Lenzless resonant transformer

[Jack Noskills]

Itsu,
Opposing parallel coils and high frequency resonance is the way to go.


Can you feed L3 using your power amp ? In my testing pushing little power had no effect, I had to use more power. I quess I was using 1-2 watts. How many amps can go through a 5 watt/12 volt halogen without any sign of light or heat in it ? That was the amount of power I was pushing in.


Don Smith used NST (35 kHz, 30 mA and 2000  volts or so), only a few turn coils on a Metglass C core (high permeability). Now we know that when L2s oppose their resonant frequency will be higher so 35 kHz drive is reasonable with only few turns. Drive signal was passed via diode so it was clipped to half cycle sine, resembling pulse.


Don also used small capacitor parallel to L3 and spark (GDT) between C and L, the well known Tesla circuit. So he was smacking the L3 with a strong pulse which gives amps and voltage for a short period of time. This will make L2s ring at their own (high) resonant frequency. When next kick comes in correct time (L3 subharmonic resonance) then L2s keep on ringing. While L2s ring, the L3 is disconnected from the system because spark is not conducting.


C in L3 must be small enough so that it charges enough to jump the GDT, this depends on the amps coming from the source.


Before we start messing around with sparks, we could check what happens in L2s when L3 is driven with a pulse. No capacitor needed in L3, possibly a rectifying diode is needed there but I don't know for sure. Purpose is to see how long those L2s ring, to get impulse response of the system. Width of the pulse could be varied. Frequency of the drive could be sub harmonic of L2 resonant frequency. It should be interesting to see what happens in L2s when the next pulse comes from L3 before oscillations have died out in L2s.


Now there is some DC resistance so I don't expect L2s to ring for long, but still longer compared to just one L2 coil only. Maybe this could be compared ?


There are two options, two LC tanks in the output or one LC tank combined from two coils. I think two separate L2s are better when driving with a pulse.


If results are interesting (tubular L2 waveform) then maybe it is time to make second similar setup but using thickest wire you got and only few turns for each coil, 10 turns is a nice round figure. If DC resistance of the coil is dropped 100 times then DC resistance is not eating up the power any more in significant amount, Q-factor will also increase giving better resonant rise.

[MileHigh]

Jack:

I don't have the right to speak for Itsu, but let me say my piece.  This one is a done deal, it's dead.  Your last posting was over the top, and all over the map, you are designing a new setup by trial and error.  At this point I suggest that you build something yourself.  If you have theories about how some kind of transformer circuit will operate, then draw up some timing diagrams and then check the operation of the circuit against your timing diagrams.  I know that "timing diagrams" is a hot potato term because people can talk a blue streak about the hypothetical operation of a circuit, but then when you ask them to make up a timing diagram they clam up right away and say nothing.  It means they are not capable of taking their "talk the talk" and actually walking the walk.

Pumping more power into a circuit will not make it all of a sudden work.  When you play with transformers you could end up saturating the cores and then the currents will jump up and get very high.  I saw on another thread where you posted and you have some quite radical ideas about transformers and related stuff and how they work.  The honest truth is that you can't get more power out of a transformer than you pump into it.  It might be a bitter pill to swallow but it's the truth.

I am not in any way trying to discourage you from doing tests and trying out circuits and all that jazz.  The key thing is that the circuit won't ever lie, it will simply do what it is supposed to do.  It's interpretation of what it is actually doing where things sometimes go amiss.

Going back to one of my earlier points, suppose you dream up some simple transformer circuit.  The challenge for you is to work out the timing diagram(s) of how it will operate _before_ you hook up your signal generator and scope probes to it.  If you can't do it that should, in the best case, motivate you to start surfing the net or buy some books, etc.  It makes no sense to design a circuit and then claim that you can't make up a timing diagram for the circuit.

Let's assume that you make a timing diagram for a circuit and then you build it and check it out with your scope.  If it checks out and the timing is as per your timing diagrams, good for you.  If it doesn't check out and your timing diagrams are way off, then that should motivate you to start probing around with your scope probe until you figure out the operation of the circuit.

Trust me, these kinds of circuits are all about the timing.  Timing is king and if you don't understand the timing, then you don't understand the circuit.

MileHigh

[Jack Noskills]

My problem is that I don't have a scope and cannot get one.


So I was asking for impulse response of the L2 coils so we all could learn from it as we have learned so far. Waveform should be tubular in this case. Why ? Because in plain LC circuit the bell shape waveform does not come from DC resistance alone, it comes mainly from self induction. For example, when DC resistance of a coil is close to zero, then ringing still dies out fast because of self induction. Higher the L then faster the oscillation dampens. Information that I have read from text books say oscillation dampens because of DC resistance only, not a word on self induction. Here we have two L2s working together and inductance is in micro henries so it will oscillate many times while at the same time they create energy in the series capacitor, which is then used by the load. Bigger the capacitor more energy we can collect.



Last night I figured it out how Don used the drive, it is not just a plain LC with spark gap but a little more. We already have L3 that rings with the same frequency as L2s and when power is taken L3 stays in resonance. This is the key and these were shown in Itsu's videos. Now this C3-L3 combo needs to be kicked with a pulse, it starts ringing and then it can be kicked at some lower subharmonic of resonant frequency. So the drive is basic Tesla type C-spark-L with steroids, L is replaced with L3-C3 and there must be diode after the spark. Otherwise C3 in the L3 will discharge over spark back to C on the other side.


Of course spark can be replaced with a solid state switch, then there is no need for pulse drive but DC is enough. The solid state switch then controls how kicks are applied to L3-C3.


Result is that there will now be free oscillation in L3-C3 tank and it will oscillate many times for free because it's L is low => self induction is low => oscillation dampening is low. Little power is then needed to keep this oscillation going on to feed the L2s which create the power which can be used.

[itsu]

Jack:

I don't have the right to speak for Itsu, but let me say my piece.  This one is a done deal, it's dead.  Your last posting was over the top, and all over the map, you are designing a new setup by trial and error.  At this point I suggest that you build something yourself.  If you have theories about how some kind of transformer circuit will operate, then draw up some timing diagrams and then check the operation of the circuit against your timing diagrams.  I know that "timing diagrams" is a hot potato term because people can talk a blue streak about the hypothetical operation of a circuit, but then when you ask them to make up a timing diagram they clam up right away and say nothing.  It means they are not capable of taking their "talk the talk" and actually walking the walk.

I have once again to agree with MileHigh here.
When you dragged in Don Smith and all these idea's about pulsing etc. i knew you need to set up your own bench and start
experimenting yourself as you are still full of vague concepts of transformers and abnormal behaviour.

I wrote up a response, but decided to sleep over it as i don't like to make impulsive reactions, but my opinion has not changed.
See that reaction below.

Regards Itsu

[itsu]

Jack,
I followed your last suggestion by powering the core in opposing parallel setup again but now with my power amp.

The problem with this is that for the PA to properly work i need to transform the low 4 Ohm output to a higher value (L3 at 32KHz = 200 Ohm)
As i have only a fixed transformer (4 Ohm to 6600 Ohm) we introduce yet another mismatch.

The bulb in the output now is much brighter, but probably due to the higher filament resistance or whatever, the resonance frequency
shifted from 17KHz to around 30KHz.

Input into the bulb about 1W.    Input into the PA is 105W!

Video here:  https://www.youtube.com/watch?v=zsHu0wOZSko&feature=youtu.be


I think it is pointless to continue this as all seems to be perfectly explainable and no abnormalities can be found.
Its becoming a neverending story with continued side steps and i see you still have a lot of new ideas, but i don't
want to invest any more of my time in it.

My advice is that if you are convinced that there must be something special you invest some money in a decent oscilloscope
and start testing/measuring yourselve.

I will continue to explore other things which will use up the little free time i have.

Regards Itsu