PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-03-29, 06:33:48
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: 1 [2] 3 4 5
Author Topic: Self running coil?  (Read 82811 times)
Group: Guest
Install an opto coupler.

If it still rises, everybody will be all over this.

   

Group: Administrator
Hero Member
*****

Posts: 3866


Buy me some coffee
Or use an opto isolated FET driver
[pdf]http://www.datasheetcatalog.org/datasheets/400/499036_DS.pdf[/pdf]
   
Group: Tech Wizard
Hero Member
*****

Posts: 1092
....
Can anyone please explain how the 3vpp measured gate leak between the source and drain is capable of keeping the circuit running NOW at 30vdc, 55KHz with 50% duty cycle without using any current and maintaining 0.56vdc on the pickup coil with 49,850 Ohm load. Thank you for your help and time

Test 5 video: http://www.youtube.com/watch?v=RhrYzBld74w

Test 6 video: http://www.youtube.com/watch?v=UflGpzijWIA

Luc

Hi Luc,

I copied my answer here too from the other forums, so that others here also could read it.

The drain source channel of the MOSFET is practically a short circuit for the ON time (actually the RDS value of the IRF640, data sheet), the ON time now means half wave duration (earlier you used 18-20% or so duty cycle, this meant less ON time, more OFF time).

The drain source channel of the MOSFET is very high resistance for the OFF time, and both the output capacitance of the MOSFET and the body diode is in parallel with the drain source, the OFF time means the other half wave duration of the full pulse time. Now the output capacitance is about a few ten to a few hundred pF only, because the drain source voltage is much higher than earlier, up to 30V.
I still think the resonant tank LC circuit is formed by the output capacitance of the FET and from the 221mH (magnet-tuned) toroidal coil.

The FET as a switch pumps energy into the tank circuit from your generator and you have to consider not only the 3V peak to peak voltage but the current spikes shown in your earlier videos as flat lines between the spikes.

I think the input energy comes from these two: the 3Vpp  and the spikes.  The flat line between the spikes is explainable from the fact that the input is a square-wave: suddenly appears across the coil then its amplitude remains more or less constant, this means no or a very little flux change, then the square wave returns to zero, this also causes a flux change in the core again, current spike appears again.

So to estimate the real input power to the tank somehow those current spikes should be studied, I believe these maintain the voltage in the caps.
Resonant LC circuits have voltage 'amplification' properties, this depends on the loaded Q factor too.  In you circuit this is modified a little, the normal Q times multiplier is not fully valid, due to the half wave rectification inside the tank.

If I can, I will address some unanswered questions tomorrow.

Thanks,  Gyula
   
Group: Moderator
Hero Member
*****

Posts: 901
Hi everyone,

Here is the most up to date circuit.

http://i944.photobucket.com/albums/ad290/gotoluc/gotoluc-circuit2.jpg
Self running coil?
« Last Edit: 2010-03-19, 03:20:16 by gotoluc »
   
Group: Moderator
Hero Member
*****

Posts: 901
This was posted at the OU topic and thought it may help others here.

Just looking at it from a layman's point of view, if I had such a circuit, and it looked like it was OU, the first thing I would do is to try to close the loop. If it stopped working when I tried it, it would at least get me working in a direction that would let me look at the reason for not being able to close the loop.

If it did close the loop, then after I got over my astonishment, I could try to find out what non-OU reason would allow me to close the loop. And it would be a hell of an incentive for everyone else to replicate what I had.

Either way, trying to close the loop would be very useful.

That may be a naive layman's view of it, but it is the reason I cannot understand why Luc doesn't even try to do this.

Hi canam101,

I personally would be surprised that the circuit (posted above) on its own would be OU

However, if we are able to make a Toroid with a magnet Resonate at next to no cost of energy since even the possible gate switching energy is being returned to the capacitor bank, then this is good.

What you and others who are not replicating may not be considering is that a pickup coil can added and real current can be extracted at no cost to the capacitor bank returned input energy.

This is where the possible OU or Free energy could be coming from. My tests so far show that when I add a load to the pickup coil it has no reverse effect on the input. JLN has demonstrated this also but he does not have his input at zero yet!

Also, there could be a possibility of multiple pickup coils added as the toroid seems to have an opposite field on each half and maybe on each side. So there could be 4 pickup coils. The other thing is the pickup coil I used is not tuned so it is not taking full advantage of the potential magnetic current.

Luc
   
Group: Moderator
Hero Member
*****

Posts: 901
Hi everyone,

thanks for all your interest, suggestions and encouragement.

I made a new video in hope it will help replicators to better understand how to tweak the signal generator for the Toroid coil to send back the most current. I also measured the current going to the gate using my scope probe across a carbon 100 Ohm resistor. The below scope shots are the results.

Link to video: http://www.youtube.com/watch?v=-pkRBe0dpa8

Luc

This first scope shot is when the coil is at the neutral point (no current used)

http://i944.photobucket.com/albums/ad290/gotoluc/1Large.png
Self running coil?


This second shot is when it is tweaked to send most energy back. One can clearly see that less current is used when sending back the most current. This is difficult for me to understand ???
http://i944.photobucket.com/albums/ad290/gotoluc/2Large.png
Self running coil?

   
Group: Guest
Luc:

Thank you for posting a schematic of your circuit.

Let's talk about when you have voltage on the capacitor and you disconnect the battery and the circuit is running.

When the MOSFET is ON, the capacitor and the inductor form an LC resonator.  The resonant frequency is 1/(sqrt(LC)).  Therefore the value of the cap and the inductor are equally important in determining the frequency.  How this frequency relates to the ON time of the MOSFET is critical.  If the LC resonance frequency is very high, it may oscillate multiple cycles during the ON time.   Therefore when you switch OFF will determine where you are on the voltage and current waves.   Perhaps what I said is more academic because I looked back at your first two clips.  It looks like oscillation of the current you show on your scope traces is exclusively due to the switching transients that I discuss in more detail below.

The MOSFET is between the capacitor and the inductor and acts like a small resistance.  Therefore in truth it is more like a an RLC resonator.  I think when the MOSFET is ON it is friendly to current passing in both directions through it, but you would have to double check that.

The bottom like is for every setup you have done the energy source is your gate driving signal for the MOSFET.  If you have ever opened up a two-way speaker you would have noticed that they put a capacitor in series with the tweeter.  The capacitor blocks low frequencies and lets high frequencies pass, which is exactly what you want for a tweeter.  Thus the capacitor lets high frequency AC energy from the amplifier flow through itself into the load, the tweeter.

In your case there is built-in capacitance associated with the MOSFET semiconductor material lets high frequency energy flow from the signal source into the rest of the circuit.

The MOSFET also acts like a diode when it is shutting off.  It's called the "body diode" and it only exists for a short time.  This is almost certainly the mechanism responsible for keeping the capacitor charged.  You have an AC signal, which represents a flow of energy, that goes through a diode and becomes a "peak detector" which will charge the cap up to the peak voltage less one diode voltage drop.

When the MOSFET switches OFF, then the coil may or may not have current flowing through it at that instant.  I there is current flowing through it, that will generate a back-EMF spike.  Depending on the timing and the the body diode, the coil may be able to charge the cap a tiny bit.  It's all academic though because the real source of energy is your gate driving signal.

Using an opto-isolator will have no affect on the operation of the circuit.  The opto-isolator will drive the MOSFET gate anyways so the same thing will happen.  measuring the current consumption of the opto-isolator itself would give you an indication of how much power is being pumped into the circuit from the gate driving signal.

In your later clips you experiment with a pick-up coil next to the toroid.  By definition, the toroidal coil with the ferrite core is designed to produce a nearly all-internal magnetic flux path.  Hence by definition you can barely pick up a signal on the outside of a toroid.  Nonetheless, you picked up some some energy.  The question is how much energy?  You should do the calculation for your load resistor.  I will assume it is something like a few microwatts.  This is an insignificant amount of power.  It is such a small amount of power that it only appears to not be affecting your circuit, but in reality it is affecting your circuit.  However, your circuit is the size of an elephant and the pick-up coil part is like a mosquito sitting on the elephant.  You can't see the effects because they are too small to see with your measuring instruments, but they are there.

I suggest you try moving your probe around as your run the circuit.  Look at the voltage across the coil itself.  If your scope can be independently triggered from the signal source, then you can work with both channels.   You could look at the coil voltage and current at the same time, or look at the capacitor voltage and the current at the same time.  That is where the real action is.  Just seeing the circuit in one dimension, the current waveform through the shunt resistor, is not giving you very much information.  You want to look at the circuit in two or more dimensions.

There is no "resonance" going on here like you think.  You are looking at an interaction between pumping energy in from the signal source on every rising and falling edge of the signal source.  Every time the MOSFET switches ON, the LC resonator wants to resonate.  When the MOSFET switches OFF, you either "catch the beat" or you don't "catch the beat" of the circuit.  When you are "catching the beat" you are getting the highest capacitor voltages.

Anybody feel free to cross-post this.

MileHigh
   
Group: Moderator
Hero Member
*****

Posts: 901
Hi All,

a new video once again.

Link: http://www.youtube.com/watch?v=sbgwlJx0zNw

Luc
   
Group: Moderator
Hero Member
*****

Posts: 901
Luc:

There is no "resonance" going on here like you think.  You are looking at an interaction between pumping energy in from the signal source on every rising and falling edge of the signal source.  Every time the MOSFET switches ON, the LC resonator wants to resonate.  When the MOSFET switches OFF, you either "catch the beat" or you don't "catch the beat" of the circuit.  When you are "catching the beat" you are getting the highest capacitor voltages.

Anybody feel free to cross-post this.

MileHigh

Hi MileHigh,

thank you for taking the time to post your explanations. The only thing that I don't get is you say "there is no resonance" but then you talk about LC resonator and "catch the beat" 

This to me sounds like a form of resonance and is what I mean when when I say at resonance the effect starts. It's about triggering the switch at the correct time and collecting back the energy.

Thanks for sharing

Luc
   
Group: Guest
Luc:

As a follow-up to your above clip let me suggest an experiment for you.

You just demonstrated when you run your circuit with the cap initially discharged nothing really happens and the capacitor doesn't charge.  Your conclusion is that the gate signal cannot be putting energy into the circuit.

Try setting the initial cap voltage to one volt and run it again.  Then for two volts, three volts, etc.

Above a certain initial cap voltage you should see the cap voltage "take off" and charge to 16 volts.  Somebody earlier stated that the MOSFET has to have some voltage across it for the capacitive effects to emerge and that sounds about right to me.

It's another property of the MOSFET, the gate/source/drain capacitances only exist if there is voltage applied.  I am pretty sure this is the case but I am not 100% certain.  The only way to be sure would be to read up on MOSFETs and go look at the data sheet for the MOSFET you are using, or just about any other MOSFET data sheet.

Anyway, if you run that experiment with the different initial cap voltages before you switch on the signal source you will see one way or the other.

MileHigh
   
Group: Guest
Luc:

Great question about the resonance issue.  I will try to clarify it for you.

Yes there is an LC resonator associated with the circuit.  However, I was pointing that out more for clarification than anything else.  If the LC resonator frequency is much lower than the switching frequency then it doesn't really resonate.  It is still worth it to be aware that it is there.  The toroid has a large inductance and you have tried large and small caps.  I suggest that you calculate the main LC resonant frequency yourself and see how it relates to your gate frequency.

In your first clips you see a high frequency resonance at the switching ON and OFF of the MOSFET.  That is an LC resonance where the "L" is still the toroid and the "C" is the junction capacitance of the MOSFET itself.  In this case the "C" is very very small, hence the high frequency oscillation.

The "beat you are catching" is that beat.  It is the transient resonant frequency associated with the switching itself.  It is not any resonance associated with your "main" LC resonator, your capacitor and your toroidal coil.  By changing your ON switching time, you interact with that oscillation when you switch off.

Let me make a disclaimer where what I am saying my not be 100% dead on, but it's main principles that are correct.  You signal source is pumping power into the circuit and charging the capacitor.  The previous experiment I suggested to you should give you convincing evidence of this.  As you vary your frequency and look for the highest capacitor voltage, its almost like you are tuning in a radio station.  Yes, you can get a higher voltage, but the circuit is not necessarily "falling into resonance."  You are just lining up a transient voltage peak of the switching spurious oscillations with the timing of the existence of the MOSFET body diode.

Hey!  For what's it's worth, that basically exactly how the early crystal radios worked with the "cat's whisker."  The "cat's whisker" was a diode, and it was being used as a peak detector to charge a small capacitor.

MileHigh
   
Group: Guest
If the opto output is sourced from the storage caps, effective isolation WILL occour .

The storage cap voltage will have to be at 12 volts or so to be capable of switching correctly

The point is to entirely remove the probability that the drive is the source before getting too excited about the rest of it.
It doesnt matter why, just remove the variable.


   
Group: Guest
Lindsay,

That's a great idea.  If the opto-isolator driver power is taken from the cap itself then the cap voltage will start to go down.  Then just by measuring the timing of the cap voltage drop you can calculate the power consumption of the setup.  It might run for quite a while with very large cap.  Or to be more precise the time-constant of the drop in voltage of the circuit will be quite long, perhaps a few minutes with a 20,000 uF cap.

MileHigh
   
Group: Guest

Could you skip the signal generator and wind a few turns on the coil to act as a gate trigger and just let the system find its own favorite frequency? Find a spot where it could run as class E?

   

Group: Administrator
Hero Member
*****

Posts: 3198
It's not as complicated as it may seem...
A very similar thing happened (i.e. energy electrostatically-coupled/transferred) with the Ainslie circuit when Aaron was running it at a ridiculously-low power level.

You may want to try replacing the MOSFET with a few 3300pf capacitors and see if you have similar same results.

.99
« Last Edit: 2010-03-19, 01:52:56 by poynt99 »
   
Hero Member
*****

Posts: 520
But how did it go up to 30 volts?


---------------------------
   
Group: Moderator
Hero Member
*****

Posts: 901
Hi all,

here is a new video demonstrating a pickup coil and a LED as load.

Link: http://www.youtube.com/watch?v=bgPR9r14zWE

Luc
   
Group: Moderator
Hero Member
*****

Posts: 901
Luc:

Great question about the resonance issue.  I will try to clarify it for you.

Yes there is an LC resonator associated with the circuit.  However, I was pointing that out more for clarification than anything else.  If the LC resonator frequency is much lower than the switching frequency then it doesn't really resonate.  It is still worth it to be aware that it is there.  The toroid has a large inductance and you have tried large and small caps.  I suggest that you calculate the main LC resonant frequency yourself and see how it relates to your gate frequency.

In your first clips you see a high frequency resonance at the switching ON and OFF of the MOSFET.  That is an LC resonance where the "L" is still the toroid and the "C" is the junction capacitance of the MOSFET itself.  In this case the "C" is very very small, hence the high frequency oscillation.

The "beat you are catching" is that beat.  It is the transient resonant frequency associated with the switching itself.  It is not any resonance associated with your "main" LC resonator, your capacitor and your toroidal coil.  By changing your ON switching time, you interact with that oscillation when you switch off.

Let me make a disclaimer where what I am saying my not be 100% dead on, but it's main principles that are correct.  You signal source is pumping power into the circuit and charging the capacitor.  The previous experiment I suggested to you should give you convincing evidence of this.  As you vary your frequency and look for the highest capacitor voltage, its almost like you are tuning in a radio station.  Yes, you can get a higher voltage, but the circuit is not necessarily "falling into resonance."  You are just lining up a transient voltage peak of the switching spurious oscillations with the timing of the existence of the MOSFET body diode.

Hey!  For what's it's worth, that basically exactly how the early crystal radios worked with the "cat's whisker."  The "cat's whisker" was a diode, and it was being used as a peak detector to charge a small capacitor.

MileHigh

Hi MileHigh,

thanks again for your detailed explanation.

I do agree with most of what you are sharing. However, I am hopeful that we can extract more energy from the pickup coil then what is leaked from the gate.

I have just posted above this post a new video attempting to demonstrate that. Please have a look.

Also, below is my reply to a user at the OU forum of how I see this coil setup.

Luc

Just looking at it from a layman's point of view, if I had such a circuit, and it looked like it was OU, the first thing I would do is to try to close the loop. If it stopped working when I tried it, it would at least get me working in a direction that would let me look at the reason for not being able to close the loop.

If it did close the loop, then after I got over my astonishment, I could try to find out what non-OU reason would allow me to close the loop. And it would be a hell of an incentive for everyone else to replicate what I had.

Either way, trying to close the loop would be very useful.

That may be a naive layman's view of it, but it is the reason I cannot understand why Luc doesn't even try to do this.

Hi canam101,

I personally would be surprised that the circuit (posted above) on its own would be OU

However, if we are able to make a Toroid with a magnet Resonate at next to no cost of energy since even the possible gate switching energy is being returned to the capacitor bank, then this is good.

What you and others who are not replicating may not be considering is that a pickup coil can be added and real current can be extracted at no cost to the capacitor bank returned input energy.

This is where the possible OU or Free energy could be coming from. My tests so far show that when I add a load to the pickup coil it has no reverse effect on the input. JLN has demonstrated this also but he does not have his input at zero yet!

Also, there could be a possibility of multiple pickup coils added as the toroid seems to have an opposite field on each half and maybe on each side. So there could be 4 pickup coils. The other thing is the pickup coil I used is not tuned so it is not taking full advantage of the potential magnetic current.

Luc
   
Group: Moderator
Hero Member
*****

Posts: 901
If the opto output is sourced from the storage caps, effective isolation WILL occour .

The storage cap voltage will have to be at 12 volts or so to be capable of switching correctly

The point is to entirely remove the probability that the drive is the source before getting too excited about the rest of it.
It doesnt matter why, just remove the variable.


Hi Lindsay,

I also agree with your idea to be a good one. This would prove without a dough that the sustaining energy was solely coming from the gate.

However, please read my previous post on the previous page to MileHigh so you also better understand where I believe the real potential of this circuit is.

Now I need to find an opto isolator that can work up to 100KHz range as I would like to try this just for fun ;D

Thanks for sharing your idea Lindsay

Luc
   
Group: Moderator
Hero Member
*****

Posts: 901
Could you skip the signal generator and wind a few turns on the coil to act as a gate trigger and just let the system find its own favorite frequency? Find a spot where it could run as class E?



Hi darkspeed,

I agree with you and I have been thinking the same thing but if you look at my test 7 video you will see that the correct frequency is verrrry sensitive to tune. However, I'll play around with that as I have time. You should also read my last post on the previous page to Milehigh to better understand where I believe the real potential of this circuit is.

Thanks for your post.

Luc
   
Group: Guest
Luc,

sorry I was not more clear earlier about the opto.

Most of them are ok to 100k

you would start with the opto source connected the the drive battery , then once you have a similar result and 12 volts at the storage caps ..switch the opto source over to the storage supply .

If that works no matter how little  power it would be enough to really get excited about.
Of course tuning the extra pickups could then be used for the dive and who knows what else?

It is  starting to look like the SM centre torroid .

My fingers are well crossed for you !

Lindsay
   

Group: Administrator
Hero Member
*****

Posts: 3198
It's not as complicated as it may seem...
But how did it go up to 30 volts?

Wattsup,

If the coil is being driven by the generator (and I suspect it is), then it would be through ES coupling via the parasitic capacitance around the MOSFET. As such, the coil will still be "switched" albeit with some differentiation, and the output power will be small. However, the voltage is kicked up as per any flyback converter.

I'll try and dig up the post I did ages ago in response to Aaron's experiments where I illustrated how the same results would be obtained with 3 capacitors and a diode in place of the MOSFET.

.99
   
Hero Member
*****

Posts: 520
@poynt99

Well with my laymen EE brain as usual I do not catch the obvious. But, you see the cap is showing 30 volts. For me this means the capacitor was receiving at least 30 volts into it until its mF rating was reached, and any more 30 volts added would not make any more difference (wasted). He could not have reached 30 volts with only 2-3 volts sent to that cap. Something had to act like a step up transformer. I don't really know but I find it very curious and would consider the toroid is doing it and not the mosfet gate.


---------------------------
   
Group: Moderator
Hero Member
*****

Posts: 901
Luc,

sorry I was not more clear earlier about the opto.

Most of them are ok to 100k

you would start with the opto source connected the the drive battery , then once you have a similar result and 12 volts at the storage caps ..switch the opto source over to the storage supply .

If that works no matter how little  power it would be enough to really get excited about.
Of course tuning the extra pickups could then be used for the dive and who knows what else?

It is  starting to look like the SM centre torroid .

My fingers are well crossed for you !

Lindsay

Hi Lindsay,

I just tried 3 models of OPTO's and what is left of the pulse signal at the mofet gate is very ugly :-\   @20KHz it has next to no pulse width and much worse @30KHz. There is no way a signal like this will switch the mosfet correctly. The other problem is, I need a 10K resistor across the gate and source to turn the mosfet off after the pulse and that alone consumes 600 micro amps. at 20KHz let alone what the gate will use.

I picked up a CMOS 555 and from what I can measure it consumes 300 micro amps, so I'll play around with that and see what it can do.
   
Group: Guest
Luc,

The optos are used  in supplies running at 100 k , dont know if they are in resonance like you have though.
Perhpas a special device is know to one of us here

low current to the oscillator  is another way to limit the source current .

Another way just might be using an igbt ..i dont know if " isolated gate " really means what it implies   .99??

What about getting it running and using  pickup coil output to replace the gate pulse  as Darkspeed suggested . It would be very sensitive to position as well as being amusing to do.

   
Pages: 1 [2] 3 4 5
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-03-29, 06:33:48