Use for the Tesla Switch

[nilrehob]

I will get back on this as soon as I finish the Multi Fuel engine I am designing now.

Hi Dave! Thanks for the reply.

It would be really interesting if you could write more about your experiences with your TS circuit. The why's and how's sotospeak.
I pursuing another trail right now, but I check this thread at least once a week, I can't really get the TS out of my head it seems :-P .

/Hob

[nilrehob]

Anyhow YES I have built this circuit and it works great !

I might as well ask some questions, if you don't mind:
* can you describe the benefits of the circuit?
* at what frequency do you run it?
* how long is the rest between "on"'s (or is it going straight from one "on" to the other "on")?
* does it require inductive load to work in the beneficial manner?

/Hob

[DoubleD]

I might as well ask some questions, if you don't mind:
* can you describe the benefits of the circuit?
* at what frequency do you run it?
* how long is the rest between "on"'s (or is it going straight from one "on" to the other "on")?
* does it require inductive load to work in the beneficial manner?

/Hob

HI Hob.

Ok I will try to give you a quick run down on the circuit ....

What I was after in this circuit was a true solid state version that could actually do something , as in run some usable power through it.
The first big problem , that ended up taking a good year or so to sort out due to the way the circuit actually works verses how the Solid state components work, was the switching on a couple on the Fets/Igbt's. I don't remember off hand what exact ones it is as I have been off this project for a while, but the reference voltage changes polarita as soon as the gate is closed and with Fets they need a gate voltage of ~15v above what it is switching to turn on hard, but as soon as it is on the source voltage changes polarity and if you have the same gate voltage applied it will fry it real quick. so getting this worked out was a Major hurdle.

This circuit is now self powered from the batteries it is switching so no other source is needed. everything is "Floating" as seen by anything else , IE the + and - rail are full floating and don't see each other, each fet and driver only see what they are switching in a virtual state and not the whole circuit.

The PWM controller is fairly crude but serves it's purpose to get the thing working but a micro controller is needed to get it to the next level.

AS for the circuit timing ...
lets start with everything OFF ..

First one series fet turns on to guarantee a 24v supply for the gate drive circuit, it stays on for 3 pulses, during the second and third pulse the opposite side fets turn on to complete the parallel connections,so they are on for 2 pulses. Then everything is off for 1 pulse and then it switches sides and repeats.

As for the freq.... it need to be tuned to the load AND the batteries being used. the circuit is just a middle man so to speak , and the load and source need to be in tune with each other for the magic to happen. Like Tesla says Match The source to the Load.


My basic theory on this ,need the micro controller to confirm, is in short form...
The electrons outside the battery flow faster than the Ions inside the battery so if the switching was right and with the proper load, the batteries should turn into an electron pump and not even realize they are being used.

Oh I thing the circuit ASH has posted might have one Diode backwards too so I will have to check that out as it is one for the Floating common and will kill the entire circuit if in wrong.

Hope this helps for now...

Later
Dave

[nilrehob]

Hope this helps for now...

Thanks for sharing Dave, good info!

Still wondering about the frequency though: If about 10^x Hz, what x?
And the load, does it have to be inductive, something with a coil?

/Hob

[Matthew Jones]

Schematic

I drew the schematic of the one I have been using, plus one change.
kent_elyue ( http://www.energeticforum.com/renewa...html#post65267 )Mentioned I was Forward Biasing my Serial Transistor. I checked. It works either way. But I guess I should do it the right way.

The latest one I building is with 2 volt cells. That the best way to create a potential spread. So it involves alot more transistor and the like. But it can be put together pretty much modular.

The frequancy of switching can be adjust with limits as well. The solid state relays are the limiting factor. I got goods ones and they cost so...

Anyway Hope it helps.

Cheers

Matt

Schematic

[DoubleD]

Thanks for sharing Dave, good info!

Still wondering about the frequency though: If about 10^x Hz, what x?
And the load, does it have to be inductive, something with a coil?

/Hob

I don't remember what the freq, range was I built into the timer I would have to dig out the notes etc. on it but there is a wide rage built in to it. I never checked to see what the actual freq was as I was looking at other things and didn't really care what the freq was at the time. After I get the Data and Function I am after then I will work on the specific conditions that create it.

The actual load doesn't seem to matter as long as the circuit is tuned to it...
When I got the Fets to frost up I had a combination of Resistive (lights) and inductive (Fan Motor) with no caps on the output (from batteries to FWBR )

This circuit will handle a few hundred amps depending on the Fet's/Igbt's used and the batteries to match. So their is no Magic Freq. that makes it work. It all depends on the set up and actual running conditions.

When I get the programmable controller together the conditions and settings can be monitored and controlled a lot better ,then it should be able to be made auto adjusting to load, but then it starts getting to be a complex circuit that is hard to replicate.



Matthews Relay circuit looks to be a nice simple set up to experiment with but the good relays do get spendy but the little automotive ones should get a circuit running to play with.

Hey Matt ... are you still working on the mechanical rotary one too ?


Dave

[Matthew Jones]

Ya good relays especially "spst's" cost big money. 20 -30 dollers. But you can get spdt's for around 2-3 dollers a peice that will do the trick. They are pretty fast.
If you intend to run big mosfets that have to be shut of you'll need to use DPDT's so you can have both turn on and turn off voltages availble.

I am playing with a way to use 3 mosfets for switch. I know I can do it because 2 mosfets and an optocoupler are used for some solid state relays but they are in one package.

I'll get it soon.

I have several mechanical ones. They work real well too. The problem, again, is the motor that drives them.
A normal BEMF producing motor will discharge your batteries over time. And the motor does not run at full potential no matter how hard you try. Because your only running off of 1 pole in the motor, the other being in a neutral state.

So until I can build a good solid motor that will run off the power source they kinda all have the same effect. 8 -20 times more use outa the load depending on the load.

The biggest thing I want is to get one that produces some real power. Thats why I started using 2 volt cells. It allows for the greatest potential spread. I wanna run some tranformerless inverter tests. They play in a pretty high voltage range. I wanna see what kind of AC loads retain your energy.

Matt

[DoubleD]

I am playing with a way to use 3 mosfets for switch. I know I can do it because 2 mosfets and an optocoupler are used for some solid state relays but they are in one package.

Matt

Matt this IS what I have done here, The "normal" mosfet switching methods do not work with this application due to the way they are manufactured.

It took me a couple years to figure out how to get them to work in this circuit and that was even working with the fet manufacturers, they just couldn't grasp the problem because this is NOT normal circuit flow involved here.

and PNP's aren't worth a crap either .. they are too hard to turn off , they don't come in very high power , and are expensive to list a few.

The original circuit I worked with to get this sorted out had all seperate components,, Opto's drivers and fets for each switch segment. once I finally got it stable I cleaned it up and used these combined components with the Opto and the driver in one unit, this cuts down on parts count and simplifies it but makes it harder to follow the actual circuit dynamics that make it function without destroying its self.

The smaller circuits on the net using the transistors work only because there is no real power run through them so the switching problems dont really show up, BUT that is why they run so hot too.
In order to get the fets or Igbt's to work effeciently they need to be turned FULL ON FAST and also Turned HARD OFF Fast and this is not an easy task especially when the voltage potentials change as soon as it is switched.



Dave

[Matthew Jones]

I understand what your saying. Thats kinda why I was using relays in the first place. SPDT can be used backwards. So the supply port becomes the power out. Then you can have 2 supplys running in and 1 out. That way you can shuttle the correct current to the mosfet.
But relays only run so fast. I wanna try higher ranges 1khz - 100khz.
Contrary to what I have heard about high speed pulse charging, I have seen some positive results in absorbtion of charge in lead acid batterries with little to no adverse effects.

Well if you get the chance Dave I would like to see you schematic, if its possible to draw.

Cheers
Matt

[Tecstatic]

Hi Matt

Here is a "work in progress" diagram.

I have used the same components before on other PCBs.

This is capable of very fast switching.

Running a small 3-phase servo motor off a similar circuit generates practically no heat. In that case just a copper area on the PCB served as "heat sink" for the MOSFETs.

This is the coming isolated power switcher board for my experimenters set, but the final PCB will be full bridge and have its own isolated power supply running off 12VDC to supply the 4 isolated and individual circuits.

The PCB can be used as 4 separate switches or as a full bridge. A coming uController board generates the timing signals for the MOSFETs.

I think this circuit is able to go above 1MHz, I have not operated above 400kHz though. I think heat losses in the TC4420 will be the hardest limit aiming for higher frequencies.

It also depends on the gate capacitance of the chosen MOSFETs.

Eric

[Matthew Jones]

Let me get my book out and start that looking that one up. I am kinda a "If it don't start on fire, it must be working" type of engineer.

Thanks I'll definatly try to learn somthing from it.

Matt

[nilrehob]

Has anyone made a successful and beneficial TS without an inductive load?

[nvisser]

In order to get the fets or Igbt's to work effeciently they need to be turned FULL ON FAST and also Turned HARD OFF Fast and this is not an easy task especially when the voltage potentials change as soon as it is switched.

Dave

By using dedicated mosfet drivers as Tecstatic uses it is easy to switch the mosfets .
TC4420 or mc34151 can be used. The last one contains 2 drivers in one chip .
A good example how to use it is from Les Banki`s Pwm circuit intended for the Bob boyce cell.

[Matthew Jones]

Depends on the definition of succesful and benificial. I have run light on mine, but they use power. Not as much though.
If you can make a 100 watt lightbulb light up for 1 hour and only use 20 watts to do it, Is that success? Is the ability to recollect unused current from your load after its done the work benificial?

If 1hp produced for 1 hour costs you 750watt hours under normal wiring and you could for the same cost make it run 8+ hours on the same energy, is that success or failure. Then take into account you could in turn make 600 watt for an hour on the shaft (Easy).

Or do you require a success to be defined as the load runs and the power supply maintains a charge?
I don't think its possible without an inductive load. I haven't seen any indication that any other load would help reproduce any energy to maintian a charge. Tuning may be the key.

And the only successful one we here about was Ronald Brandts car motor, which drove around for years without needing a charge.

Matt

[nilrehob]

Depends on the definition of succesful and benificial.

successful = circuit tried out, load still performs, nothing burns
beneficial = load consumes less energy from battery while performing the same

Or do you require a success to be defined as the load runs and the power supply maintains a charge?

maintains a charge? do you mean overunity?

I don't think its possible without an inductive load. I haven't seen any indication that any other load would help reproduce any energy to maintian a charge.

I agree, so far it seems like all really efficient circuits contains a coil,
it's all the common basic circuit with an as sharp discontinuation of the current in the coil as possible.
If so, I think the TS might be made much simpler and more efficient.
Im planning on a new setup, but I need faster diodes and they haven't arrived yet.

/Hob