Use for the Tesla Switch

[StevanC]

Has anyone involved with this project put together
a verbal description or explanation of what is
desired in the way of performance objectives or
operational parameters for the final goal?

Apart from being as 'simple' as possible and
as efficient as possible?

Something along the lines of an RFP?

How about this:

1. peak output power 2kW
2. continous peak output power 1kW
3. scalable up
4. MTBF (minimum failure interval) 1-year (the more the better)
5. serviceability (parts tha wear off should be made swappable)
6. output rating: 12V or 24V to a battery or a transformer (AC) (1000/12=83.33A as average)
7. Average power yield a day: 24kWh =>2kAh @12V x 83.33A
8. license allowing public replication and building
9. license allowing commercial application worldwide without possibility of a monopoly (non exclusive rights only) such as the GPL-v3 does.
10. License should restrict safe keeping or censorship of the device/blueprints

for a start?
Stevan C.

[Matthew Jones]

If you were using say 100 amp hour batteries you would have to 24 of them to put out 1 kilowatt continues at or around a C20. And thats the bare minimum.
I am working on getting that many.
But for the average experimenter.... I am not sure if thats a fare expectation from anybody at this point. Especially in light that most people haven't built one that can deliver 100 watts.

Matt

[StevanC]

may the SOURCE be with You

-._.-
"strong the force is in him"

for those able to "C"

/*	Copyright (C) <2010>  Brent Burns burnsb55@gmail.com
        addons and further development: GPL v3 cest73@yahoo.com 2010
    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 3 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#define F_CPU 12000000UL  /* 12Mhz crystal */

#include <util/delay.h>
#include <avr/io.h>

// here we define constants
#define TIMING		432 //total time of an event
#define BITS		  2 //number of bits to sequentially toggle
#define QUE		420 //timme used in on state 
#define UDELAY		 12 //time lapsed in off state
#define TRUE		  1 // a constant

/* a little bit of ANSI here:
   | is OR
   & is AND
   ^ is XOR
   ~ is NOT
   << is bit shift ( "this" << "times" (left))
   >> right shift ("unsigned" x2^"times" shifted)
*/

// here we make macros that speed up code writing
#define bore		_delay_ms(UDELAY); // the micro delay
#define patience	_delay_ms(QUE); // the micro delay
#define clearC		 PORTC = 0;  // take all pins low on PORT C
#define clearB		 PORTB = 0;  // take all pins low on PORT C

#define BIT(x) (1<<x); // this is an ancient ANSI C macro really

// here ve define variables
char bit = 0;

// here we difine subroutines
static void
action (unsigned char x) //here we do action
{
		PORTB = (1<<x); 	//we rise one bit up (we "Que" it)
		  PORTC = 0b00000001; // we show we are "in Que"
		patience		//we wait...
		clearB			//we null all bits out (we "de-Que" all)
		  PORTC = 0b00000010; // we show we are "out of Que"
		bore			//waiting for nothing is boring...
}

// here is the actual looping code
int main(void)
{
	DDRC = 0b00000011; /* pins C0, C1 are outputs */
	DDRB = 0b00000011; /* pins B0, B1, B2 are outputs */
	  clearC		// we clear port C
	clearB		// we clear port B
	while (TRUE)
	{
	/* this is a ZIG-ZAG algorythm here
	usefull for doing cycles of 3 or more pins
	*/
	/* this is commented out (won't compile)
	    for ( bit = 0; bit < BITS;bit++)
	    {
	    action(bit);
	    }

	    for ( bit = (BITS); bit > 0; bit--)
	    {
	    action(bit);
	    }
	*/
	/* here we really are toggling just two bits
	indefinitely ;)
	*/
	    for ( bit = 0; bit < BITS;bit++)
	    {
	    action(bit);
	    }

	}
}

The files led to the pulsing of two pins:
advpulser-01
for those that can't C (pronounced "see" )

One needs the complete tool-chain (AVRstudio or like).
more in depth info regarding AVR:
AVR Freaks

Also there will be my helping hand for those faring on GNU/Linux with various recipes "en route" like:

# avrdude -V -v -c usbasp -p m48 -U flash:w:advpulser.hex

the command that led to uploading it to the AVR via the sibling 07302 from JYETech



Best regards,
Stevan C.

[StevanC]

Right ON

Hi Matt,
I'm all "open" for corrections here

You have a strong point there:

"what is practical"

My point of start was:

"what is sufficient to power a home"

Ideally we should meet somewher in between.

The question is where exactly?


Stevan C.

[John_K]

Gee I love Serbian humour

I was only shooting for the "1W challenge"!!

Back to the bar, oops, I mean bench...



John K.

[StevanC]

Gee I love Serbian humour

I was only shooting for the "1W challenge"!!

Back to the bar, oops, I mean bench...



John K.

John,
We ain't supposed to reach that in a step (a giant step that would be in deed ), but the 1W would be just fine as a beginning...
therefrom we would just scale up (x1000 )


Stevan C.

[ldissing]

You are getting something done

@SeaMonkey,

I'm impressed. I asked the question of "what was the goal" many moons ago, and got nothing. If nothing else, you've got some response, so that people may start to work toward a common goal with the TS. That must be the first thing, to get people working together instead of everyone on their own path, power in numbers and mind, with the total mind power focused, it can be accomplished.

@all,

If everything was switched in a home to 12V, then life is simpler, but expensive to do, this is probably not an option. I'd think that 5KW peak would be the goal, not 2KW but whatever you guys want to do.

My calculation of 1KW average = 83.3333 amps /C20 = ~4.15 Amps / battery ~= 20 100Ah batteries batteries assuming very little loss. On a 12V (24 max on a side) TS at 50% duty, you'd need 10 X 4 batteries or 40, if I haven't lost my mind. On a 36V TS, you'd need 10x6 or 60 batteries. In my tests I had about 6 times the power output of a TS over straight batteries, so those 40 batteries in a TS could last 120 hours instead of just 20 hours.

JB once said....if you want power, just hook up to the batteries (roughly). You don't need the TS for power (unless we can get 6-8 times the power out of the same batteries). If you could get the TS to charge the batteries by themselves without external power, you'd still need 40, so you could bank switch, charge one set while using the other for power, but the TS could be smaller....just a charging device, not a "powering" device. I.e. you don't need incredible power out of the TS in charging mode.

JB said the TS has three uses. My initial thoughts were 1.) for charging, 2.) for powering, and 3.) Don't have a clue.

Anybody have any thoughts or interest on just figuring out how to make it be a charger? If we could figure this out we "might" not need high current devices (just need the inverters, etc. which would be required anyway for powering 120V and 240V appliances). Of course, you could roll your own inverters too!

Anybody come up with the 3rd use for the TS? I haven't, unless it was for refrigeration i.e. (Dave's cold running TS, or JB running it underwater. I assume that JB was going to run it underwater because water would dissipate the cold better than air, but it is just assumption.)

Leroy

P.S. I finally built my HV PIC controlled battery charger/discharger, so I may be able to get into the swing of working on the TS again.

[citfta]

Goal for my TS

Hi All,

My goal has been to get the batteries to charge themselves or at least charge other batteries without losing their own charge. So far my best effort has been to get the TS to run for several days while powering itself without losing any charge. The only load was a few LEDs. So I think I have a long way to go to get any decent load to run or to get some charging done without losing charge on my TS batteries. I agree with Leroy if we can get some decent charging from a TS then we could just use it to keep our batteries up and not have to have two large banks of batteries. This seems like the easier way to go if it can be done.

@ Seamonkey, several of us have tried relay switching at one time or another. Matt and some of us have determined from testing that we don't need really high frequency of switching to get pretty good results if we are just running a load. I have used relays controlled by a PIC controller and gotten some ok results. Do you think the Mosfets will switch faster than relay contacts? I am asking because if relays are as fast as Mosfets then we are still missing something to get the results we are looking for. John B has suggested we only need switching speeds of around .5 Hz or so. Relays can easily handle these speeds yet we still don't seem to get any charging and our batteries eventually run down if we load them very much. I think Matt has been the only one to be able to run a reasonable load for an extended time. Thanks for the suggestions you have made and the resources you provided.

Carroll

[Matthew Jones]

@all,
If everything was switched in a home to 12V, then life is simpler, but expensive to do, this is probably not an option. I'd think that 5KW peak would be the goal, not 2KW but whatever you guys want to do.
My calculation of 1KW average = 83.3333 amps /C20 = ~4.15 Amps / battery ~= 20 100Ah batteries batteries assuming very little loss. On a 12V (24 max on a side) TS at 50% duty, you'd need 10 X 4 batteries or 40, if I haven't lost my mind. On a 36V TS, you'd need 10x6 or 60 batteries. In my tests I had about 6 times the power output of a TS over straight batteries, so those 40 batteries in a TS could last 120 hours instead of just 20 hours........
JB once said....if you want power, just hook up to the batteries (roughly). You don't need the TS for power (unless we can get 6-8 times the power out of the same batteries). If you could get the TS to charge the batteries by themselves without external power, you'd still need 40, so you could bank switch, charge one set while using the other for power, but the TS could be smaller....just a charging device, not a "powering" device. I.e. you don't need incredible power out of the TS in charging mode...........
Anybody have any thoughts or interest on just figuring out how to make it be a charger?
Leroy

Leroy, I cut bit of your message out of the quete.

5kw is my goal. But getting the batteries together is what is taking me the most time. You can run a rather large load and just use mechanical relays in parallel. Switch anywhere from every few second up to 10 minutes.
Inevitably I would like to see a TS version that knew when to switch to keep the potential on the inverter.

Bits came up with that charging scheme with a smaller setup but the batts wouldn't charge over a certain level. Thats the impedance in the battery. I would only guess but if you had the means to tune the switching characteristics based on the batteries impedance you could go further. But I don't have clue how.
I have the same problem trying to back charge a system. At certain point the system will not take any more charge and its alot lower than I wanna be at. But with an external charge while switching and running a load you can fill the batteries back up.
Thats what I wanna do with it is use it as an island system. The biggest thing I want to come up with means to switch it solid state and not have to worry about the potential drop in the system getting in the way.
If it works I am also really worried about building it with highly available cheap parts. The device would need to be very simple to reconstruct and repair.

Matt

[nvisser]

Seamonkey
I think you got a good idea what we want out of this switch.
Most of us were probably introduced to the TS by Patrick O Kelly that makes everything always sound so easy with quotes like: "The switching circuit is said to be able to power its load indefinitely"
Then he presents the tesla switch build by the Electro Dyne company that worked with a commutator and ran for 3 years while driving a load till they stopped it.
When JB came along and promised that he will hang around till we all got it working most of us started building.
He posted this circuit and told us to choose the correct load to get the batteries to charge. See "bedini TS charg2" below
He also post this drawing of the negative spike that the transistor should give that will cause the charging. All this was at a low frequency of about 2 to 4 hz. See TS signal.
He also post a circuit where he removed some of the diodes that really cost most of us some components.
I also include some of his quotes.
Needless to say nobody really had real success with it.

I have to add that everytime I read that bedini quotes document I just want to start all over again. It sounds so easy!!
I think it will be easier to get it to work using a SSG ,that does not draw to much current, as load instead of trying to drive inverters that eats current.

[Matthew Jones]

....I think it will be easier to get it to work using a SSG ,that does not draw to much current, as load instead of trying to drive inverters that eats current.....

I think if we get the switching to work correctly we will see a mass of current, or something like it, that will drive any load.

Matt

[vallentin]

Thank you Vissie!

We all need a restart. A reboot.
You managed to do that. Thank you again.
So, back to basics.
I remember that the scalar charger is a rough half of TS.
IF we manage to work it out, and UNDERSTAND it, only then we can understand the proper TS functionality and extract the best out of it. Remember the 3 battery test that JB wanted us to try...and understand?
Experimenting with the cap pulser I found that the ~18V left is fair enough to charge a small battery.
Yes, I know, timing is an issue. The time it takes to charge is an issue, load is an issue, and many, many more things to take into consideration.
I want to try the scalar charger with an SSG as a load, to see what happens and IF it works properly.
We will figure it out someday...
I owe you a keg!

@leroy: welcome back!

Valentin

[Matthew Jones]

@SeaMonkey

How do you decide inductive over protection and under protection?

Matt

[ldissing]

@leroy: welcome back!
Valentin

Thanks to you, John K., Mathew, Stevan C. and a few others for the nice welcome back.

@Mathew,

Yes, I know of your exploits and the ability to use relays...done it myself and seen some good stuff from them at 8 ms (50% duty) didn't try higher, because I was trying to get them to go as fast as possible without screwing up.

@All,
One of the reasons I built the battery charger/discharger was to see what happened on the battery at different frequencies and duty cycles. It is interesting to see the wave forms on the charging battery side, although, right now I'm hitting them with a pulse off a coil, so the voltage is pretty high (SSG style but computer controlled), but I haven't measured the source side yet. I'm trying to get a feel for the batteries natural resonance and I can see both capacitive and inductive behaviors. I'm still not sure which we are looking for here with the TS, but it is an insight into how the battery responds to current and voltage. I'll play with this for a while to understand when a battery is in charge mode, how frequencies change the behavior (if any) and voltage/current relationships. The PIC I'm using is a 16F887 with PWM, so that is how I'm using it...no need for code to get in the way of the frequencies...and it is spot on as far as frequency and duty cycle. I've run 10000Hz at 10% to 50% duty and the scope says it is dead on at those frequencies. Tested some others and I can do about any freq down to 488Hz or lower if I use the slow oscillator that is built in. 8MHz is the one I'm using, but I can also use a 32KHz oscillator on the test board, so I could go pretty darn slow.

It is really a test bed and battery rejuvenater based on some of JB principles, but it is different too and is not in any of his patents, so it was an interesting little project.

We need to understand how the batteries react to what we are doing. When we can understand how they are reacting to what we are doing, I think we can have a go at building the system correctly.

Hope to contribute something soon. Anybody in the Dallas area with a UV light for making PCBs?


Leroy

[SeaMonkey]

@SeaMonkey

How do you decide inductive over protection and under protection?

Matt

Decide first whether you wish to 'capture' the
EMI/RFI or simply 'dissipate' it to protect the switching
MosFets and associated circuitry.

Diverting the spikes into an energy collection array
or a battery will generally render it harmless.

If you wish to use a transient energy absorber for
a particularly troublesome 'spike' then select one
with a breakover voltage equal to the device being
protected, or lower.

For a MosFet rated for 100 Volt service, use a
transient absorber rated for no more than 100
Volts breakover.

Most MosFets (and transistors for that matter)
will exhibit an 'actual' breakover (or avalanche)
that is slightly higher than their rated values;
the transient absorbers will have a precise
breakover voltage as listed so will grab the
'spike' before it hazards the MosFet.

Snubbers may also be used in select circuit
locations to minimize spike effects. Usually
to prevent 'false' switching or 'latchup.'

- - - - - - -

Has anyone yet done the lead acid battery
pulse discharge test in order to identify the
'overvoltage' pulse which results?

To do this you'll need a high current MosFet
to switch battery current into a 'load' of
0.5 Ohms. Drive the MosFet with a pulse
generator (and appropriate Gate Driver
chip to get sharp pulses) with varying pulse
widths of from 5 microSeconds to 100 microSeconds
at a frequency of about 100 Hz.

Pulse widths may be increased to about 500
microSeconds if needed in order to see the
best result.

"Scope" the battery terminals as the battery
is being pulsed into the 'load' to see the
Overvoltage Pulse that the battery produces
immediately following the discharge pulse.

By adjusting the pulse width you should be
able to find a maximum or peak overvoltage
pulse for each battery being tested.

Do you know what we're looking for here?

Will any who have the ability to perform the
testing and provide feedback to the project
please do so?

The test may prove interesting for other
kinds of batteries as well (NiCd, NiMH, LiPO)
to determine whether they'll have a
'resonant' characteristic.