This
would determine if the 'effect' is the result of vagaries associated
with the wiring.
No !
You definitely need wires
to connect your parts.
If you want to upscale your device for higher
currents and load - you will need more diameter for the wire.
20Ah
batteries have an internal resistance of 15mOhms. Thats almost in the
range of the wiring.
For an upscaled device - you probably want to
choose 100Ah or 400Ah batteries.
Whats the resistance of the wires in
the demo device ? Do you know ?
Maybe there is a minimum resistance
needed for save operation ? Do you know ?
Wires are inductors - what
inductivity have the wires in the demo setup ? Do you know ?
If
you´re engineering something - you have to deal with all that stuff.
Whats
wrong with it ?
This
would determine if the 'effect' of a recharge may not be consistent
with actual recharge.
No !
As already
mentioned different batteries have different internal resistance,
depending on temperature, charge condition and pulse durations involved.
Actual
batteries are not rated for that purpose - means you have no data nor
guarantee that values will be different - for the same lot - or the same
battery model. So you have to collect the data on your own. Even if the
effect is proven - it is of no value if it works just with a single
battery.
So if you want to build lots of things like that - you have
to get a grip on your key components. The battery is a key component.
You
may add overvoltage and deep discharge circuits and so on...
This
would determine if the switch is responsible for what we claim is 'self
resonance' or 'preferred oscillation.
What I have
seen on your scope traces is a sporadic 50MHz glitch introducing a
"short" cycle.
All "normal" cycles don´t have this glitch. This
glitch comes initially from the NE555 power supply rail.
But how does
it come there ? inductive coupling ? On experimenting - you can find
out.
This
would determine if the resonance can be 'imposed' on the circuit
This
would be prefered, because you cannot solder 10 NE555 on top of each
other to increase the drive level to drive more mosfets with increased
gate charge.
So you probably want to replace that NE555 with
something that automatically adjusts and seeks the right properties for
that oscillation.
Not
sure what a DS-cap is. I take it that R1 is the load
resistor. The control signal DOES NOT jitter. I take it -
nonetheless - that would determine whether the load inductance was
responsible for the oscillation.
The gate of a
mosfet has almost infinite resistance against source and drain.
But
the gate forms mutual capacitors with source and drain.
This
capacitor ranges from 100pF up to 100dreds of nF depending on the used
part, how much in parallel and so on.
This capacity is somewhat
determined in the datasheet - but has significant tolerances.
In a
"professional" design - you want to get rid of those uncertain
conditions.
In your demo circuit - the gate capacity plays an
important role - because it forms an RC low-pass with R1.
This is why
you would have to match R1 every time you change the mosfet.
In
first order - the gate capacity against source (in combination with R1)
limits the amount of time needed to charge up the gate and to discharge
it - which finally controls the figure of the output resistance varying
with time on switching on and off.
Otherwise we have that
mutual drain-gate capacity (DS was a typo). If you discharge the gate
capacity (turning off) - the back-emf of the inductive load will lift
off together with the drain voltage. Because of the DG capacity - we
have a flow of charge from drain to gate on switching off.
In a
normal circuit you overcome that by having a low resistor from gate to
ground - and an extra protection diode to protect the gate.
You can
break the mosfet by having a higher gs voltage than rated - typical 15
volts.
If you would switch an inductive load with a mosfet - and
would disconnect the gate immediatley on turning off - the back emf on
the drain will lift the gate via drain-gate capacity - and the mosfet
would be dead.;-(((
But this means that the DG capacity can
operate as a feedback path.
Not
sure what you're recommending here. Presumably whether or not it
could be determined if the switching circuit alone could generate the
'effect'.
This would not work for the reasons that I've
explained. We've 'scaled it' as far as it can go with the
MOSFET. We've tried MOSFETS in series. It's too brittle.
a
predicted change in '5 dimensions'. 'Go back to 3'. 'the only
outcome that it doesnt work'. What part of this is experimentally
relevant and how much of this is determined as required precisely
because of that predicted outcome 'it doesn't work'?
The
only proof would be to try it out. (using tunable external
pseudo-random oscillator).
Getting rid of that 555 and the mosfet and
the tuning of R1 is essential for scaling up.
If the mosfet
(dg-capacity)+ R1 + coupling spike back into 555 is the feedback path -
then you will run into problems changing that configuration.
(because
that path is broken then)
If the outcome of further investigation is
that you have to insert a short cycle if there is a special signature
in the load current (already mentioned 50MHz glitch) - then we can
design a circuit for that triggering as huge mosfets banks as needed.
But
right now the chicken and egg thing isn´t clear.
If this glitched is
caused by an intermittant NE555 output stage overload effect - well a
pseudo random oscillator would do the same job.
What
you have listed here Fritz are the very questions that were addressed
by our accreditors. The experimental evidence was required in
terms of the thesis. The experimental results speak to the
thesis. There is NO other interpretation. Else we would not
have got that accreditation. This is precisely why I do not want
to waste more time on this thread with more experiments related to proof
of concept.
And
exactly what is it that I still do not understand?
That
there is a different point of view.
You found something, invented
something, there is a proof of concept.
But transforming that to
an easy replicable and scalable "technology" is a job on its own.
If I
use my oscilloscope to find an intermittant glitch crashing my
controller - this doesn´t mean that I dont´t trust proof of concept.
What is as clear as daylight is that you doubt the results related to
proof of concept. I could spend another year researching this to
your satisfaction and still you would have doubts.
Is
it necessary to wipe away my doubts if I just want to help you with the
driver stage ?
A hands-on experience would wipe away doubts anyway
if they really exist.
It's the nature of the claim that causes this. Those
results. They do NOT make sense in the context of known physics.
OK
there is something, a battery, a load, a switch, excess energy. A
miracle happens.
But BTW: I don´t think that there is a bubble in
spacetime which surrouds your circuit causing everything to work
completly different.
In that case I cant help you anyway
because I dont know how electronic components in a spacetime bubble
operate.
Regards,
Fritz