Tesla Switch need help

[plengo]

@ALL
I wish we would never piss people off!!!

@gyulasun
I understood what you said, even tought I think there was a little math error on your calculations. I am still working on the simulation to see at what level things get the "charges" transfered from cap to cap and how much is lost or not.

You see, I think it is impossible for the 'charge' to get lost or waisted but i understand that "energy" is lost during the process, I dont quite agree with it tough.

In real lab tests so far I did notice that the conservation of the "charge" is there and transfering from 2 parallel caps to 2 in series back and forth DOES work but looses about 15% or so of "energy" in each transfer, still one can loop this thing for a long time with just one "charge" from the battery source to the caps and let the caps feed a load until the whole "charge" is lost. (I keep saying "charge" in quotes so that all see that I mean "charge" = displacement of unequal distribution of charges on space, such as the plates of the caps or poles of a battery terminal, in other words, charges are the particle carriers of its energy sign and they not necessary move around space even tought they can, but their energy makes the space around be changed).

What is puzzleling me is the fact that 2 caps in parralel would transfer to 2 other caps also in parallel without ANY LOSS, being that all have the same Faraday capacity. That's something. So I am trying to figure out, before my next lab test, what is the BEST compination of transfering those charges back and fourth with the minimun loss as possible. Off course I am trusting (assuming) that the simulator is taking care of the mathematics for me to the absolute perfection known to the developers.

Fausto.

[Groundloop]

@All,

I deleted my posts because they are no longer relevant to this discussion.

@Plengo,

Have you seen this document: http://home.no/ufoufoufoufo/FECAP/cond.zip

Groundloop.

[plengo]

@Groundloop,

man THANK YOU that document is exactly inline with what I am thinking. Even the concept of charges they think the same that I just explained before. So you know that I am looking for now is what that document is showing, how to get the "energy" not lost and distribute that over the caps but in a way that we can use it again and keep that loop, looping.

I think also that when they explain about the Whimshurst-Machine is what happen in the battery ( in different ways off course) and which allows the "energy" to be not lost in the process of switching the batteries.

I really think we are at something here. Good theory, some experiments done. Let's accomplish more and the best, knowing "why" it works.

Fausto.

[Groundloop]

@Plengo,

Now that we know what we are up against, let us take this in small steps. My first goal will be
getting my HEXFET based switch up and running. We really need a good test bed that waste
no power at all. (Or at least as little as possible.) I have tried this before with my FECAP device
with poor results. So this time I will tweak my circuit until I get it right even if I have to turn
some HEXFETs the wrong way around and use higher gate voltages. I'm also thinking on using
some FET drivers ICs. I will have my next prototype PCBs in house around the 17 th. so I
guess I have to wait until then before I get any further with this project. I will be testing on my first
prototype while waiting for those PCBs to see what configuration I need to have for the best result.
I have also started to look into using small signal transformers instead of opto-couplers. The
transformers I have will handle signal rates up to a maximum of 15 MHz pulse ramp time. This should
be fast enough for this project. The big advantage with those transformers is the low power needed
to operate them. A opto-coupler waste up to approx. 20 mA at 5 Volt. A transformer do not. So sometime
in the near future I will have a good working loss-less switch for this project. I just need to slow down a
bit and think through this concept at a different angle.

Groundloop.

[gyulasun]

@ALL
I wish we would never piss people off!!! 

Hi Fausto,  I fully agree with your wish!

@gyulasun
I understood what you said, even tought I think there was a little math error on your calculations.
What is puzzleling me is the fact that 2 caps in parralel would transfer to 2 other caps also in parallel without ANY LOSS, being that all have the same Faraday capacity. That's something. So I am trying to figure out, before my next lab test, what is the BEST compination of transfering those charges back and fourth with the minimun loss as possible. Off course I am trusting (assuming) that the simulator is taking care of the mathematics for me to the absolute perfection known to the developers.

Fausto.

Well,  would you mind pointing out my little math error in my calculations?  Often it is easier to see others' mistakes than realize our own mistakes... 

Regarding circuit simulators, one has to be careful with the results received from them. It all depends first on the component models the simulator is using. In case of capacitors it is advisable to check if the model has got a series (or parallel) equivalent resistor to account for the losses that surely appear in a real electrolytic capacitor for instance. Because if your simulator just uses an ideal capacitor model then results will differ from real practical behavior. I suggest to connect normal resistors in parallel with all four capacitors in the simulation for your test 1 and run it again and see then the voltages on each capacitor. The value of the resistor could be 10 MegaOhm each.

In Groundloop's link, in the cond.doc there is a good point: to transfer charges through a coil (or transformer) so that a gain of 1.41 can be received.  I do not know yet if this is going to manifest in practice, I wish it would!
This same circuit was a topic on other Forums (for instance by Marcus Wagner in a Bedini yahoo group if I recall correctly) but unfortunately I have never ever seen it built and measured and what results may have been received from it.

Keep up good work,

Ciao,  Gyula