HV pulse transformer
lamazoid, Wed Mar 21 2007, 08:56PMHi everyone!
I want to write a program for calculating pulse transformers like ignition coil, those used in stun guns etc.. But unfortunately i can find no theory for them, exept common description like this one:
Can anyone help with this?
Re: HV pulse transformer
Steve Ward, Wed Mar 21 2007, 11:42PM
This seems like a tricky one.
I believe you can model them similarly to a tesla coil, except the primary is "untuned". The basic operation is storage of energy in the primary inductance of the transformer. Then by interrupting the primary current, energy is forced out of the secondary winding. By conservation of energy you know what the output energy is, and if you know the L and C paramters of the winding, you can figure out what the peak voltage and current will be. For better predictions you can factor in winding resistance and possibly losses in the magnetic core material, but i dont know too much specifically about that. You also have to be sure you arent saturating your core, otherwise the calculations go right out the window.
The biggest problem will be determining the capacitance of the secondary winding. This factor is as important as any, so without knowing it, you wont have much idea as to the output voltage/current (though you can still know the potential energy there). I dont know of way to calculate this capacitance, it seems to be something that you measure experimentally (by knowing the inductance - which is easy to measure - and applying a step or pulse to the coil, you can find the resonant frequency and work out the capacitance).
Steve Ward, Wed Mar 21 2007, 11:42PM
This seems like a tricky one.
I believe you can model them similarly to a tesla coil, except the primary is "untuned". The basic operation is storage of energy in the primary inductance of the transformer. Then by interrupting the primary current, energy is forced out of the secondary winding. By conservation of energy you know what the output energy is, and if you know the L and C paramters of the winding, you can figure out what the peak voltage and current will be. For better predictions you can factor in winding resistance and possibly losses in the magnetic core material, but i dont know too much specifically about that. You also have to be sure you arent saturating your core, otherwise the calculations go right out the window.
The biggest problem will be determining the capacitance of the secondary winding. This factor is as important as any, so without knowing it, you wont have much idea as to the output voltage/current (though you can still know the potential energy there). I dont know of way to calculate this capacitance, it seems to be something that you measure experimentally (by knowing the inductance - which is easy to measure - and applying a step or pulse to the coil, you can find the resonant frequency and work out the capacitance).
Re: HV pulse transformer
Sam, Thu Mar 22 2007, 02:26AM
if You are just looking for testing, You could use a micrometer to adjust a spark gap, then You could estimate the output voltage, as for current, maybe a current clamp?
Sam, Thu Mar 22 2007, 02:26AM
if You are just looking for testing, You could use a micrometer to adjust a spark gap, then You could estimate the output voltage, as for current, maybe a current clamp?
Re: HV pulse transformer
lamazoid, Thu Mar 22 2007, 12:15PM
Steve, could you give some exact formulas? E.g. for capacitance-inductance-voltage relations?
lamazoid, Thu Mar 22 2007, 12:15PM
Steve, could you give some exact formulas? E.g. for capacitance-inductance-voltage relations?
Re: HV pulse transformer
ragnar, Thu Mar 22 2007, 01:51PM
Lamazoid, please could you spare us the hassle and try a Google search or your local library? We'll be happy to help if you get stumped.
ragnar, Thu Mar 22 2007, 01:51PM
Lamazoid, please could you spare us the hassle and try a Google search or your local library? We'll be happy to help if you get stumped.
Re: HV pulse transformer
lamazoid, Thu Mar 22 2007, 05:57PM
I did a huge search before posting here. And all i found relates to left or say, low voltage side. This is well described on coilgun sites like this:
For example, i have modelled a typical stun gun circuit (1kv; 0.33uF; 0.1 Ohm; 15uH) This results into pulse width of about 8us and peak current of 140 A.
So, for the secondary winding we have pulse width, inductance, resistance (dc) and propably capacitance too. And we need to find optimum ratio between them. That's what i am asking about. Because *sorry* i was unable to find any information!
Maybe you will experience more luck, or you can just tell?
lamazoid, Thu Mar 22 2007, 05:57PM
I did a huge search before posting here. And all i found relates to left or say, low voltage side. This is well described on coilgun sites like this:
For example, i have modelled a typical stun gun circuit (1kv; 0.33uF; 0.1 Ohm; 15uH) This results into pulse width of about 8us and peak current of 140 A.
So, for the secondary winding we have pulse width, inductance, resistance (dc) and propably capacitance too. And we need to find optimum ratio between them. That's what i am asking about. Because *sorry* i was unable to find any information!
Maybe you will experience more luck, or you can just tell?
Re: HV pulse transformer
Steve Conner, Thu Mar 22 2007, 06:12PM
It's all just standard transformer theory. Or Tesla coil theory: a Tesla coil is just a pulse transformer with an air core, and a pulse transformer is just a Tesla coil with a ferromagnetic core.
The equations you would use are the same differential equations that a circuit simulator like SPICE uses.
You may not be able to rearrange these equations to get an analytic solution for the parameter you want to optimize. SPICE doesn't even try: it solves them by numerical integration.
My own interest was in finding analytic solutions to optimize solid-state Tesla coils, but I never got anywhere. I found one interesting transfer function that wasn't generally known, but I failed to explain it to anyone else. Terry Fritz made much better progress with a program that numerically solves the Tesla coil circuit over and over again, walking the solution space to find the optimum. I think: I never quite understood what he did.
I personally gave up on my analytic stuff and used parametric runs in PSpice to optimize my Tesla coil designs.
Steve Conner, Thu Mar 22 2007, 06:12PM
It's all just standard transformer theory. Or Tesla coil theory: a Tesla coil is just a pulse transformer with an air core, and a pulse transformer is just a Tesla coil with a ferromagnetic core.
The equations you would use are the same differential equations that a circuit simulator like SPICE uses.
You may not be able to rearrange these equations to get an analytic solution for the parameter you want to optimize. SPICE doesn't even try: it solves them by numerical integration.
My own interest was in finding analytic solutions to optimize solid-state Tesla coils, but I never got anywhere. I found one interesting transfer function that wasn't generally known, but I failed to explain it to anyone else. Terry Fritz made much better progress with a program that numerically solves the Tesla coil circuit over and over again, walking the solution space to find the optimum. I think: I never quite understood what he did.
I personally gave up on my analytic stuff and used parametric runs in PSpice to optimize my Tesla coil designs.
Re: HV pulse transformer
Dr. Shark, Fri Mar 23 2007, 11:32AM
I think the main problem with ignition coils and the like is that in general you will at best have a vague idea about the capacitance and inductance of the secondary side given just geometric parameters. It's a lot easier to get accurate estimates of your circuit parameters with a nicely wound Tesla coil.
If you can measure these parameters, the maths is relatively straightforward. Look at weakly coupled oscillators, a concept that frequently comes up in physics, to see how the energy sloshes back and forth between primary and secondary side.
Dr. Shark, Fri Mar 23 2007, 11:32AM
I think the main problem with ignition coils and the like is that in general you will at best have a vague idea about the capacitance and inductance of the secondary side given just geometric parameters. It's a lot easier to get accurate estimates of your circuit parameters with a nicely wound Tesla coil.
If you can measure these parameters, the maths is relatively straightforward. Look at weakly coupled oscillators, a concept that frequently comes up in physics, to see how the energy sloshes back and forth between primary and secondary side.
Print this page