Benchtop Pulsed EM Rig: 3 lb Jump Coil Inductance & Driver Arch

[2ndchair]

Benchtop Pulsed EM Rig: Optimizing coil inductance (uH) & driver architecture for a 3 lb pulse jump

Hi everyone,

I'm developing a benchtop kinetic positioning prototype that uses pulsed air-core electromagnets to actively levitate and pulse-kick a 3 lb (1.36 kg) aluminum/permanent-magnet mass roughly 1 to 2 inches off a non-ferrous deck.

The core goal is fast pulse rise/decay so the coil doesn't drag on the mass or dump massive heat into the deck surface.

Current Planned Specs & Hardware:
- Payload: ~3 lbs (1.36 kg) target mass with embedded N52 Halbach array.
- Power Bus: 24V DC main supply -> Constant-Current Buck Charger -> 4S Supercap / Pulse Cap bank (~10.8V max).
- Coils: Custom flat/pancake air-core coils (looking at heavy gauge, low turn count to keep inductance down).
- Control Loop: ESP32-S3 running a 1 kHz loop with analog inductive proximity / optical height feedback.

Where I need your sanity-check & guidance:

1. Coil Inductance (L) vs. Pulse Decay: To achieve a clean 15-20 ms kick without inductive lag, what inductance range (uH) should I be aiming for? My back-of-the-napkin target was ~200 uH, but I'd love input on optimizing turn count vs. peak current surge (I^2R losses).

2. Driver Protection: For switching high instantaneous currents into a low-inductance load without blowing up gate drivers or frying MOSFETs/H-bridges, what specific flyback/snubber topologies or switching setups (e.g., dedicated SCRs for pulse jumps vs. MOSFETs for baseline hover) would you recommend?

3. Capacitor Discharge Architecture: Any recommended protection practices when tying a buck-regulated charge controller directly to high-current pulse discharge caps?

Appreciate any advice, circuit topology suggestions, or math sanity-checks!

[moffy]

You might want to start with the magnetic field calculations, because they will be what is doing the levitating and overcoming gravity. They will also tell you the field strength necessary and from that you can try and work out a coil size/inductance/current that is practical. A magnetic Finite Element Method(FEM) might be handy, I know there are some freebies but I've never used one.

[jwet]

I agree with Moffy- figure out the magnetic circuit first- driving an inductive load hard is not difficult- basically it requires voltage to overcome the inductors basic behavior= V/L=di/dt.  The magnetics calculation will tell you what is required and the inductance will fall out of that.

Also, don't forget that its not just ferrous materials that will react to big fast puises of current and magnetic fields.  Any conductor (Aluminum, eg) will see some huge recirculating currents in a strong field.  "Eddy Currents" are why swinging a permanent magnet on a pendulum up to a piece of Al will create a braking effect before it can touch the Al.

Have fun.

[2ndchair]

Thanks Moffy and JWET — this is helpful.

JWET — that V/L = di/dt relationship makes total sense for why we need high voltage on the cap bank for the current rise time (di/dt) without running into inductive lag during the 15-20 ms window. Also appreciate the reminder on Eddy currents and aluminum. i had forgotten about that. (now i'm worried about eddy's forcing my lift laterally.) i'm going to keep the deck non-conductive (probably some composite) right around the coil array to avoid back-EMF/damping issues.

Moffy — i like the idea of starting with the magnetic circuit first and letting the inductance "fall out" of the force calculation. i'm going to find a FEMM calculator right after this.

Quick follow-up:
1. When driving pancake coils for fast pulses, do you generally prefer low turn-count with thicker wire (12-14 AWG?) to keep resistance and inductance minimal, or maybe multi-strand Litz wire to lower the skin effect during fast discharges?

Thanks for steering me in the right direction!

[moffy]

Hard to tell, it depends on the amount of current you need to put through the coil, induction cooktops tend to be a Litz pancake coil, but you might need something beefier and use an insulated rectangular cross section copper winding. Pity you can't use some permanent magnets as well as the coil to repel the weight.

[jwet]

There was a demo gadget in one of the labs at my old school (Univ of Illinois) in the late 70's that they would bring out for open houses to amaze freshmen and civilians.  It would launch heavy metal rings very accurately over 20 feet, could consistently shoot them 25 feet into a paint can.  It was an old demo probably from the the 1940's, it used very heavy wire- like 6 Gauge IIRC.  In the same lab, there was an old demo of a levitating frying pan that would get hot enough to boil water in 30 seconds.  It used some monster coil.  The memories are very vague 50 years on but thought provoking.  The lab was a very dusty place in the basement of the Old Everitt EE Bldg- they're were big motors and generators bolted to the floor- ala Frankenstein's lab.   I'll bet you its still in use there, they were memorable demos.  Whatever they taught in that lab is what you need to know- the Electric Power guys took that course.  You could probably figure out the course and the instructor with some googling and a course catalog.  Its likely that the course notes are posted in this day and age and maybe even some you tube videos.  Have fun, be careful.
NB- couldn't stand it- googled around my self, and I think the course # is EE330, a senior level elective for Power Engineers.  My memory failed me, the crazy lab is on the fourth floor of the Everitt Bldg- its is very dusty though. The whole place was built like a brick s'house- really stout, the fourth floor was as stout as any basement. I'll leave it to you to find the class notes, etc.

[2ndchair]

moffy and jwet,

thanks for sharing that UIUC lab story, jwet — stuff like that really makes me wish i'd have gone into engineering. i bet the EE that built that could build my entire test rig in 10 minutes.

gemini says that classic demo is called an "Elihu Thomson Electromagnetic Ring Launcher" or a "Thomson Ring Flinger". (i'm getting all the help I can get since high-power electromagnetics definitely isn't in my ballpark)

that 1940s setup using 6 AWG wire and induced eddy currents is spot-on to what we’re aiming for. except we’re pairing those pulsed fields with N52s embedded in the puck to give it even more launch punch per amp.

moffy — i just now got my first full 2D FEA simulation running properly in FEMM! After dialing in the boundary conditions and running it in axisymmetric mode with a 100A pulse through an air-coil, it modeled a direct upward launch force of ***57.7 N (~13 lbs of peak force)*** against my test puck.

after i tweak it a bit, the next step is finding a solid EE to help build a quick, dirty benchtop POC. even though i'd like to jump straight to the end, i'm guessing no solid EE would want to build the big rig without testing the real-world pulse timing, thermal load, and dynamic stability first.

"why did that thing just fly sideways through the wall?"

i really appreciate the help and advice from both of you!

[jwet]

I'll follow this thread if you'll thank me with the thanks button.  Its my perverse OCD way of measuring my impact here.

[Dominiz78]

Hey, I actually posted a related project of yours today. Feel free to check out my blog post, I’d love to exchange ideas about your project sometime. Best regards, Dominic https://www.eevblog.com/forum/projects/working-magnetic-levitation-system-next-project-a-clock-with-a-levitating-hand/

[moffy]

Just an FYI secure the coil down, every action has an equal and opposite reaction.

[2ndchair]

jwet — i smashed that thanks button. i didn't even know that was a thing until you said it. appreciate you sticking around.

dominiz78 — thanks for dropping by. that levitating clock hand project looks amazing! it's great to connect with other folks messing around with active magnetic vector control and field dynamics.

moffy — excellent point on the reaction force. i'll be sure to bolt the coil down to a heavy wooden baseplate before firing any cap pulses.

appreciated!