Large Toroidal Coil

[Peterae]

OK i have weird things happening now, I found There seems to be a frequency and phase combination that lights a 240VAC 5 Watt bulb dimly connected to the 4 turn coil around the periphery, if i alter the drive frequency then i have to relocate the phase to get it to light.
The weird things that have been happening are strange low hertz buzzing from somewhere near the coil, i don't seem to be able to pin point the noise, and my scope is occasionally doing the strange waveforms, but when the scope shows the waveform the bulb does not get any brighter, if i disconnect the usb lead to my battery powered laptop the bulb get's just slightly dimmer, if i disconnect any of the scope BNC's from the scope the bulb goes out.

I cannot rule out that the scope trace is showing garbage, although before when i first discovered the effect and took the shot up the top of the page the scope screen was changing in time while showing High peak voltages.

[Grumpy]

grounding through the scope?

Try an earth ground to the ring.

[ion]

Peterae....a couple of observations and precautionary measures:

Your driven coils have very low inductance, probably less than 100uH, as such you need to insure that the driving pulse width is extremely narrow or you will generate excessively high peak currents and any glitch from your generator source will fry the FET's.

With such available noise and EMI in the vicinity of your pulse generator, there is a possibility it could glitch and restart, possibly causing a wider pulse delivered to the FET's cooking them. I recommend some small lamps or resistors to limit the current or set the current limit feature on your power supply to a low value.

Your diodes look like they are twisted to the windings but not soldered. If these diodes lose connection for an instant, the flyback pulse can be rather high exceeding the FET voltage rating, possibly breaking down the FET.

As always, I recommend a 10 volt Zener from gate to source to clamp any EMI induced energy and keep it from breaking down the gate.

[Peterae]

Thanks for the advice ION, we are running ok at the moment,  the pulses are about 100nS wide after rise and fall delays in the fet and driver.

@ G the scope is not earthed nor is the laptop it's running from a battery.
Interestingly i can stop that light bulb from lighting by touching one side of it with my hand.

[EMdevices]

I like the infinite loop approach, and coupling to it from both ends with short pulses.
We know from EMI board design rules that pulses running in parallel traces couple to each other.
Now, DC pulses injected will induce DC pulses in the adjacent infinite loop, traveling in opposite directions.
The right frequency will cause resonance and huge DC pulse buildup.
I would suggest adding 10 ohm series resistors to the FET drains, for some impedance matching and safety.
Also, like ION suggested, protect the FET gates with Zener diodes.
Don't think only unipolarity, but consider what would happen if the Drain went to -300V while your Gate is at ground potential, think about it.

[MileHigh]

Peter:

I have a few suggestions for your consideration.  I am assuming that you have a main drive coil and then the other coils are considered secondaries.

If you have a signal generator then put a pure sine wave into the main drive coil and then observe the outputs on your secondary coils.  You can change the frequency and you can also experiment with different resistive loads on your secondary coils.  Sweeping the frequency will give you an idea of the frequency roll-off in the secondary coils.  Changing the load resistors will give you an idea about the power coupling between the drive coil and the secondary coils.

Then switch your signal generator over to square waves and experiment some more.  When the square wave frequency is low enough you will see on the secondary coils a kind of double-spike where each spike has an exponential decay.  You can try seeing what happens with different load resistors here also.  It's actually possible to measure how much energy can be transferred between the main drive coil and the secondary coils like this.

If you do this, at least you will have a sense of how the coupling looks like between the drive coil and the secondary coils in terms of frequency response, phase, and coupling.  You don't have to get too fancy here or make any precise measurements like I alluded to.  Rather, the idea is to get an innate sense of how it functions through basic experimentation first.

Once you do that then go crazy with your nanosecond pulse generator and stuff like that.  The simple testing should help you understand the results you get with the pulse generator testing.  The most important idea is to start simple and then get more complicated afterward.

MileHigh

[Peterae]

Cheers EM i have diodes across the coil and fet drain/source which would hopefully stop any negative swing across the fet.

I am really annoyed i have not been able to recreate the exact effect i had yesterday, i am left with doubt in my mind that the scope was telling the truth about the large pulses and dc offset, i have only seen that once today and that was with the laptop cashed out, which could mean the scope was feeding duff data to the laptop from trace drawing at the point it crashed, at the moment i need to dismiss any scope shots, and maybe pull out my old bench scope(although this is earthed)

I currently believe yesterdays effect is a combination of many parameters, wire positions, lengths, even the scope probes and laptop connection seem to be involved, phase and frequency, although the laptop and scope are powered i fail to see how any energy could be transferred into the system that i was seeing yesterday.

[MileHigh]

I'll add a little addendum:

To really understand what's going on here you need to put one channel of your scope on the source signal that's going into your main drive coil and the second channel of your scope on whatever other signal that you want to look at.  It is of prime importance to look at your output signals relative to your input signal.

I often say this:  You get 10 times more information from looking at two simultaneous traces on your scope display as compared to looking at just a single trace.  You have to see and understand the timing relationship between your input signal and your output signals.

MileHigh

[Peterae]

the setup above MH works by sending 2 very fast short pulses around an endless coil, and then delaying the phase between them by 0-1350nS, the 2 pulses are independent from each other so i have 2 drive coils each is wound on each side of the toroid.

So really the output is a result of resonance, phase of the 2 inputs and drive frequency, i don't really need to monitor the 2 drives as i know exactly what they are doing, as in pulsing with a time delay between them.

The output result in the endless coil is complex and has very little resemblance of the drive pulses, on top of that one of the output coils is wound 90 degrees from the drive coils so conventional induction has little to play, it's not so much a case of trying to predict whats going to happen, but more or less a matter of looking for anomalies on the output coils, which is what i thought i had found yesterday and then looking closer to try and understand it, i am trying to move away from convention and discover new things.

[Grumpy]

touch the coil and tell me if it tingles

[wattsup]

I think the looping coils are generating much higher frequency then your applied frequency because they will start to slightly overlap and perpetuate inside the loop. That increased frequency can cause re-coupling backwards inside the driving coil and cause a reverse action on your diode inside the fet. I have seen this recently with voltage rising very high.

wattsup

[giantkiller]

touch the coil and tell me if it tingles

Not a good idea!

Get a guppy fish in a small bowl and watch to see if the fish swims frantic or tries to get away. Or use small pet birds, flys or bees.
Put a small neomag wrapped loosely in tinfoil on a wooden stick. Move it in and out of the coil vicinity. Watch your dog perk up or shy away.
Or wad up a bundle of 6" long of 30awg wire with a 1 foot lead. On the end of the lead put a Hall effect sensor to a scope. Bring the wire bundle end in the vicinity of the coil. Look at the scope.

Or just tell us about the nuclear headaches the day after!   That warm feeling is the cooking of the microwavable synapses.

I think the looping coils are generating much higher frequency then your applied frequency because they will start to slightly overlap and perpetuate inside the loop. That increased frequency can cause re-coupling backwards inside the driving coil and cause a reverse action on your diode inside the fet. I have seen this recently with voltage rising very high.

wattsup

Exactly. This it what caused the 12mhz out from the 43khz input in my Cu, Steel coil resonate test. This is the start of the feedback loop. Not current but EM.