they're doing is legal. This is because the application is deep in snake-oil, alt-medicine bulls**t territory, as signified by:

Which is pseudo-scientific nonsense*. The whole device in my opinion is a waste of materials and as pointed out a major nuisance to anything relying on radio transmission in the neighborhood. It should not be built, period.

But to each their own. Let's assume you want to go ahead with this sharade, then my concern would be that the collapse of the magnetic field in those coils would send back a spike to the power supply that potentially fries it. So it's not just the switching FETs that need some thought, but also the PSU itself.

Furthermore, I'm mystified how a 24V PSU is supposed to output 18V or 15V, unless that's a little artistic freedom in the schematic and a different PSU is actually used depending on what voltage is required.

I don't see your gate drive setup yet. @Koepel pointed out right away that you need gate drivers to make that poor little Nano switch a bunch of heavy-handed MOSFETs. Have you identified a suitable part yet, or drafted a schematic of the required totem-pole drivers?

(*) Unless this is about transcranial electromagnetic stimulation (TEMS), which is indeed promising - but also something YOU SHOULD NOT BE TRYING AT HOME, AT ALL!!!