Claimed OU circuit of Rosemary Ainslie

[Hoppy]

I have tested Aaron's latest mod with the ten turn pots. This mod was quite stable but again I cannot get down to a 3.7% duty cycle at the mosfet gate and the adjustment is very limited. I am still getting good correlation between the temperature of the 'control' and circuit at equivalent power into the control resistive inductor, so I cannot see how the high COP is arrived at. 

I also tested the theory that the battery is being charged with heavy current spikes as claimed by Aaron. To do this, I connected a high voltage diode in series with the battery to block any spikes that may be coming back into the battery. I then carefully monitored the discharge rate of the battery with the diode in circuit and then switched a short circuit across the diode and again monitored the rate of discharge. I could not see any evidence of charging with or without the flyback diode connected across the 8.9uH, 10R resistive inductor.

The amplitude of the spike could be adjusted up and down with the gate pot but at no point did I see any of Aaron's 'oscillations' when scoped across the battery. They may have been there but my waveform resolution is suffering because of a drifting focus control. However, if they are present, IMO they are certainly not producing any noticeable charge into the battery.


Hoppy

[poynt99]

Hoppy,

Aaron is going to refer you to his video showing HV spikes going backwards through a string of diodes.

He would be correct in stating that this will happen and this is the reason you are seeing no difference.

I've stated it a few times and I'll re-emphasize again, when dealing with high frequency and high rise/fall time circuits we can't think of these devices and how they operate in terms of DC behaviour. Consideration has to be given to transient conditions and under such, all the reactive components within and around these devices need to be taken into account. For example, there is a junction capacitance across all diodes, and although small in a reversed-bias condition (but is non-linear and slow), it is there and represents a path for spikes to conduct.

Similar consideration must be given to the MOSFET.

.99

[spinner]

Oh, dear...
This thread is still alive?
Can't believe it....

"OU FET heaters"   



It's well over 100 pages now...

What exactly is the problem? That the circuit isn't OU? Anything else?

Jeeez....

[forest]

Hoppy,

Aaron is going to refer you to his video showing HV spikes going backwards through a string of diodes.

He would be correct in stating that this will happen and this is the reason you are seeing no difference.

I've stated it a few times and I'll re-emphasize again, when dealing with high frequency and high rise/fall time circuits we can't think of these devices and how they operate in terms of DC behaviour. Consideration has to be given to transient conditions and under such, all the reactive components within and around these devices need to be taken into account. For example, there is a junction capacitance across all diodes, and although small in a reversed-bias condition (but is non-linear and slow), it is there and represents a path for spikes to conduct.

Similar consideration must be given to the MOSFET.

.99

EXACTLY. Do you know maybe the way to damage internal MOSFET diode but let FET still working ?

[MileHigh]

Hi Hoppy,

I also tested the theory that the battery is being charged with heavy current spikes as claimed by Aaron. To do this, I connected a high voltage diode in series with the battery to block any spikes that may be coming back into the battery. I then carefully monitored the discharge rate of the battery with the diode in circuit and then switched a short circuit across the diode and again monitored the rate of discharge. I could not see any evidence of charging with or without the flyback diode connected across the 8.9uH, 10R resistive inductor.

Are you indicating that you were working with two diodes, a series diode at the battery positive, and the more familiar parallel diode across the coil-resistor.  You tried with the series diode or bypassing the series diode with no noticeable difference across the shunt resistor?  Then as a separate test you tried with/without the parallel diode across the coil-resistor and got similar results?

I have a new tentative theory about the high voltage spikes that Aaron is seeing at the battery positive terminal:  It's back to basics again because you can't forget that the MOSFET is not conducting so current can't flow in the circuit.  That's the only way to get the high voltage pulse in the first place.

The fact that the inductive energy from the discharging coil-resistor has "nowhere to go" because the MOSFET is switched off results in the energy going to three places:  1) Some of the energy is capacitively coupled through the MOSFET junction capacitances of the gate and source pins and goes into the 555 circuit and the shunt resistor because of dv/dt to ground.  2) Some of the energy is dissipated through the MOSFET "off" resistance and goes through the shunt resistor to ground.  3) Some of the energy is reflected back to the battery positive terminal and travels through the battery to ground.

For the battery, the real issue is how do you model it for high-speed transients.  Like Harvey stated earlier, there may be a type of battery "inductance effect" for high-speed transients.

I am going to suggest something similar to that based on input impedance.  The battery has a "dynamic impedance" model for when high-speed voltage transients enter at the positive terminal.  For something on the order of a few microseconds, when the battery is first hit with a voltage transient, it acts like a very high input resistance and a very small input capacitance.  We know what happens with the input capacitance, you get dv/dt current to ground.

I want to focus on the dynamic input resistance.  This only lasts for a few microseconds before you drop down to the input charging resistance that might only be 0.5 ohms or less.  So the 500 nSec high-voltage spike hits the battery and is "dissipated" in the battery as a v-squared/R dissipation, where R is a very large value for just a few microseconds.

It is highly doubtful that the battery is actually charging under this voltage transient condition, because you normally associate charging with a very low battery input resistance.

In theory, this can be confirmed by measuring the current going into the battery when the 500 nSec high-voltage transient hits it.  If what I am saying is true, you should measure relatively little input current.  You can actually measure the battery input resistance under the voltage transient condition since you have the transient voltage and the transient current.

Exactly where this transient spike energy goes in the battery is perhaps for somebody else to answer, I am not a battery expert.  I have a feeling it is perhaps more than 90% energy dissipation as heat and less than 10% battery recharging.

The net result of this theory, and I stress that it is a theory only, is that the big voltage transients going into the battery that are about 500 nSec wide (to be confirmed with measurements) do not really charge the battery.  The individual spikes have very little energy in them, and almost all of it becomes resistive heat.

The net result is that the 500 nSec voltage transients at the battery positive terminal are simply too fast and too short in duration to charge the battery.

You can contrast this with a typical Bedini motor configuration.  In this case the spikes going into the charging battery are much much longer in duration, so the spikes are "slower."  The charging battery has more than enough time to react to these slow spikes and really gets charged.  The battery reduces the charging spikes to low-voltage current spikes, that are just a few volts max above the battery voltage.  That's in distinct contrast to what Aaron is seeing in his recent clips, where the spikes are "too fast" for the battery to react and the spikes don't get muffled down to a low voltage at all.

Some food for thought for everybody!

MileHigh