Self running coil?

[Hoppy]

Its interesting that if a NE555 timer is used to supply pulses to a mosfet which is switching a coil connected between its drain and supply rail, whilst the 555 output is high, current from the supply rail will be fed through the mosfet drain to gate and through the 555 pin 3 output into its supply rail. If a cap is placed across the 555 supply rails, the current fed back will charge the cap sufficiently to power the 555 timer, so that the normal supply to the 555 can be disconnected. I first noticed this when I was pulsing a DC contactor from a 30V supply. The 555 was supplied from the 30V power rail with a 15V zener stabilised supply, smoothed with a 100uF cap.

Hoppy

@Gotoluc,
Hi,
I've not posted on this forum for over a year I reckon, but your experiment looks very promising.
Have you checked to make sure that no power can be added to the output through the gate of the mosfet?
There is a lot on this thread to read through.
Also I am still not sure if your coils are wound so the 2 coils appose each other, just like in a common mode choke.
In common mode chokes you can have split winding and parallel winding, yours is split.
The fact that your inductance is so high (more than the sum of the two) would indicate these coils are not wire in common mode config but are wired in series.
I have lots of cores to test out this idea with and a dual pulse circuit to allow the pulse current to be made square. (Thanks to Paul Lowrance and his mini orbo pulse circuit).
http://globalfreeenergy.info/2010/03/

Rob

[gyulasun]

....
Also I am still not sure if your coils are wound so the 2 coils appose each other, just like in a common mode choke.
In common mode chokes you can have split winding and parallel winding, yours is split.
The fact that your inductance is so high (more than the sum of the two) would indicate these coils are not wire in common mode config but are wired in series.
I have lots of cores to test out this idea with and a dual pulse circuit to allow the pulse current to be made square. (Thanks to Paul Lowrance and his mini orbo pulse circuit).
http://globalfreeenergy.info/2010/03/

Rob

Hi Rob,

Re on common mode choke: you are correct the winding technique Luc used for his toroidal coils is really the one as the so called common mode chokes are made BUT the big difference is the way how they are connected: Luc connected the two coils in series aiding i.e. the MUTUAL inductances of the two coils add to the sum of the individual inductances in series,  so that the resultant inductance is nearly the 4 times of a single coil, L_resultant=(L₁+L₂+2M where L1=L2=L in the equation (the two coils are assumed to have the same inductance which is nearly true and M is nearly L because in ring cores the coefficience of coupling is nearly 1 ).  (A useful link on this is here: http://www.daycounter.com/LabBook/Mutual-Inductance.phtml )

I hope this helps clarify your doubt above.

rgds,  Gyula

[mscoffman]

...
But Luc has been putting the 100 Ohm resistor in series into the gate circuit you must have missed that.   He showed actual scope shots already taken directly across the series 100 Ohm, see here the waveforms (click on the quote's link in bold just in next line):
...

rgds, Gyula

OK yes, I see. Thanks for bringing that up. That makes the 2Sxx mosfet
look much better with only 1.4Vpp across the 100ohm resister...
You know what, if gotoluc looks on his signal generator he will see
the output impedance specified probably like 60ohms. If he put a
60ohm resistor where where the signal generator leads terminate
to ground right before the circuit he may find that it squares up
the input signal trace significantly. You can see that the transistor
is beginning to convert voltage variations to current meaning that
it is seeing the input signal variations at the top of the square wave.
I think a signal generator termination resistor might square that signal up.
The signal generator will be working slightly harder but the level control
will bring the signal voltage back up to where it's needed

I wonder what the total circuit behavior *is* with this 2Sxx transistor?
The darned thing still has a high Ciss of nearly 1000pf or more so I don't
want to spend to much time on it if we could get one with a lower Ciss.

IRF640 drive max at 360mW peak
2Sxx drive max at 42mW peak

So we are driving it with about 1/9 of the power, via drive voltage variation
alone. Man I wonder where that lower figure would end up with smaller
Ciss? We are getting impedance matching probs between the signal
generator and circuit. So I wonder how the circuit is behaving with this
transistor? Gain in the bulk caps or not?

:S:Mark Coffman

[gyulasun]

Its interesting that if a NE555 timer is used to supply pulses to a mosfet which is switching a coil connected between its drain and supply rail, whilst the 555 output is high, current from the supply rail will be fed through the mosfet drain to gate and through the 555 pin 3 output into its supply rail. If a cap is placed across the 555 supply rails, the current fed back will charge the cap sufficiently to power the 555 timer, so that the normal supply to the 555 can be disconnected. I first noticed this when I was pulsing a DC contactor from a 30V supply. The 555 was supplied from the 30V power rail with a 15V zener stabilised supply, smoothed with a 100uF cap.

Hoppy

Hi Hoppy,

This is an interesting observation and you nicely described what you found,  however, some further notices may be in order.

The energy coming from the drain side back to the gate is caused by the drain gate interelectrode capacitance and depending on the type of the power MOSFET, this capacitance can be quite high to be able to cause the situation you described.  IF you choose another MOSFET type that has a few pFs only between the drain gate, then this situation cannot really happen as severely as you found, of course this also depends on the working frequency too. In Luc's circuit the frequency has been only as high as 50-60kHz or so and the IRF640 drain source cap value, C_rss is about 53pF at 25V drain voltage, this gives about X_L=40-50 kOHm capacitive reactance so high that significant current cannot really flow through it. Especially when there are no voltage spikes at the drain side because of tuning for resonance there, I think of the lack of high voltage difference between the drain gate -- this is the second reason why in Luc's circuit this is not experienced. (of course if he did experience it, it would be an unwanted phenomena because it would steal energy from the output resonant circuit.)

So the energy coming back via the C_rss also depends much on the instantenous drain voltage wrt the gate, and also, the higher this cap value, the bigger the energy coming back, at a given frequency.

You surely can recall other members finding that the body, (the casing) of the 555 integrated circuit sometimes gets heated up or suddenly simply go wrong after working for a few minutes, when directly drives the gate of MOSFETs that switches a coil via the drain, especially if the coil at the output has a resonant frequency that is far from the input drive frequency, hence big, nasty voltage spikes can couple back to the gate to find the output circuits inside the 555, beyond its Pin 3. 
While the amount of the unwanted energy coupled back can really supply a 555 as it were a separate power supply, I think it is an unwanted situation in most of the applications and you would have to make sure to limit the amount of the current coupled back, afterall it steals from the output energy, and make sure the 555 or the 15V Zener for instance will not overdissipate.  Also, when you wish to switch ON such a driving circuit that gets its supply from the drainside, it is not at all a 100% sure it can start again after a switch OFF because it would have no or very little supply voltage due to the initial lack of the big spikes at the drain side (a static 30V at the drain cannot couple back much current via the drain gate capacitance to start up the 555).

Sorry for the long 'rambling' I felt it interesting to discuss.

rgds,  Gyula

[mscoffman]

Hi Hoppy,

This is an interesting observation and you nicely described what you found,  however, some further notices may be in order.

The energy coming from the drain side back to the gate is caused by the drain gate interelectrode capacitance and depending on the type of the power MOSFET, this capacitance can be quite high to be able to cause the situation you described.  IF you choose another MOSFET type that has a few pFs only between the drain gate, then this situation cannot really happen as severely as you found, of course this also depends on the working frequency too. In Luc's circuit the frequency has been only as high as 50-60kHz or so and the IRF640 drain source cap value, C_rss is about 53pF at 25V drain voltage, this gives about X_L=40-50 kOHm capacitive reactance so high that significant current cannot really flow through it. Especially when there are no voltage spikes at the drain side because of tuning for resonance there, I think of the lack of high voltage difference between the drain gate -- this is the second reason why in Luc's circuit this is not experienced. (of course if he did experience it, it would be an unwanted phenomena because it would steal energy from the output resonant circuit.)

So the energy coming back via the C_rss also depends much on the instantenous drain voltage wrt the gate, and also, the higher this cap value, the bigger the energy coming back, at a given frequency.

You surely can recall other members finding that the body, (the casing) of the 555 integrated circuit sometimes gets heated up or suddenly simply go wrong after working for a few minutes, when directly drives the gate of MOSFETs that switches a coil via the drain, especially if the coil at the output has a resonant frequency that is far from the input drive frequency, hence big, nasty voltage spikes can couple back to the gate to find the output circuits inside the 555, beyond its Pin 3. 
While the amount of the unwanted energy coupled back can really supply a 555 as it were a separate power supply, I think it is an unwanted situation in most of the applications and you would have to make sure to limit the amount of the current coupled back, afterall it steals from the output energy, and make sure the 555 or the 15V Zener for instance will not overdissipate.  Also, when you wish to switch ON such a driving circuit that gets its supply from the drainside, it is not at all a 100% sure it can start again after a switch OFF because it would have no or very little supply voltage due to the initial lack of the big spikes at the drain side (a static 30V at the drain cannot couple back much current via the drain gate capacitance to start up the 555).

Sorry for the long 'rambling' I felt it interesting to discuss.

rgds,  Gyula

Right, the Crss spec value is the reverse source to gate capacitance.
In the IR640 it is typically 60pf. Makes no sense for a cap, but it
does for a semiconductor device. The positive spike of the Ids
caused by the coil can come through that so you essentially have a
output powered NE555. I think I would put a 1n914 diode from the gate
drive line to pin3 to guarantee what was going on. We estimated that
cmos NE555 took only 300uW to run and the zener diode doesn't turn on
until it's back biased to 15V protecting the NE555 from too much kick.
Would it work in our case as yours was driving into 30Vdc? It wouldn't
hurt to try and maybe would drain off some gate charge. A normal
NE555 I'm pretty sure it wouldn't. We would have to guarantee full
drive voltage and current.

Bad news; in the STS2DNF30L the Crss parameter it is only 11pf!

There is something called reverse amplifier gain or reverse
transconductance. And this Crss provides the basis for
that. This can also lead to self oscillations in linear AC amps.

:S:MarkCoffman