Understanding electricity in the TPU.

[wattsup]

@groundloop

Thanks for your great help in getting this circuit up and going. I made mine and put two photos below. It does not look like your since I am making everything on a terminal block and no soldering in case I blow things. Force of habit. lol

I am using two IRF840 (n-channel) and one IRF9540 (p-channel) and unfortunately my EE surplus place does not have those fast switching diodes but I will find them. For now I am using standard didoes but the circuits seems to work OK. I can see pulsing on both sides of the p-channel drain and source, something I could not do before so this tell me it works.

For me, this will open a whole new world of pulsing potential. I have been thinking of the optimal coil build for this pulsing method.

Once I have mastered the circuit usage I will make a video to show guys what this thing can do. More fun.

Thanks again.

wattsup

[Groundloop]

@wattsup,

Great build! Yes it is smart to use a "rat nest" configuration when testing. I just blew one of
my mosfets and has to solder in a new one. :-) Just remember that if you want to switch at
high frequency then you need to keep those wires as short as possible. Also remember that the
maximum current you can switch is limited by the diodes used. Also, this circuit will start to
bias the transistors at around 5,6 volt (4 volt with signal input), so keep the Ub voltage at a
safe level.

Groundloop.

[wattsup]

@Groundloop

Thanks for your e-mail and will post here an answer.

Been around just reading threads while I test this circuit.

Have tried it on a variety of coil configs but I think the pulsing is not working as it should. Putting my scope at the D and S of the P-channel mosfet, since this is where the main switch transfer to the coils will occur, I see both waveforms are pointing in the same direction, or, rising and falling together, but it seems to me they should be opposite. One should be rising and other falling.

Also, applying DC input does nothing to the circuit so this is telling me the mosfets are not pulsing the way they should.

I am getting the same led lighting results just with the FG positive (no negative connected - lol).

So there is something wrong.

I still ave not found a local supply of the BYV29-500 diodes but I did try a whole host of different diodes and this does in fact make major difference in output but still no pulsing of the mosfets. Diodes are very important.

I will have to make a video of this so you can get a real sense of how I am testing this.

[Groundloop]

@wattsup,

Thanks for taking time to answer my questions.

The first thing that pops into my mind is the coils. Have you tried to invert one of the
coils? This will invert the negative kickback from the coil. Also, how big are the coils.
Did you select coils that have a resonant frequency for the FG frequency you use?

>>Also, applying DC input does nothing to the circuit so this is telling me the mosfets
>>are not pulsing the way they should.

This is worrying. In my circuit I could "pump" enough power into the coils to
light up a small light bulb. (Must add that I needed low frequency and lost
of turns on the output coil.)

What DC voltage did you use on the circuit?

I do not have a o-scope at home right now but I will borrow one and test a little
more on my circuit this weekend.

Groundloop.

[wattsup]

@Groundloop

Thanks for your response. Much appreciated and also all the work you put into getting this "starter" circuit. Starter because I think this will eventually become more complicated.

Although the initial idea of having two coils wound in the same direction with a switching point at dead center seems simple enough, it is actually more complicated then one would expect and kind of hard to grasp the overall concept of how these coils really act at their switching points.

You see, when the coils are connected in series it makes one long coil and the p-channel D and S are actually swimming in the coils blotch region where polarity is very difficult to define in absolute terms. Those D and S points need a definitive polarity to switch.

When the Gate has no pulse on it, the D and S are not connected just like the two n-channels are not connected, so in this off pulse, the two coils are not connected and cannot actually make one center blotch wall. When all the mosfets get a pulse, then all are connected and I think the p-channel is then acting like a master bypass canceling the n-channels attempt to bring new polarities to the switch point. So all the circuit is doing is connecting and disconnected the blotch wall without bringing new polarities to the switch points.

Man, this is very complicated because it is not like just making a specific circuit for a specific linear task. You need to have a more in depth understanding of what is happening at those switching points. I guess that is why mosfets are mainly used at the coil ends and not in the middle. lol

A question I have is..........

Is there a mosfet that works in reverse and I don't mean the D and S going from positive to negative or from negative to positive. What I mean is when there is no pulse, the D and S are connected, instead of being non connected at no pulse. Like in a normal relay has N/O and N/C contacts, is there a mosfet that can have one or the other, or are all mosfets made to work the same way. Always N/O.

What a freak out this is.