W9NK90Z MOSFET Switching High Current with 2nF Cap and 4.418uH Air Core Inductor

I don't want to purchase a MOSFET with a higher VGS breakdown voltage at this moment, so I'm stuck with the W9NK90Z and its 30VGS breakdown voltage. I have the source connected to ground and the drain connected to a 2nF capacitor in parallel with a 4.418uH primary of an air core transformer. Those aren't committed values, but I don't want to lower the inductance by too much because I imagine it would lower the coupling coefficient. The series resistance is minimal. At the other end of the capacitor and inductor is a voltage source high enough to keep the MOSFET saturated when turned on. The MOSFET is turned on by a

The W9NK90Z has an Rds ON of 1.1 ohms. The 6.59 amps tells me your drain supply voltage is only about 7 volts. Is this correct?

No one would run 30 volts on a MOSFET gate. Looking at the datasheet shows 10 volts being enough to turn this device on hard.

This is the schematic of it. The Rds is accounted for. The ESR of the primary is estimated but probably exaggerated because I'll be using relatively thick copper wire. Not shown is a 30V gate-source Zener, but it shouldn't make a difference with 30V at the gate. The clock simulated has a 95% duty cycle at 100Hz, but the frequency will be variable, and the duty cycle can be changed.The two 47 kohm resistors are there to allow the primary voltage to bleed out a little. The reason that I'm using the maximum voltage allowable at the gate is because the current across a saturated MOSFET is very dependent on the gate voltage.

Cody, what is that wire-like thing connecting the +100 V to the 0.01 ohm drain resistor?

Again, 30 volts of gate drive is never heard of and is a good way to break a MOSFET. 15 volts is really as high as anyone ever drives a MOSFET with 10-12 volts being more than enough to saturate the device.

The .01 ohm resistor is the ESR of the transformer, not a drain resistor. The drain resistors are the 47 kohm resistors in parallel. The "wire-like thing" is the transformer that this program draws horribly. Yes, that amount of voltage will saturate the device, but as I said before, once saturation has begun, the current across the MOSFET is heavily dependent on the gate-source voltage. If I lowered the voltage at the gate, I'd be lowering the current across the MOSFET, which is the exact opposite thing that I want. I'd happily lower the gate voltage if I could get a higher current using that lower voltage.

Have you got a secondary and a secondary load on the transformer?

The picture shows the secondary. The load is a small capacitor to ground. This is the circuit for a MIDI-interrupted dual resonant solid state Tesla coil, not showing the circuit that will interrupt a constant voltage according to a MIDI input. I will also use this circuit with a variable 555 timer input for my high school science fair, but with much larger values for the capacitors and transformer. I have it majoritively figured out except for the MOSFET not switching enough current.

OK, to drive primary current (and power) your circuit needs a resistive secondary load. Otherwise all the load is is the primary winding magnetizing inductance.

As the circuit drives the resonant circuit there will be energy exchanged between the inductance and the capacitance and there will be loss in the circuit resistance.

So, your transformer model needs the proper primary inductance, secondary inductance, and the coupling factor. Or skip the coupling factor and use a series inductance to represent the leakage inductance. A Tesla coil has loose coupling - high leakage inductance - and this allows the secondary to swing through a huge voltage range without the primary loading it too much and reducing the circuit Q.

I found out this might just solve a problem that I was having

I understand how a Tesla coil works. The coupling coefficient is modeled to be .2. That's an extremely rough estimate. Since the primary inductance will be pretty small and the spark output will also be relatively small for this specific build, I'll be winding it vertically or conically rather than in a spiral to increase the coupling. The transformer model has the secondary inductance and ESR of a small prototype secondary that I wound. The primary inductance and ESR are estimated. With that and the parallel capacitor remaining resonant with the secondary circuit, I plugged in varying inductance and capacitance values, and the secondary voltage was highest with a lower primary inductance (and the correspondingly higher capacitance in parallel with the primary). I would expect that because of the lower impedance and higher inductance ratio. Anyways, I imagine that a lower impedance would mean higher current, so I don't understand why that's a problem. I just want the MOSFET to be switching more current. Since the MOSFET is being saturated, the current being passed is limited by the maximum gate-source voltage allowed by the built-in Zener and the threshold voltage. I've also simulated the MOSFET with the same threshold voltage, drain voltage source, on resistance, and gate-source voltage only. No load beside the on resistance. The current was just 200mA higher. Lowering the on resistance kept the same current. Adding a resistive load at about 20 ohms reduced the current slightly, which is to be expected. I want to know how to exceed the limit placed by the gate-source voltage. There must be a way if this MOSFET has been tested to have a maximum continuous current of 5 to 8 amps. With all due respect, I really do not need help with the rest of the circuit. I've been studying spark gap Tesla coils intensively for months, and I just want to try out a small and simple solid state one with the same principles as a spark gap Tesla coil (except the location of the MOSFET switch is in a different location than the location of the spark gap switch because the output voltage was even lower with the MOSFET there). The only problem I'm running into is the MOSFET not passing enough current, and I'd appreciate it if you could help me out with that.

I don't have a complete picture of your SPICE model is so I cannot run it. I don't know the transformer parameters.

But from what you are saying the problem with your SPICE model is the load and not the MOSFET. There must be a resistive load on the transformer secondary to reflect a resistive load to the transformer primary. Presented with an impedance of less than 15 ohms the MOSFET will switch more than the 6.59 amps you simulate.

Building a working SPICE model often consists of building small parts of the circuit, getting them working per calculations, then adding more of the active circuitry. If you'd like to eliminate the MOSFET from the SPICE model a square wave voltage source can be substituted. After getting the voltage source and Tesla transformer working the MOSFET can be added.

I'm not using SPICE. Typically, I run my circuits in a live circuit simulator for convenience before rebuilding the circuit in LTSpice for accuracy. When I have problems in the live simulator, the same problems arise in LTSpice. When I have no problems in the live simulator, easy-to-fix problems may arise in LTSpice, but it's not a common event. Anyways, the problem seems to not be with my load. That's what I was trying to point out with my simulation of the MOSFET with no load. At the gate, I had 30V. The source was grounded. The drain was given any voltage through a 1.1 ohm resistor (the on resistance). The max current I could pull with 30V at the gate and with a 3.75V threshold voltage was 6.89 (I meant to say 300mA higher earlier, not 200) amps, and that occurred at any drain voltage that saturated the MOSFET. This simulator often agrees with LTSpice (within some small tolerance due to the longer time step of the live simulator). I can't find any SPICE models of the W9NK90Z. Otherwise, I would simulate it in LTSpice to check if it's simply a problem with the equations that this simulator uses for MOSFETs. I just found two IRFIB7N50A MOSFETs lying around, but I can't find SPICE models for those either. However, I still believe that I'd be getting a similar result in LTSpice. When a MOSFET is saturated, the current isn't limited solely by the on resistance and load resistance. Like I've said a couple times, it's mainly the gate-source voltage and the threshold voltage. There's a complicated equation for it at http://ecee.colorado.edu/~bart/book/book/chapter7/ch7_3.htm. Simply search (ctrl+F kind of search) for "saturation," and it'll be the first thing. Wikipedia--using two other sources--gives the same equation but with an additional channel-length modulation parameter. If I knew what each of those parameters were, I'd do calculations myself. Speaking of parameters, you said that you don't know the transformer parameters. It has all been given.

Primary inductance: 4.418uH
Primary ESR: .01 ohms
Secondary inductance: 2mH
Secondary ESR: 10 ohms
Coupling coefficient: .2

Replacing the MOSFET with a square wave voltage won't allow the voltage to oscillate because a voltage source mandates a certain voltage, as I should. I tried it anyways, and the results were horrid. Extremely high current but extremely low topload voltage. I then had the square wave control an analog switch with the same on resistance as the MOSFET. The result was 90 amps through the switch and an adequate (but adjustable) 16430V out of the simulated topload. Limiting current is much simpler than increasing current. Anyways, that leads me to believe that the problem is either the MOSFET or the simulator's idea of a MOSFET.

I just took a look at the graphs on the datasheet. There must be something horribly wrong with the simulator. With only 30V at the drain and 10V at the gate, this thing is supposed to put out 18A. Now I know why you said 10V would turn this MOSFET on hard. I wish there was a SPICE model for this device so I could really see how much current it would pass in this circuit without having to buy a better multimeter. I guess I'll just test it out using values from the datasheet after I figure out the MIDI-input interrupter subcircuit. I'm thinking about using some kind of leading edge detection circuit, but the schematic I found requires 4 n-type and 4 p-type MOSFETs, which I don't have. Anyways, thank you for helping me and putting up with my ignorance. I've learned a handful of things from this. The two biggest being the reason why Tesla coils have low coupling (I always thought it was because of primary arcing and corona) and the fact that I should always compare simulation results with the curve characteristics on datasheets.

Highly upgraded AEGs require a lot of power, and they need the MOSFET to be capable of handling this power. While any MOSFET provides a huge advantage over the stock trigger contacts, most MOSFETs aren't designed to switch the most demanding AEGS. Airlab's MOSFET is designed specifically for highly upgraded AEGs, making it capable of switching extreme currents without heating up. This capability makes it extremely reliable in any AEG, so it is a good choice even for modest setups.
Thanks!!!!!

OK now that I have the secondary inductance I built an LTSpice model.

Using a similar MOSFET - the STW11NM80 (already in the LTSpice library) the circuit drain current does ramp up as expected by the formula V = Ldi/dt.

There appears to be something wrong with your simulation. I suggest moving on to LTSpice.