# LTSpice problem: higher voltages without inductors?

Hi, since this problem doesn’t let me sleep, I tried simulating it and finally got LTSpice up and running. Unfortunately, the results puzzle me.
The first pulse is as expected: It sends a shorter pulse over the cap to RST. But what the hell is going on after the second pulse? Where do 6V come from? Is this a simulation artifact and how do I avoid it? I would expect that V(emitter) goes to 3.3V-transistor dropoff, and at the next LOW pulse gets

Everything I configured should be visible in the schematic. Since some curves overlap:

• V(p1) follows V(d0), except for the small, visible transient at the beginning of the first LOW pulse.
• V(d0) is as configures: starts 3.3V, drops to low after 1s, then does a cycle of 3s LOW, 2s HIGH.
• V(rst) is at 3.3V all the time, except for the visible LOW spike at 1s.

Looks ok.... you've simulated a charge pump - the first stage of a Cockroft-Walton multilier.. The transistor is acting as a simple diode.

regards

Allan

Hm, ok. I kind of start to understand what is happening. Basically, I set the potential at the emitter to 3.3V/2.7V via the diode (transistor) while the other side of the cap is at 0V. Then I push the potential on the emitter side to 6V when the other cap side goes to 3.3V?
I was able to get the expected behaviour (regular short pulses to RST) with a Schottky diode from the emitter to 3.3V.
Is this a viable way?

Look up the Cockroft - Walton multiplier. They used it in the Cavendish lab in Cambridge to make the first linear accelerator. It generated about a million volts.

Allan.

Ah, I first looked at the German Wikipedia article about the accelerator. The English one is much better. Looks like I understood about right.

Ways to increase voltage:

inductive reactance - V proportional to dI/dt switched-capacitors (including Cockcroft-Walton - diodes count as switches) transformer (which is another way to use inductance) pump a series LC resonant circuit

There are probably others.

Yes, I knew all that involved inductances and falsely assumed you actually needed them. Thanks.