Partnered Output Coils - Free Energy

[TinselKoala]

@Vortex1: How much does the effect depend on the speed of the diode? Is that part number MUR8320? I can't quite read it on the image, and I can't find a datasheet for MUR8320. Can you post the data that the sim uses for that diode? Thanks in advance...

[TinselKoala]

Ah, is it MURS320? That's a fast diode indeed, 25 ns, but surface mount! Too small for my clumsy fingers to manage. I need simulated fingers!

ETA: In the second schematic the MUR460 is similar in speed to the UF5402 that I used in the actual EMJ-Meyer circuit, but a little higher power rating.


Thanks for your work on this, it's nice to see sims agreeing with reality!

[MarkE]

The large spike / ringing in the EMJ setup is from the leakage inductance of the " partnered coils". I can show you the same spikes , up to 4kV (with a few tweaks) in a simple circuit simulation or even one that does not use partnered coils but substitutes a single inductor in the diode clamping output network.

For simplicity, here is a simple circuit (without tweaks) with just the ten ohm resistor and a 100uH coil L4 which simulates the resistor inductance.

No "partnered coils" in this one and no coupling between L1 and L4. The HV spikes are almost 2.7kV. measured at the collector. Drive frequency is 1.73 kHz and drive voltage from simulated FG (V2) is 5 volts.

I also have EMJ's actual circuit with the "partnered coils" simulated if anyone is interested.

Sure, because ignoring leakage inductance, you've got the equivalent of the load branch including the output diode across the primary.  The leakage inductance yields the big spike.  The "partnered" coils add some paratici capacitance which contributes to the ringing.

[Vortex1]

@Vortex1: How much does the effect depend on the speed of the diode? Is that part number MUR8320? I can't quite read it on the image, and I can't find a datasheet for MUR8320. Can you post the data that the sim uses for that diode? Thanks in advance...

Earlier post was an MURS320. The last post was an MUR460. I tried a few diodes without much effect. I need to update my library as I'm using just the parts that came with and older version of LTSpice.

I'm sure some of the guys here are much better at sim programs than I, was just trying to make a small point.

Biggest effect seems to be the leakage inductance e.g with a magnetic coupling factor of 95% the program probably inserts the 5% as uncoupled inductance or what we call leakage inductance.  The leakage inductance, besides appearing between the primary and both "partnered coils", also appears between between the partnered coils, so there is not a perfect bucking null of inductance.

I'll see if I can dig the part out, but it is in the standard LTSpice library if you have that.

[Groundloop]

.