Use for the Tesla Switch

[ldissing]

How do any of you truly BENEFIT from my work?

How would that improve your ability to think for
yourselves? To accept a challenge? To create
a solution? To learn something new by way of
one's own efforts?

Dave has the ability to make his diagram better.
Yet he has chosen to 'bail out.' Depriving many
of an opportunity to unravel the complexity of
his diode steering array.

Or, possibly Dave is unable to decipher it himself.

I'm sorry, but you do not seem to fully understand
what your experience has demonstrated to you.

This is one mighty deep rut that y'all are in...

Mr. Monkey,

Please start signing your posts with your real name, it is absurd to talk to a fiction.

I told you I would not write to you again, but you keep on posting and posting and writing and slamming and ....

You are right on with some of your points, so learn us sompin'. Teach us to be better engineers, get us out of the rut. Contribute man, contribute. You act and sound like you are an authority, so be one. Suggest the alternatives that you so eloquently write about. Why do you continue to write things that have no "real" meaning, just to say that we are in a rut. That should be obvious, really. This has been going on for 2 years before JB and 1 year after JB, and JB did contribute a thing or two. Would you be willing to actually build something and/or contribute?

How does one get out of a rut? Test, test, test, but then again if one doesn't understand what one sees and the results one gets, then that is all mute. Make a freakin' contribution and quit telling us we are in a rut....stating the obvious is folly.

My contention is that this is all about resonance and none of know that much about it, so why don't you in your intellectual superiority educate us on what resonance means. I'm not even being snide with the previous sentence, you seem to want to help, (maybe just the newbies), but it is the builders and the old timers that want this to work more than the newbies. So, help the old timers and you WILL help the new guys, but not with the crap you've been spouting.

You seem to like putting down Dave, or Matt or me, or whomever, but this is doing nothing and you are trying to play mind games. I know what you are and others will finally decide this too if given enough time. You could help those who have built the TS while others just watch and wait for someone to tell them they "have it". Is that what you are also looking for? Perhaps, you can help, but do you have it within you?

Are you willing to suck up your ego? Are you willing to put an idea forth? Do you have any...I think you do, don't be so afraid. We won't tell you you are an idiot for trying....trying is the name of the game in an adventurous game...is this not a game?

Leroy

[SeaMonkey]

See thats your typical response. When faced with what you have not seen and you know little about. "We are the Fools", Nothing We see is the truth its only a lack of knowledge. Yet thats all you have for an explanation. Over and Over again.

I have to find out a few things.

Where did you get the info on Ronald Brandt? You said And I can go back and find the posts if for some reason you think I mistaken.
Ronald Brandt only got 4-6 time more work for the energy in his Batteries.
Where did that come from.?
Ronald Brandt had to desulphate his batteries before he could charge them.?
Where did that come from.?

All I want is the reference to those 2 things. Thats it. You've handed out reference material. Please give me the reference for those 2 statements.
And don't go getting them off of the first 2 pages of Google like you have everything else. Cause I have already read all that stuff.

I am positive Vissie and Stevan will be successful. They are hard working humble people ready to help and prove themselves. Humility is the key ingredient. Do you know that word Humility. The act of being Humble.

YOUR NOT!!!

You should take Vissies advice and start your own thread. You'll find interested people initially but sooner or later they'll need proof as well. Unfortunately since you have showed up we have force to share in the background through private message.
No one right now trusts you and almost everybody who frequents this thread has posted there discontent and mistrust for you.
We believe you came hear to derial this conversation and any progress on it.

I am only going to ask one time, Please start your own thread and leave this one alone. If you care about the newcomer you will.

Matt

It's interesting how dis-info gets started. It's
almost like the game of 'telephone.'

Ronald Brandt stated in an interview that his device
extended the useful life of his batteries per charge
but that it would ultimately run down. The batteries
would require periodic charges to keep them going.
The device that he was describing was the one he
built for his automobile which was converted to electric.
This interview is available from at least two magazine
articles which appeared approximately 20 years ago.

It was not stated that his batteries required desulfation.
Only re-charging to compensate for the eventual
run-down of his device. It did not produce perpetual
motion.

Not a problem! The atmosphere in this particular rut
is a bit stale anyway.

For those who may want the 'secret' to a smoothly
functioning "Brandt-Tesla-Switch" which utilizes MosFets
as switches:

(1) Ordinary Low Side driver chips, preferably with an
opto-isolated input for isolation, are all that is necessary.
One for each of the six N-channel MosFets.

(2) A small toroidal transformer is utilized to bring power
to each of six driver circuits. The toroids are driven by
a high frequency oscillator of 50 ~ 100 KHz with the
output winding feeding a full wave voltage doubler to
produce a DC potential of 15 ~18 Volts to power the
MosFet Driver Chips.

The numbers of turns for the transformer (1:1) will be
determined empirically with 3 turns per volt as a starting point.
There are several chip devices which will serve well as
the Oscillator and Driver for the 6 toroidal transformers
whose primaries are parallel connected and fed through
'twisted pair' wiring to each. The total amount of power
required is not great.

Small Driver Boards are created for each of the MosFets
to be placed in close proximity to them.


(3) A Pulse Width Modulator Controller such as the 3525
is utilized to produce the drive pulses for the two banks
of batteries, switching the banks alternately between
"discharge" and "charge." The 3525 enables operation
over an adjustable range of frequencies with adjustable
pulse width 'duty cycle' and dead time.

(4) Some diagrams show capacitors in the output line
to the load. Their only purpose is to prevent a continuous
DC current path to the load in the event of some sort
of switching malfunction. 1000 uF may not be large
enough; you may want to use 4700 uF with low ESR.

The 'polarity' of the 'loop' as shown in early drawings is
incorrect. Use the polarity provided in the simple drawing
which shows the positions of the MosFets in the 'loop.'
This will assure that each MosFet is oriented to properly
conduct 'loop current' and avoids 'conflict' with the
MosFet body diodes.

Yes, it does work.

How long will it run? Depends upon your skill and patience
in 'tuning' it.

[DoubleD]

For those who may want the 'secret' to a smoothly
functioning "Brandt-Tesla-Switch" which utilizes MosFets
as switches:

(1) Ordinary Low Side driver chips, preferably with an
opto-isolated input for isolation, are all that is necessary.
One for each of the six N-channel MosFets.

(2) A small toroidal transformer is utilized to bring power
to each of six driver circuits. The toroids are driven by
a high frequency oscillator of 50 ~ 100 KHz with the
output winding feeding a full wave voltage doubler to
produce a DC potential of 15 ~18 Volts to power the
MosFet Driver Chips.

The numbers of turns for the transformer (1:1) will be
determined empirically with 3 turns per volt as a starting point.
There are several chip devices which will serve well as
the Oscillator and Driver for the 6 toroidal transformers
whose primaries are parallel connected and fed through
'twisted pair' wiring to each. The total amount of power
required is not great.

cool About time !

Now tell me HOW you figure this is SIMPLER than mine ??

Instead of pulling the Boost power from the already existing source in the cct. for the 2 fets that HAVE to be switched as High side, you are adding a transformer and another chip for oscillator to EACH Fet's sub cct.

Oh and my Drivers ARE opto isolated too

Sorry but I don't see it.

I can see the need for something to this effect IF we were working with much lower voltage system where the available voltage is no where near enough for good switching though.

Actually this is the same thing used on some of my big IGBT H-Bridge motor controller ccts
I can post some pics if anyone wants to see.

.

[nvisser]

Are 60nsec output rise and fall times fast enough to drive the fets of a TS effectivlery or do we need faster?
Bearden said that we will never get the battery back poppers to work if we do not switch in about 2nsec!!!!
I have not see a fet driver with that specs.
What microvawe components is he talking about?

[Matthew Jones]

When he says Microwave, he talks about Microwave Communication. Signal Switching. And from what I can gather when It comes time to switch that fast and shuttle a bit of power you have to be able (In the circiut) Start the switching process before you even need it, While the prior process is still happening. The amount of discharge may be that fast but the actual process to get there may take more time.

Either way he is more or less talking about duplicating something along the lines of Transient spike which has no width with out a coil.

That how I understand it anyway.

Matt

[StevanC]

:splat on the forehead:

cool About time !

Now tell me HOW you figure this is SIMPLER than mine ??

Instead of pulling the Boost power from the already existing source in the cct. for the 2 fets that HAVE to be switched as High side, you are adding a transformer and another chip for oscillator to EACH Fet's sub cct.

Oh and my Drivers ARE opto isolated too

Sorry but I don't see it.

I can see the need for something to this effect IF we were working with much lower voltage system where the available voltage is no where near enough for good switching though.

Actually this is the same thing used on some of my big IGBT H-Bridge motor controller ccts
I can post some pics if anyone wants to see.

.

@DoubleD,
1. What's the difference of Rds(on) under near to maximum rated amperage in the function of gate voltage?
1.1. And especially in time limited "out of spec " (=above) Amperage levels (100A for IRF3205)?

(i suppose You can decipher a data-sheet?)

HINT:
what's the difference in between 12V and 18V of source to gate voltage level in regards of the ability of the MOSFET to carry current?

"CHEWED UP" (=for layman):

the BOOTSTRAP procedure requires a capacitor large enough to maintain full saturation voltage levels for the while the device is afloat in full conduction and under load.
The limiting factors are the ability of the charging diode (1N4007 on my circuit) to deliver enough charge to a capacitor large enough (10uF in my cct) to sustain the "floating" period abowe both the undervoltage lockout and the most beneficial voltage (near Vgs(max) (is 20V in my circuit) ).
In my case (420mces) is is simply not ample enough! My bootstrap keeps falling through the 9V level all the time!!!


While on the other end (floating supply):
A. We CAN easily produce a stable 18V
B. We can find 9nsec/6A(and better) low side drivers more easy
C. short leads driver to power-stage increase performance
D. Complexity has to be somewhere, If we load itall to driver-stage, then the "brains" and power-stage remain bare and KISS (=good IMHO)

So,
IMHO:
I. Floating supply of 18V rectified AC for opto-isolated (or opto-fed) low side drivers

II. Floating driver stage of 18V +/- pulses driving the power-stage directly with co-axial (RG) leads of exact length with 20V (=10V-10W+10V-10W) Zener clamped MOSFETS.


or we can't do it "just right"


Stevan C.
P.S.

We didn't even come to consider the LOAD:
It is a lowFVD transformer (300A rated primary?) with substantial core capacity (my gut tells me that..)
That is a "beast of its own"...

[StevanC]

Are 60nsec output rise and fall times fast enough to drive the fets of a TS effectivlery or do we need faster?
Bearden said that we will never get the battery back poppers to work if we do not switch in about 2nsec!!!!
I have not see a fet driver with that specs.
What microvawe components is he talking about?

I would opt for going to 9nsec if possible at all?
And traces within 15cm (6inch) or a fraction of it - the shorter the better?
There are 9nsec rated (for a 1nF load) drivers out there...
The rise time is key factor, the propagation time is of lesser significance AFAIK?

Stevan C.

[nvisser]

Do you know of a opto-fet driver with that specs?

[StevanC]

Do you know of a opto-fet driver with that specs?

Why opto-fet (=bad IMHO)?
Motorola has some and HP seems have some (seek MOC30xx or like?)

But the problem with optos is they either can't drive directly or are not fast enough or both (I wonder why?)

There are excellent high side drivers (I picked a 600V tolerant one) that just need "juice" to be fed to their "Vboost" and "Vg" (which are "floating end") and which then provide that to the respective Vg "leg" to the device being controlled.
My "favorite" has even 20V clamp (but limited) and works in both bootstrap or "floating" and as a low side device.

Best regards,
Stevan C.

[StevanC]

THE floaing supply

Why is this floating supply so horrible to accept?


1. Can we make a 555 driven "push pull" primary part of a 6 secondary windings floating PSU (the low side driver scenario)?
this one is more tempting, as we are not limited to nsec edges, we merely need 18V on the capacitor off the rectified side. BJT will do. We then merely "Que" drivers...


2. Can we make a 3525 driven dual transformer each in push-pull with 4 secondaries each double floating 4 MOSFET controller (the SeaMonkey's "final answer" scenario)?
Here we grab the lion for its tail (he he...)

It can't be that bad?

Stevan C.

[DoubleD]

@DoubleD,
1. What's the difference of Rds(on) under near to maximum rated amperage in the function of gate voltage?
1.1. And especially in time limited "out of spec " (=above) Amperage levels (100A for IRF3205)?

(i suppose You can decipher a data-sheet?)

HINT:
what's the difference in between 12V and 18V of source to gate voltage level in regards of the ability of the MOSFET to carry current?

"CHEWED UP" (=for layman):

If you think you can actually pull 100 amps through that TO20 package Go for it.
Also the amp rating is for only @10v too if you read the data sheet. actually that is about where Most are rated.

I am not here to fight and argue over this I gave you a Proven working cct that will handle most peoples needs to further their experimenting.

Every cct is a compromise of one sort or another.
I compromised a little on the Low side gate voltage for simplicity of design for the average use. There is still plenty of Voltage there to turn the fet Full on if you hit it fast.
Now if you use this for a Heavy Power application then it will need to be modified accordingly , just like any other device out there( the same cct for a 500watt inverter will not run a 5kw inverter withoput modification even if it used the same design !!

The Idea of the transformer supply (now that it was FINALLY put out there) Is great for the applications that require it. But be prepared for the extra complexity that come with it.
This is why I chose NOT to go that route here ( K.I.S.S. )

[StevanC]

A picture is worth 1000 words...

And a motion picture is worth a 1000 still images!

The FET pulser in action

This is off my anciet phone, so expect nothinng really, its just proof i got it and it works...

And few still "shots":
(regards< Stevan C.)

[nvisser]

Any diagrams?

[StevanC]

If you think you can actually pull 100 amps through that TO20 package Go for it.
Also the amp rating is for only @10v too if you read the data sheet. actually that is about where Most are rated.

Its standing there it can handle more (repetitive pulse) if 2% D/C
and one single pulse of 100usec 100A 20V
and 10usec 200A 20V as far as the junction is concerned,

And TO-220 package can handle 75A DC in pulses and up to 62A for the IRF3205 (in question) as DC.
The actual limits are subject to temperature derating, but bottom line:
I shoot not at 100A, but i really appreciate the overhead reserve

I am not here to fight and argue over this I gave you a Proven working cct that will handle most peoples needs to further their experimenting.

mine is working just fine up to 5A (5A x 20V)
but I see how I can improve it, to be better

Every cct is a compromise of one sort or another.
I compromised a little on the Low side gate voltage for simplicity of design for the average use. There is still plenty of Voltage there to turn the fet Full on if you hit it fast.
Now if you use this for a Heavy Power application then it will need to be modified accordingly , just like any other device out there( the same cct for a 500watt inverter will not run a 5kw inverter withoput modification even if it used the same design !!

that is correct, I agree

The Idea of the transformer supply (now that it was FINALLY put out there) Is great for the applications that require it. But be prepared for the extra complexity that come with it.
This is why I chose NOT to go that route here ( K.I.S.S. )

I agree here too:
it's a road wi woill be forced to take if we want to scale up.

Or,
If we want to squeeze the devices we picked with their top most performance?
or both?

Stevan C.

[StevanC]

Any diagrams?

Now this was ambivalent?


of what?
my YT circuit or the floating supply?
or both?


Stevan C.