Source Impedance

Jess

what do you mean:

"when the oscilloscope probe was connected across the capacitor."

are you using a differential probe?

you can't connect a scope probe that has one side grounded "across" a component unless that component also has one side grounded....

Mark

Yes, I am using diferential probe. jess

The simulation files contain models that are encrypted with Pspice. They can't be made to work in LTspice, which doesn't support Pspice encryption.

Which average current?

I can only imagine that you are referring to the series resistance of the tuned circuit. As I said, I've ignored capacitor losses, all the resistance is in the inductors, Right click on each inductor to open its parameters, where you'll see it, 0.560812 ohms each inductor, calculated from my finite element analysis model.

If I knew the exact dimensions of your Helmholz coil pair, I could get nearer.

Hi,

I meant average current for the power supply. If 12A peak to peak is flowing through the circuit than how much current 12 V power supply is supplying if the duty cycle is 2.2usec or 5usec. I still do not know how did you calculate the 11watts power dissipiation acorss each transistor.

jess

"If I knew the exact dimensions of your Helmholz coil pair, I could get nearer. "

Both coils are rectangle coils. Length = 18 inches, width = 9 inches and connected in series. The distance between the coils is 8 inches. I know the two coils are not separated correctly but this is the setup I have right now. The wire guage is 38. turn / coil = 7. Number of strands per coil = 1050.

Can I use the finite element analysis model, if yes than how? How do you construct the coils in your last LT Spice model that you posted. Plus the voltage across two inductors in your model were out of phase. Are they not suppose to be in phase?

jess

Let's get the terminology clear. In general parlance, "length" of a coil means length along its magnetic axis. For a circular coil, we have internal and external diameter, for a rectangular coil we have height and width, if it's square, we sometimes say "side".

Is that 38 AWG or SWG? They're different. You talk about strands. Is this 150x38 gauge litz? It looks like it might be, 7*150=1050.

Get and install FEMM (Finite Element Magnetics Modeling). On the 'Net, it's a steep learning curve.

It's the same model I've used all along.

Axisymmetric model of two "pancake" coils, internal diameter 2", external diameter 3", length 0.25", loosely filled with 25 turns of 22AWG magnet wire, separated axially by 4" between outer ends.

Number of turns and wire gauge fudged to get somewhere near your stated inductance and Q.

Mutual inductance derived from integrating the field from one coil only, over the cross sectional area where the other coil would be.

That was for the circuit in your drawing, with capacitor across one coil only. Forget about it.

The voltages across the two inductors in the model I posted are in phase. They have to be, they're in simple series.

Plot the power supply current, and integrate it, See below for how

Varying the duty cycle will amplitude modulate the coil current.

I still do not know

That was for your silly series-parallel circuit, one coil in parallel with a capacitor, with the other coil in series. Per your drawing.

I thought we'd put that to bed, ignore it:-)

Alt-Ctrl-Left click on each device. That gives you a plot of the dissipation in each device. Then, in the plot window, Ctrl-left click on each trace's legend. That will calculate and display the integral.

The last simulation I posted has .meas cards for dissipation in each device. Run the simulation, then view the Spice error log to see the results. (M1dis, M2diss, etc.).

RTFM.

Yes it is Litz wire. I will get back to you > It's the same model I've used all along.

Did you create the coils in LTSpice? If yes than how?

jess

You meant 40nsec or 40usec.

I tried it for the source. The average power dissipiation came out to be -102.42 W for the interval from 0 to 500usec. Average current is minus 8.6 A . 11.23 A. RMS. So, does it mean that I need a power supply of 12 volts and 8.5 A or better 9 A?

If I use the FSM to model the coils than can that model be imported into LTSpice?

jess

Hi,

The 12V power supply is displaying 2.59 A current draw when the duty cycle is 1.25usec at 100KHz. And the transistors are getting hot. Though according to simulation they will only dissipiate 1.8 W.

jess

That isn't *dissipation* in the source, it's power sourced plus dissipation in the (arbitrary) 0.1 ohm internal source resistance.

It's current flowing *out*, so, by convention, it's negative.

In a word, yes.

No. you have to extract the necessary constants, and use them to build a Spice model.

Not exactly. I drew out a FEMM model as I described, ran it, and used data extracted from the field plot to make the Spice model coils you see, which include mutual inductance and series resistance.

I *could* make a FEMM model from a .dxf of a (third angle projection) drawing of the setup.

The simulation uses an idealized value of C to get resonance. You can't do that with a real circuit. You're stuck with whatever capacitor you can get.

You need to adjust your exciting frequency for minimum sinusoidal voltage across the LC circuit. Ignore spikes. That will give true resonance.

It would be better to adjust for maximum voltage across C, but that needs high voltage probes.

Adjust for a dip across the LC.