Need Help with LTspice Circuit for Simulating a Spark Gap

[bobfrancis1980]

LTspice does not have a native spark gap component so I copied the one from this video on youtube:

/>
If I click on the Cap text in the schematic the scope shows a climb up to 3.3kV after about 14ms which makes sense since I the dielectric strength of air is defined as a constant in the schematic.

I have two problems. One, I can't figure out how to get it to show the output voltage pulse for when the spark gap breaks down.

Two, there is also only one capacitor charge up when there should be many, not sure what I am doing wrong. I have attached the LTspice schematic in a zip file, I would appreciate any help you can give me

Bob

P.S. I enabled visuals for all voltage related phenomenon and I can see that the 6,000V is constant for the entire time frame under the label V(n001). That probably refers to the voltage source V1 in the schematic.

[davekni]

I have two problems. One, I can't figure out how to get it to show the output voltage pulse for when the spark gap breaks down.

That circuit has no output.  Spark gap shorts capacitor.  If you want an output, you need to model the load of that output and connect it in series with the spark gap.

Two, there is also only one capacitor charge up when there should be many, not sure what I am doing wrong. I have attached the LTspice schematic in a zip file, I would appreciate any help you can give me

Circuit didn't simulate at all until I reversed instance name and value for Sgap.  Once I reversed those two labels, worked fine.  Produces a series of cap discharges at ~68Hz.

That model includes a slight breakdown voltage increase with time.  Is that to model electrode wear increasing gap?

What features of a spark gap are important to you?  In other words, what needs to be modeled?  Gap resistance is quite non-linear rather than fixed 100 ohms.  Small bits of noisy corona discharge likely occur before gap triggers.  If retriggered rapidly, remaining ionized air from previous trigger will reduce breakdown voltage.  A simpler model should be possible if only basic gap switching is sufficient.  Adding more effects adds complexity.

[bobfrancis1980]

Thank you so much!!!

The Youtube image had the S1 on top and Sgap on the bottom.
I also figured out why the V(cap) was going off at 3300volts. Because that was the breakdown voltage. If I want to see ~6000volts from the charging capacitor I have to increase the breakdown voltage (and in a marx gen the electrode distance) to compensate.

The youtube author had two simulations for the spark gap, one that assumed no growth rate and one that did. I will be using compressed air from an air tank so I can probably go ahead with the no growth rate circuit.

Now I am going to turn the circuit into a component and use it in the marx gen circuit I made.

Thanks again, you saved my bacon

Bob

[bobfrancis1980]

I am stuck again

I am not seeing the voltage rise times when I run the 2 stage marx simulation while attempting to use my componentified Spark Gap.

I am just trying to get the basic marx circuit working, then make it 20 stages and play with the resistor and capacitor values to figure out my design.

Any help you can offer would be greatly appreciated 

Bob

[davekni]

Spark gap circuit still has values referenced to ground, not all floating.

For Marx simulation, it would be simpler to use plain switches.  Wire all controls to a common pulsed voltage source.  Set voltage source for say 100us pulses every 10ms or whatever fire rate you need.  Then adjust charging circuit to get desired voltage at firing time.  The important part of Marx simulation will be modeling wiring inductance and especially parasitic capacitance between stages and to ground.