Bipolar Avalanche Pulser

[ghost896]

Hello everyone, I am trying to make avalanche based high voltage switcher. Before this circuit, I made an avalanche and marx-based switching circuit (Symmetric Avalanche Marx Generator) but I couldn't get the output characteristic I wanted due to the different switching times. It was nice to see the high voltage and low rise time though. That's why I want to do the switching in a stack structure instead of staggered, but I'm having trouble setting up the symmetric structure. I was able to simulate the positive pulse from 1000V to 0V in less than 5 ns, but I cannot show the negative pulse. How do I set up the negative stack?
I will be grateful if you could help me.
Best Regards

 

[BradtheRad]

I think you want the negative pulse to be referenced to 0V ground?
So that you get double amplitude to your load?

Use PNP to create a symmetrical negative section.

This is done by:

* Copy your positive section. Change all NPN to PNP (while keeping emitter in same leg).

* Provide negative supply rail for negative section PNP transistors.
Thus 0V center ground should contact all NPN & PNP emitters.

* Turn ON PNP by pulling its base terminal more negative.
Shut OFF PNP by pulling its base terminal more positive. Do not assume that leaving base disconnected shuts off a transistor.

 

[FvM]

Unfortunately there are no pnp avalanche transistors available.

A positive going pulse can tapped at Q9 emitter. The trigger can be bestapplied through a transformer.

 

[stark43]

As @FvM said these transistors (Fmmt417) have no pnp series. I simulated the image I shared and I'm getting symmetrical output. All I want is to design them as a stack, not a marx generator-based cascade.

A positive going pulse can tapped at Q9 emitter.

Could you elaborate a little more, sir?

--- Updated Nov 21, 2022 ---

Hello everyone, I am trying to make avalanche based high voltage switcher. Before this circuit, I made an avalanche and marx-based switching circuit (Symmetric Avalanche Marx Generator) but I couldn't get the output characteristic I wanted due to the different switching times. It was nice to see the high voltage and low rise time though. That's why I want to do the switching in a stack structure instead of staggered, but I'm having trouble setting up the symmetric structure. I was able to simulate the positive pulse from 1000V to 0V in less than 5 ns, but I cannot show the negative pulse. How do I set up the negative stack?
I will be grateful if you could help me.
Best Regards
View attachment 179820
View attachment 179821

It doesn't drop to exactly 0 V, it drops to 700 V.

 

[ghost896]

The trigger can be bestapplied through a transformer.

I found such a scheme, it seems worth a try, but unfortunately I don't have much time. Is 5V enough as a trig? I think you are talking about 1:1 pulse transformer, right?(@FvM )

--- Updated Nov 21, 2022 ---

The up and down structure is also formed with the same transistor directions. I don't understand how to generate symmetrical wave. Any ideas?

 

[FvM]

Could you elaborate a little more

I'm referring to the post #1 schematic.
Short R13 at Q7 collector, insert the load resistor between Q9 emitter and GND.

 

[ghost896]

The trigger can be bestapplied through a transformer

Hello sir, are you talking about 1:1 pulse transformer? Is the goal to isolate input and output? Is there any pulse transformer?

 

[FvM]

An avalanche transistor with floating emitter can't be well triggerred by a ground referenced voltage source, mainly because the generated avalanche pulse would run back to the trigger generator.

The transformer needs isolation according to the pulse voltage, low capacitance, winding ratio appropriate for the driver circuit e.g. 1:1. A small toroid core with tripple insulated wire should work.

 

[ghost896]

An avalanche transistor with floating emitter can't be well triggerred by a ground referenced voltage source, mainly because the generated avalanche pulse would run back to the trigger generator.

The transformer needs isolation according to the pulse voltage, low capacitance, winding ratio appropriate for the driver circuit e.g. 1:1. A small toroid core with tripple insulated wire should work.

I can't find a specific product when I google the features you mentioned. Today is my last day to order from digikey. There is a 1:1 pulse transformer, but when you type toroid core, different things appear. I trigger the build with 5V and switch 2500V. Can you send me a link to a similar product (for example, from digikey) so I can research the parameters myself. Thank you sir.
Best Regards

--- Updated Nov 25, 2022 ---

An avalanche transistor with floating emitter can't be well triggerred by a ground referenced voltage source, mainly because the generated avalanche pulse would run back to the trigger generator.

The transformer needs isolation according to the pulse voltage, low capacitance, winding ratio appropriate for the driver circuit e.g. 1:1. A small toroid core with tripple insulated wire should work.

The voltage reflected on the trig pin as you said. If I use 1:1, won't it still be reflected to the source at the same voltage, sir? (1:1 Pulse transformer Digikey) Is it something like this you're talking about?

--- Updated Nov 25, 2022 ---

https://www.digikey.com/en/products/detail/pulse-electronics/PA2001NLT/2265708

 

[FvM]

The trigger voltage shall be applied between base and emitter of the transistor. That's not possible without a transformer,

 

[ghost896]

The trigger voltage shall be applied between base and emitter of the transistor. That's not possible without a transformer,

Thank you very much. But finally, do I need to isolate the trig source and the ground of the transistor?

 

[FvM]

As you know, the trigger transformer suggestion is related to the original question how to generate complementary pulses with avalanche transistors. My circuit suggestion is in post #6, it requires a trigger transformer. A design with grounded triggered transistor don't need a transformer, at least if the control circuit uses a common ground with the pulser.

Unfortunately you didn't yet sketch your final design idea.

 

[ghost896]

Unfortunately you didn't yet sketch your final design idea.

 

[stark43]

@FvM Which one is correct to make the base pin of T1 transistor high? The schema that @ghost896 shared seems wrong to me.