Quest for fastest and sharpest solid-state switching

[nilrehob]

Hi everyone,

After looking at http://www.energeticforum.com/renewa...ng-batery.html and the related videos I got a severe craving for a hex-controller

But since it seems like these controllers are hard to get and what is really needed for a project as above is very very very fast and sharp switching I figure the hex-controller is overkill (sigh!).

So I'm humbly asking the members of this forum for schematics or links to the same for working circuits (component list included) that would give me the fastest and sharpest solid-state switching on the planet (given a hobbyist budget).

/Hob

[nvisser]

Here is the Les Banki circuit intended for the Bob Boyce toroid. The circuit is for around 42 KHz, 21 kHz and 10 kHz but you asked for a fast switching circuit, the mosfet driver part and the fets. That can be connected to any oscillator and you apparently should use the 3 same frequencies.
The TC4420 is a 6a high speed mosfet driver with rise and fall times of 25nsec!. That means if you feed it with an oscillator with a slow rise time like a 555 it will take that signal and convert it to one with a 25nsec rise time and it will switch the mosfet on fully.
You could use TL494`s for your 3 frequency inputs
These parts are cheaper and the driver’s pinout are the same as the ones bob uses.
As for the fets. Use IRF540n. Also cheap
The toriod tuns you can get on chapter10 of Patrick O kelly`s site. PJK book ( Bob Boyce )
Even Bob said that he thinks the future for his technology lies in electricity generation and not in hydrogen generation
I hope this is of any help

Les Banki starter kit!

[nilrehob]

Thanks!
This will take some time to digest though.
I'll try to figure out what to omit and what to keep.


/Hob

[boguslaw]

25ns is too slow and FETS have a protection diode builtin but your searching is going to be in RIGHT direction.
I think it can work with a few mosfets right on the same wire switching with almost the same frequency or with the same frequency but slightly shifted in phase.

THIS IS WHAT TESLA DID ! LOOK FOR MULTISPARKGAP !

[boguslaw]

We need switching time of PICOSECONDS

[nilrehob]

We need switching time of PICOSECONDS

So you're saying the above wont work?
Do you have any other solid-state suggestions?

/Hob

[nvisser]

I dont even think the microwave EE can do that. Bearden talks about 5nsec.
We are talking about the rise time .
In the old days they apparantly used high voltage spike and clamp it on low voltage. That would gave them the fast rise times
Here is a way of speeding up switching time to the order of 2 ns!!, have a look at US Patent 5332938 from 1994, But it needs high voltages in the range of 100v.
Where do you get that information and components that can do that?

[boguslaw]

So you're saying the above wont work?
Do you have any other solid-state suggestions?

/Hob

I wasn't easy but I found it : Transient voltage suppression diode - Wikipedia, the free encyclopedia

though the actual circuit would be quite hard to properly make because it's a voltage controlled switch

There are also other surge protectors even for a few kV range. Don Smith is using them btw

[boguslaw]

Actually the answer lies in multispark gap used by Tesla.
I bet that Bedini is using it also : two frequencies !

Get two square wave or AC signals of the same frequency or different and shift one in phase.Then merge and you can even got sub-picosecond switch.

Am I correct ?

[nvisser]

Not to sure how you plan to use a transorb to do the job. The only way i can think of is to drive it over its protection voltage and use that switch protection pulse
How will you control the pulse width?
Any way I do not think you need to have such fast rising times. The Boyce and Banki circuit use the drivers with 25nsec rise and fall time and the fets can handle that to
Boyce said to make the pulse lenghts around 2.5usec (duty cycle)

[nvisser]

This voltage suppressor diodes are used on the fets of the Banki circuit I posted but for protection only
I looked at the datasheet and think I can see how it is possible. They come from 6v to about 300v. So if you choose a say 9 v one it will go into protection when your input pulse reach 9 v and it will keep on rectifying or switched till the signal drops below 9v .So the pulse width follows the input signal pulse width
Does that sound correct?

[boguslaw]

how about that way :
you charge a capacitor and discharge to the coil by simply rather slow mosfet.while there is a discharge you charge capacitor again or use second capacitor already charged.
now mosfet closes but there is a bemf from coil (coil must be large) which is over 300v and opens transorb in 1 picosecond whcih allows almost instant capacitor discharge ,right ?

1 picosecond disruptive discharge of capacitor means a very sharp gradient on coil.

As you see you must be very cleaver and find the correct timing

It is all about timing.

[nvisser]

Have you got any references to Don Smith`s work . Diagrams maybe?

[boguslaw]

Search for donsmith.pdf

[sucahyo]

That means if you feed it with an oscillator with a slow rise time like a 555 it will take that signal and convert it to one with a 25nsec rise time and it will switch the mosfet on fully.

I don't think it will after seeing the result of my one transistor circuit and two transistor circuit. The output of two transistor circuit is more than half of one transistor circuit. The second transistor is used to trigger the first transistor using 555 signal. This will sharpen the signal and make it stronger too (info from Harvey at Rosemary thread).

I think fast transistor driven with slow rise time will be slow as well. By using cascade configuration the switching is improved considerably. It help mine to reach 350V at the LV part (the HV part is out off my meter range).

I am using high voltage high speed transistor scrapped from computer power supply, KSC5027. Have no success with mosfet.

YouTube - Self voltage adjusting of radiant oscillator