I've got a 12V relais. a small one.
I want to connect it to 12V battery and let it make and break its own supply.

the back emf of the coil should go into a step up transformer. all in series.

the secondary probably will have a to high resonamt frequency, but it could be tuned down with a capacitor over it.

one side of the secondary should be to ground.
the other side rectified into a capacitor, and then back into the battery to recharge it.

the triggering frequency of the relais should be matched to the resonant frequency of the secondary. so i will try diffent capacitors.

A few words on using a relay such as this as a make and break device.  First, the contact separation
is very small so any high voltage spike will jump the gap like a spark gap.  Second, the second you
discharge a cap through a low resistance coil you will weld the contacts together and bye bye relay.
Third is that the relay coil will not allow you to drive it fast before it turns into a steady on device. My
advice to you is to use a Vibrator.

thanks chuff1 . it might brake but Thats Ok. maybe I can redirect the bemf so it wont damage it.

Reed switches are also good while experimenting since they are sealed in glass and usually contain some kind of inert gas.  You can get several million open and closures before the contacts become inoperable.  And while experimenting, you can look for the ominous green glow of the spark; that will be a big clue as to whether you're heading in the right direction or not.  Just keep the power levels down and watch for the effects you are after.  We can always figure out how to scale-up once the system is understood.

agreed. thats why I learned so much. I played safe.
but still... got burned and shocked. hahaha.

upscaling to higher voltages might reveal more effects. like the pressure felt by hand, and at the same time metal attraction.
weight/time shifts?
sparks in differnt colors and shapes.

I think sometimes I just missed things because the voltage was to low.

one more tought. the resonant voltage rise of the bifilarpancake coil has no magnetic component.

I tried several ways to detect the magnetic component. but it simply isnt there.

this is the idea
1 and 2 are the relais coil connections.
5 and 6 is the 12v dc power supply (battery or super caps).
3 and 4 are the switch connections. to 5 and 6 or 7 and 8(energised)

the step up transformer is in series with the rrlais coil.

when the coils are energised (magnetic field) via 5 and 6 the switch turns to 7 and 8. disconnecting the power supply.
the coils are then connected to 7 and 8 the capacitor.
the bemf of the collapsing magnetic field is stored in the capacitors dielectric field. and is able to shortly oscillate between magnetic and dielectric field.

maybe the contacts will stick on 7 and 8, maybe not (the contact is made without charge in the capacitor).

the back  magneto motive force of the capacitor discharge, is the current that is stepped up in the transformer.
on the high side the magnetic field is able to create a b emf voltage again rectified into a capacitor and load (back to battery)

damn... sideways...

Your schematic as shown will not work as expected.  Your running the step up transformer through
the relay coil which will increase the resistance to the transformer significantly.  Also the voltage
generated from the back emf of the step up transformer will charge the capacitor once, then instead
of discharging it back into the coil it will just receive another impulse from the transformer and the
capacitor circuit will be for nothing unless you replaced it for a load.  In addition it may be wise
to use a bridge rectifier instead of the half wave. 

oops secondary should be: one side to ground otherside half rectified.

and maybe place the capacitor to the primary side.
no spark gap just let it ring and discharge. this might be a problem when opening the relais. better start without a cap maybe

the coil could also be in parallel
edit; not parallel the video says the coil needs to be inseries else it shorts it out.

my relais is much smaller than the video one. but it should work. only with a smaller load. a few leds in series making 12V probably will be good

Back Electro Motive Force from the collapsing magnetic field (inductor)
Back Magneto Motive Force from the collapsing dielectric field (capacitor)

A Bifilar coil has both fields. It even seems to have a third field.

We can split a bifilar coil into 2 distanced windings, and place a dielectric between the windings. the dielectric will still be between the 2 windings. But the two windings can now be influenced separately.
One with the BEMF(low voltage high current magnetic field) of a coil next to it, And the other one with the BMMF from a coil next to that one(fed with a high votage discharge of a capacitor).
As long as the BEMF and BMMF are simultaneous, the split bifilar coil should react, and altough I believe the resonant frequency should prove the best, it isnt necessary. As long as the bmmf and bemf are combined into one signal.

A relais is needed, a solid state relais would be best. but a simple 12V relais could make things more understandable.

Using my Universal Switch (which is actually just a high-speed SSR), the coil behaves quite a bit differently and back EMF climbs real quick.  Using this switch instead of a strong drive transistor allows the exciter coil to find its SRF, which by the way changes based on how closely the collector coil is placed next to the exciter coil.

So now with only a capacitor connected to the collector coil and no capacitor on the exciter coil, I can see both resonant frequencies of each coil.  What is also interesting is watching the exciter coil's SRF and Q-factor change as the two coils are brought closer together.  I did not see any of this before using this switch.  Essentially the difference is the switch only closes the circuit during the on-time of the set duty cycle; the coil is completely disconnected from any closed circuit the rest of the time.  The other difference is the switch allows you to switch either the positive or negative side of the power source, so it's possible for the coil to remain earth grounded.  I haven't tried this test yet to see if the exciter coil SRF changes, but will have a look shortly.

great work Matt. I'll look into that ssr.

I hooked up my IGBT pulse driver to a 12V relais (dpdt). I used 100Hz square pulses at 95% duty cycle.

relais switched a power supply 0.5V 0.04A
To a step up transformer.(95% charge at the primary, 5% off generating bemf)

highside was grounded and one side probed.

the spikes are off scale. around 750V

this looks like what nelson uses.

so with a solid state relais these HV bemf pulses  can be used to pulse a bifilar coil.

next is seeing what happens with a capacitor bmmf

small changes to the same setup.

removed the ground of the secondary. moved the probe to the primary side.

removed the power supply (!) replaced it with a 82pF mica capacitor.

still pulsing at 100Hz 95 duty.

so the capacitor is 95% connected to the step up transformer and 5% off.

the relais is also 95% magnetised and 5% off.
it seems the bemf of the relais is somehow feeding the capacitor.

the pulses on the primary are 90V.
the secondary also reads pulses of 90V(!) so the voltage isnt stepped up this time. only a bemf does this. a bmmf doesnt step up(!)

the capacitor and transformer are never connected to the power supply...

the only source of energy is the collapsing field of the relais inductor.

I bend the mica capacitor towards the relais. now Im getting 110V pulses.

left yellow black is the igbt pulser to relais coil

right black and red is to primary of the probed step up transformer.

the capacitor is bend towards the relais.

Evo, that relay looks like a DPDT type, but I can't see where the Common, NO and NC connections are.  Can you scratch out a simple schema for me to look at.   Having a hard time understanding how this is working for you and why on earth just moving the capacitor is changing the voltage readings you are getting.

here you go matt
its a omron g5v-2 12vdc relais. pdf is on the web some where.

photo of the relais shows the connections better

I think the collapsing magnetic field of the relais produces a dielectric field.
the capacitor captures the dielectric field and stores it.


the cap then is shorted out over the primary coil and measured there.

funny that the secondary also gives the exact same pulse signal as the primary. no stepup here. so the pulse is different.

the bmmf from the dielectric field collapse is differnt then the bemf from the magnetic field collapse, (it does step up the voltage)

when i remove the transformer the signal is still there on the relais pins.

but when I remove the capacitor, the signal is gone. the the capacitor is vital in capturing the energy.
this should prove interesting with a bifilar coil with its large capacitance.

I also probed both switched connections to see if they alternate but the signal is read in sinq. pulsed at the same time...

even if only one lead of the capacitor is connected, all pins show the pulse.

WHAT?

ok small correction.

with only the igbt to the relais all pins are showing the pulse (without any capacitor or inductor connected).

this makes sense.
the magnetic field collapses (relais coil) and the metal parts of the relais are induced by the bemf dielectric field. producing a short burst in voltage.

when the cap is connected it can store the energy.

I also retried the capacitor charging discharging setup.

connected the 12v dc power supply to the 95 % on
and the transformer to the 5% off.

the cap should charge up from the 12v dc
and discharge via the coil.
but no... it wont work, why???

I try the same with an identical relais.

I see an interesting concept with your setup--two switches in series.  If these two switches do not open and close at EXACTLY the same time, there exists a possibility of a very small spike, possibly faster than your scope can detect.  That might have something to do with the effects you are seeing.

Quote from Matt Watts on June 28th, 2017, 08:32 AM

I see an interesting concept with your setup--two switches in series.  If these two switches do not open and close at EXACTLY the same time, there exists a possibility of a very small spike, possibly faster than your scope can detect.  That might have something to do with the effects you are seeing.

hmm maybe. I dont know.
if nothing is connected to the switch, there still is the pulse from the relais coil bemf as the igbt turns it of.

Okay, I have a foundation to work with here--two switches controlled by my signal generator.  One charges the cap; the other fires it--basic CDI.  One side of the coil and cap remain grounded at all times.

For the two scopeshots, the first one is pretty self explanatory:  Yellow trace is charge, pink is fire.

In the second scopeshot, yellow is the ring-back on the exciter coil; pink is the signal on the collector coil.  Notice there is some power transfer going on here, the output LED is lit.  Exciter coil is on the bottom; collector coil is on top.

The capacitor I'm using is probably way too big, 100uF.  I'll try to tune it better, but what's interesting is the ringing on the exciter coil--it's almost as if it doesn't care about the capacitor and instead rings based on its own internal SRF.  That ringing changes drastically if I lift up the collector coil.  Back EMF shoots way up even though I'm only charging the capacitor with five volts.  The frequency and Q-factor of the ring also changes, so the collector coil is definitely communicating with the exciter coil--they don't work independent of each other.  The fields emitted by both coils are entangled.  Vortex rings mixing maybe...?