Use for the Tesla Switch

[Matthew Jones]

I think you just use somthing like
If (IN1 = 1) then *Action*

The IN command open's the pin up to receive. The biggest thing you have to run a line to ground at at say 1k ohm resistance. Then use like 500 ohm to the pin. A voltage divider.
The pin comes on at 1 volt. So if you pumping it with 12 volt per say you have to do a voltage divider to knock the voltage down to 1 volt. I don't think the pin can handle over 3 volt on an input (But that in your pic docs for your particular version). The closer you get to 1 volt the better, because it take less time to turn it off.

Hope that helps
Matt

[nvisser]

Great idea. Then feed the trigger into pin 1?
Can you give me an idea how to write that on the picaxe.
Never done input commands before

[Matthew Jones]

http://www.energeticforum.com/renewable-energy/3609-big-joule-theif.html

Its that circiut. I used a coil that was 500 turns of 20 awg for the trigger and 1000 turns for the power. 1 inch iron core. I DID NOT use the "C" winding hooked to the bridge in the schematic.
I took the output and ran it to a 100 volt 10,000 uf cap. Big cap, but just about any cap with medium uf range, works.

Be careful it runs the cap up pretty fast and it will give you a good zap on dry skin.

Matt

[Bit's-n-Bytes]

Here is the diagram of a fet SSG. I post it here as this is where we discusses the fets
I have not done any measurements or even looked at the pulses on a scope yet, but i can tell you that it runs like hell and charge well to.
It is a low side switch like the bipolar ssg. The pulse width and current draw gets adjusted by moving the read switch in and out.
Much could still be done like use a monostable digital chip to adjust the pulse width .
Any mosfet driver chip can be used. I used this one as that is the only high, low side chip available here and I got lots of them.
A hall switch could be used instead of the read switch but I understand it uses more energy.
I could not get it to work with the normal trigger coil, but did not try to much.
The plan is to use it together with a tesla switch configuration that gets switched only when the parallel batteries reach about 14V. It will be switched by a latching relay that was provided to me by Bits together with other very nice goodies. Thank you again Jeff
More to follow.

Thanks Vissie, You could try the pic using the "pulsout 7, 200" (illustrative purpose only). This may allow you to tune for the sweet spot of the fet.

Jeff

[ldissing]

This video shows the spikes that show up on serial switch. The circiut though is not a Tesla switch its a joule thief. But the very same thing happens in the Tesla switch, if your looking at a serial switch.

YouTube - capcharge2.MPG

This is looking at them on a 20 mhz scope, which right now is all I have. I bought 100 mhz usb but nothing in the book for it told me not to ground to an IC that was hooked up to serial port. So its dead or I would show you what Bedini drew for us a while back.

When they do show up they grow alot further. This is what starts to drive the motor. That spike represent almost a doubling of the power (Mostly amperage) in the system. You can't measure it though off of a scope. It will show no potential on a meter, even a really good one.

Now here's the kicker that I keep reiterating based on experience. The only way I have been able invoke that spike is with mosfet based Non inductively protected solid state relays. The relay come with a data sheet and the data sheet does show to use inductive protect and HOW MUCH. I have several times included the protection hoping to invoke that spike. But it does not show up on the scope. And It does not show up on the power side of the switch.
I have one switch that runs off of 3055 (silicon ones I can't remember the whole number). It has motor cam that turns and contact fire small fast relays to make the base connection. At full speed a very small spike shows up. IT ran 6.5 months at 12 volt +- with a .75 amp +- draw, on 4x 5 amp hour batteries.

After seeing this thing several time I am certain I know what to look for and how to invoke it but you have to have something small or if you go large you have to have something that holds up.

I have tried 2 time now to use a small automotive mosfet. IXTA200N. I used them is circiut for PWM that runs my Hydro system on the car. I had extra and tried them out.

They don't hold up to even the smallest load unless inductively protected. But at that rate I never seen any response from them that would expected.
They may not be good for this application. They work in my PWN though real well.

So thats the delima I have when scaling up. Transistor at low cost do not allow for a large load. And just about everything misses the ball completly.

How to invoke it???

You take 4 batteries and charge them up. Fill them to the brim. Start your switching and run the Tesla switch with the a large load. Put the cycles into a mode that allows one set of batteries to climb fast and the other to discharge fast. IE. One set is at 13.5 and the other at at 12.50. If you are switch at 10hz then they should almost immediately switch positions. If you can get it going faster then do so. After about 10 hertz is when you'll see the spike start showing up. The faster you go the bigger it gets, If you can keep it going you will notice your load is running harder at your potential at the "Bridge point" is very low.

Now one thing I do to all my batteries is I run them on my simple motor until the charging from it give me a 1:1 discharge charge ratio. This I believe lowers the batteries impedance at higher charging voltages. It formats the battery if you will. A monopole does the same thing maybe better but it takes longer to see the effects. IMHO. Unless of course you a big ten coiler or something.

Most of my switch's are like the Brandt switch wiring. Other than the ISCC which is strictly a slow switch for extending the use of the power in a given set of batteries.

Thats all the beens I got to spill.

I have several switch's going at any time. I have no problems building something simple or complex and I have no worries about the cost at this time.

If for instance we could get s joule thief running a mosfet switch to start acting like a JT with a MJL21194 which is what is shown in the video, then I could transfer that to a TS real quick. So could other people. I have no worries about what I use as long as it shows those results, and can be scaled up.

So it someone else's turn now, tell me what works, please !!

Matt

So I can understand what you are doing in the vid...can you point me to the post that describes the setup you are using generate the waveform?

Thanks,

Leroy

[nvisser]

Mosfet driven SSG

Here is the diagram of a fet SSG. I post it here as this is where we discusses the fets
I have not done any measurements or even looked at the pulses on a scope yet, but i can tell you that it runs like hell and charge well to.
It is a low side switch like the bipolar ssg. The pulse width and current draw gets adjusted by moving the read switch in and out.
Much could still be done like use a monostable digital chip to adjust the pulse width .
Any mosfet driver chip can be used. I used this one as that is the only high, low side chip available here and I got lots of them.
A hall switch could be used instead of the read switch but I understand it uses more energy.
I could not get it to work with the normal trigger coil, but did not try to much.
The plan is to use it together with a tesla switch configuration that gets switched only when the parallel batteries reach about 14V. It will be switched by a latching relay that was provided to me by Bits together with other very nice goodies. Thank you again Jeff
More to follow.

[SeaMonkey]

SeaMonkey,

I liked the description of how the inductive discharge can differ in low and high impedances. Ive been fiddling around with different loads on the back end of the SG and this seems to fit my understanding.

I like to take a fresh view however on the comment "collapsing magnetic field" however true it may be. There is just something in that description that seems so plain and boring to me. I prefer Eric Dollards description:

Very interesting (and dangerous) phenomena manifest themselves when the current path is interrupted, thereby causing infinite resistance to appear. In this case resistance is best represented by its inverse, conductance. The conductance is then zero. Because the current vanished instantly the field collapses at a velocity approaching that of light. As EMF is directly related to velocity of flux, it tends towards infinity. Very powerful effects are produced because the field is attempting to maintain current by producing whatever EMF required. If a considerable amount of energy exists, say several kilowatt hours (250 KWH for lightning stroke), the ensuing discharge can produce most profound effects and can completely destroy inadequately protected apparatus.

Erics descriptions almost makes me think the inductor has become a signal generator that rapidly rises to a ridiculous frequency in an attempt to maintain current flow in the inductor. Maybe this is one and the same, he uses the terms in his sentence (field collapses) but a little more description in HOW it collapses I feel provides much insight.

I saw a chap using a mosfet on an SG once, he didnt use any drivers, just a reed switch to connect the battery directly to the gate at the correct time. Not sure how well this works, but it rotated and gave inductive kick, so he was half way there I guess. I guess you probably dont have the time to teach it to us all, but something to replicate would be a start. We could then begin our own study, develop our own understandings, and perhaps one day agree with you


Regards

I agree - the explanation by Eric Dollard is
illuminating. Indeed, if the 'switch' in its
'off' state is a near perfect insulator then
the flyback potential will increase to an
impressively high level. That it happens
at the 'speed of light' is no exaggeration.

It is possible to use a MosFet without a
driver chip - providing the Gate Capacitance
is charged and discharged suitably fast by
the powering circuit. The on and off transitions
are controlled by the speed with which the
Gate terminal is brought to 10 Volts (ON) and
decreased to 0 Volts (OFF.) Since the Gate
functions as a small capacitance there is a
significant charge and discharge current which
must be provided by the 'driving' circuit as
quickly as possible. There are a number of
very excellent Data Sheets and Applications
Notes which provide detailed information on
the techniques of attaining high speed switching
with the MosFet.

Here are few links to get started looking:

International Rectifier - Technical Information Application Notes Design Tips Papers

Application Notes - Fairchild Semiconductor

Fairchild Semiconductor Reference Design

International Rectifier - Application Notes

There are many others provided by the other
semiconductor manufacturers.

Peruse the web sites of the manufacturers to
look for technical resources/documents.

[Matthew Jones]

This video shows the spikes that show up on serial switch. The circiut though is not a Tesla switch its a joule thief. But the very same thing happens in the Tesla switch, if your looking at a serial switch.

YouTube - capcharge2.MPG

This is looking at them on a 20 mhz scope, which right now is all I have. I bought 100 mhz usb but nothing in the book for it told me not to ground to an IC that was hooked up to serial port. So its dead or I would show you what Bedini drew for us a while back.

When they do show up they grow alot further. This is what starts to drive the motor. That spike represent almost a doubling of the power (Mostly amperage) in the system. You can't measure it though off of a scope. It will show no potential on a meter, even a really good one.

Now here's the kicker that I keep reiterating based on experience. The only way I have been able invoke that spike is with mosfet based Non inductively protected solid state relays. The relay come with a data sheet and the data sheet does show to use inductive protect and HOW MUCH. I have several times included the protection hoping to invoke that spike. But it does not show up on the scope. And It does not show up on the power side of the switch.
I have one switch that runs off of 3055 (silicon ones I can't remember the whole number). It has motor cam that turns and contact fire small fast relays to make the base connection. At full speed a very small spike shows up. IT ran 6.5 months at 12 volt +- with a .75 amp +- draw, on 4x 5 amp hour batteries.

After seeing this thing several time I am certain I know what to look for and how to invoke it but you have to have something small or if you go large you have to have something that holds up.

I have tried 2 time now to use a small automotive mosfet. IXTA200N. I used them is circiut for PWM that runs my Hydro system on the car. I had extra and tried them out.

They don't hold up to even the smallest load unless inductively protected. But at that rate I never seen any response from them that would expected.
They may not be good for this application. They work in my PWN though real well.

So thats the delima I have when scaling up. Transistor at low cost do not allow for a large load. And just about everything misses the ball completly.

How to invoke it???

You take 4 batteries and charge them up. Fill them to the brim. Start your switching and run the Tesla switch with the a large load. Put the cycles into a mode that allows one set of batteries to climb fast and the other to discharge fast. IE. One set is at 13.5 and the other at at 12.50. If you are switch at 10hz then they should almost immediately switch positions. If you can get it going faster then do so. After about 10 hertz is when you'll see the spike start showing up. The faster you go the bigger it gets, If you can keep it going you will notice your load is running harder at your potential at the "Bridge point" is very low.

Now one thing I do to all my batteries is I run them on my simple motor until the charging from it give me a 1:1 discharge charge ratio. This I believe lowers the batteries impedance at higher charging voltages. It formats the battery if you will. A monopole does the same thing maybe better but it takes longer to see the effects. IMHO. Unless of course you a big ten coiler or something.

Most of my switch's are like the Brandt switch wiring. Other than the ISCC which is strictly a slow switch for extending the use of the power in a given set of batteries.

Thats all the beens I got to spill.

I have several switch's going at any time. I have no problems building something simple or complex and I have no worries about the cost at this time.

If for instance we could get s joule thief running a mosfet switch to start acting like a JT with a MJL21194 which is what is shown in the video, then I could transfer that to a TS real quick. So could other people. I have no worries about what I use as long as it shows those results, and can be scaled up.

So it someone else's turn now, tell me what works, please !!

Matt

[Bit's-n-Bytes]

This one with mosfet

The original circuit configuration of Ron Brandt
used two banks of 3 batteries each (see thumbnail.)

Each bank of 3, when series connected for 36 Volts,
would discharge through the opposite bank, parallel
connected (12 Volts) with germanium diodes, and into one
half of an output transformer primary. By controlled switching,
the banks would alternate, each providing one half of the
output cycle (24 Volts) to the transformer while 'charging'
the opposite parallel connected bank.

While charging batteries which are parallel connected is
not the ideal, I can understand why Ron used that sort
of connection. The three batteries together would
present a high current path with much less loss while
receiving the load current 'charge' pulse thus making
maximum power available to the load.

Ron said he adjusted the switching frequency until
he hit a battery 'resonance' at about 900 Hz.

While the batteries would eventually run down and
require external charging, equalizing and desulfation,
they would power his automobile for about 400 miles
of driving on a charge.

By the addition of an 'H' bridge to produce alternations,
it would be possible to re-configure the Brandt circuit
to achieve low frequency electro-mechanical bank
switching with a higher frequency 'battery resonance'
output frequency. That is, if an alternating current
output is desired.

For pulsating DC the solution is even simpler.

Yes, for maximum efficiency it is customary to use a
MosFet with the lowest suitable voltage rating for a
given application. This is to assure that the Rds(on)
will be the very lowest possible for minimal conduction
losses (heat generation.)
To protect the MosFet from an overvoltage transient
(flyback) it is necessary to provide some means of
protection. Once the MosFet goes into avalanche
it will run very hot and potentially self-destruct.

Needless to say, I've 'melted' one or two because of
this need for adequate protection. Fortunately, the
protective measures are not difficult and can be as
simple as a transient absorber from Drain to Source.

Similar to the measures taken to protect the bipolar
transistor from 'punch-through' transients.

@chainmailleman, this one with mosfets or any others that SeaMonkey may have..

Jeff

[chainmailleman]

Bring it on SeaMonkey, You are in the company of some of the finest Entrepreneur's in this world, and the folks specifically on this forum and thread are always up for a challenge. We have shown to the world our accomplishments and have made huge strides in achieving our goal. So let's go--Give us your thoughts in a schematic. Let us know what parts you want to test. Heck, half of us may have the parts sitting in our labs, (I have three, if you'll allow me to count the kitchen table), but I am ready and I know a plethora of engineers that are also ready, so in your words "Once someone is ready to take on the challenge" I am that someone, just give me a schematic.

Jeff

What type of schematic are you looking for?

[Bit's-n-Bytes]

My turn

One way or another it will be done. You'll either
realize how simple it is and do it yourselves (ideal)
or we'll work together to arrive at a workable
solution (also ideal.)

Once someone is ready to take on the challenge.

Bring it on SeaMonkey, You are in the company of some of the finest Entrepreneur's in this world, and the folks specifically on this forum and thread are always up for a challenge. We have shown to the world our accomplishments and have made huge strides in achieving our goal. So let's go--Give us your thoughts in a schematic. Let us know what parts you want to test. Heck, half of us may have the parts sitting in our labs, (I have three, if you'll allow me to count the kitchen table), but I am ready and I know a plethora of engineers that are also ready, so in your words "Once someone is ready to take on the challenge" I am that someone, just give me a schematic.

Jeff

[SeaMonkey]

@SeaMonkey:

If I wasn't so stupid, I wouldn't be here! If I was brilliant, I'd say...this is HOW you do it and WHY.

Show us the circuit, tell us how and why it works, then you can begin your lecture and instruct us on WHY you are THE man with the knowledge.

Let me clear up a few points....
1. I could NOT make a circuit that would get cold unless I put it in the fridge.
2. I understand and agree completely with the inductor resisting current change and HV potential spike upon cessation of current. I have been saying this for several years.
3. I agree that MOSFETS have superior switching characteristics in terms of speed and amp carrying capability. At least my study of them seems to indicate this.

So, where do we start. JB talked about negative resistance in this forum. He said that Naudin showed it, but had not figured out how to use it. He did not say anything more on Naudin, except that he thought we could figure it out. I do not believe that I have figured that out. I know how to make a BJT "go negative" but have not figured out how to use this to my advantage.

I do not know if the MOSFET can act as a negative resistor (negistor). I have never designed a circuit with any type of FET. Is it your contention that the negistor is "not needed" or that a MOSFET can be a negistor?

You indicated that Dave's circuit was too complicated. Why do you not suggest where it is too complicated and suggest alternatives. (The circuit is actually fairly simple and the controller part of the schematic was not shown in the post). The controller is actually more complicated. Is it that you do not think the 4 battery TS is useful? You said in one post that it was "old technology"....does that mean that should not be studied?

JB indicated that he preferred to use BJTs, not FETS in this application. Perhaps JB was not telling us because he did not want us to use MOSFETs. Perhaps MOSFETs would be superior and work better? This seems to be what you are saying, although, I have not "studied" everything you have written in this forum, but read casually through it.

Perhaps you can help me to understand finesse. Perhaps you can "show" me how to test the circuits I build and understand what is going on with limited resources, i.e. equipment, time, money, brain power and intellect.

Like Matt says, I am not opposed to trying new things. But again, a picture us worth a thousand words, i.e. a circuit. Budget does have constraints though.

Leroy

You're a very cerebral man with a ton of questions!

Very good!

The 'Charge Pump' is old technology - but still very
useful.

Daves schematic is 'unnecessarily' too complicated.
There are simpler 'solutions' to this problem.

The 6 battery Brandt arrangement is the better one.

Negative Resistance (except for the Lead Acid Battery)
is unimportant. Many devices can be made to
demonstrate a 'negative resistance' curve but with very
little practical value.

I collect the PV units from small yardlights. Find neighbors
who are wanting to dispose of some old 'no longer working'
units and 'cannibalize' the small solar battery. While most
are only capable of 40 mA or so of current, several of the
small solar batteries series connected can provide substantial
power for experimental purposes. I really like CHEAP!

In due time! Was ROME built in one day?

[SeaMonkey]

So,
no one is actually interested to the TS-PVAmp schematic I posted anymore?


Or, are we ready to try compare the actuall losses BJT way vs MOSFET way and move on any time soon?

And are there any ways to circumvent the MOSFET inherent flaw of onidirectional current braking, so we could fanally use it's inherent advantege of ultra low FVD at at least four times the AMPERE density a BJT has?

Or could we find BJTs that are 40V/60A-DC rated that handle 120A surges and have a hFE of at least 100@60A-DC while below 2V (120W burned out of 2.4kW passing) VCEsat?

Or shall i seek other company?


Stevan C.

Nope! Stay where you are.

We'll get there in due time!

[SeaMonkey]

Now if it where me walking in on the discussion in which I though I could contribute I would come with example and the willingness to at least understand that the existing crowd of people probably know a little about what they are doing. If I was convinced my way was better I would go far out of my way to prove it. Putting a small circuit together is not that big of challenge.

So I hope you can understand when I say, I am not trying to burn anything, SeaMonkey did that when he walked in here. And If I opinion is skewed in someway I apologize to all.

n the three years I been hear this typical type of person does not contribute in the long run they only delay. If the goal is to contribute simply hand drawing a proposed schematic and stating case is all that is needed. I am always willing to try something new. Thats the primary reason I have so many builds. I am always willing offer an civil explanation for or against any proposed project solely based on experience. But If my understanding of some thing are flawed then I can easily alter that. And again I'll try anything.

Like I said this is the last time I'll get into this. Educate matters, people are not so willing to change the course of the research they are doing based on a few pretty words and some subtle insults. No one should for that matter....

Matthew Jones

Matthew,

I understand the source of your frustration
and your emotionalism.

Making 'OU' work is not simply putting together
a circuit that someone provides. Bedini has made
that very, very clear.

One must be prepared to devote considerable
time to truly understanding what it takes to squeeze
maximum performance from the 'transistor' or the
'MosFet' or the 'IGBT' or the 'chip' or whatever...

And we must break out of the realm of false beliefs
such as:

"only THIS transistor will do it!" or

"the diodes are reverse biased to permit RADIANT
ENERGY FLOW" or

"OU is a whole different animal than conventional
electricity or electronics" or

"you gotta use the original parts" or

"more turns on the coil means more energy."

It is so easy for those who have deficient technical
understanding to fall for the false dis-info which
abounds and to look for answers in the wrong places.

Often spending large sums of money for 'parts'
which are far too costly and completely unnecessary.

That you have done much work, and continue to
research and experiment, is commendable.

Hopefully, you will - "And again I'll try anything."

[SeaMonkey]

Now SeaMonkey the question you haven't answered directly is WILL YOU HELP DESIGN A MOSFET CIRCUIT?

Mark

One way or another it will be done. You'll either
realize how simple it is and do it yourselves (ideal)
or we'll work together to arrive at a workable
solution (also ideal.)

Once someone is ready to take on the challenge.