STEORN DEMO LIVE & STREAM in Dublin, December 15th, 10 AM

[Omnibus]

@teslaalset,

The graphs above show the principle possibility to design E-t curves of the input with a negative slope within a full period. That's based on the fact that you can apply voltages differing in offset while due to the peculiarities of the RC circuit the resulting current loses that offset. This allows for a favorable I-V juxtaposition leading to the observed curves with negative slope. Now, of course, when I say negative slope I have in mind the slope obtained by the least squares method. If we're finding the slope the way you suggested, using the beginning and the end point, the slopes in this case will not be negative. Now, the question is, can you lose per unit time more than you gain (which the least squares method gives) along the way and nevertheless end up with a gain at the end of the day (as the slope based on the two end points suggests)? This is where I'm coming from in suggesting this method of offset.

On the other hand, as was seen, if there's a favorable load (the capacitance and the resistance are just right) on the power supply, as is the load applied by the 1X probe, then there's OU even in absence of offset and that can be seen even if the slope is to be determined from the end points (and not using least squares method). Of course, here I assume that the voltage the 1X probe measures is the real voltage applied to the RC circuit (true current and true resistance being guaranteed in all cases). In absence of offset the 10X probe appears to be lacking the proper combination of R and C to load the pulse generator favorably, that is, to load it so that OU can be produced. That's how I understand it at present. I tried to model an additional load on the pulse generator, by trying various R and C's while measuring with the 10X probe but still haven't been able to design the effect seen with the 1X probe. The only way so far I could observe OU with the 10X probe (if that's really OU, of course) is by playing with the offset.

Seems to me, all in all this is all a matter of design, a matter of combining various elements (R, C, L) and probably playing with the offset, with the idea to shift the phase of the current slightly more with respect to the voltage than the shift that the capacitance in the circuit would cause naturally. This applies to the current and voltage too -- say, have voltage (for a given current) which not only will differ from the voltage derived from Ohm's law but also different from that expected for the given RC circuit. Success in terms of OU with electrical devices will depend on whether or not such design is possible and is actually found out. This is the obvious search for asymmetries (in opposition to the symmetric CoE) we've been after in every concrete rendition of an OU device we've been exploring so far. @Omega_0 also mentioned that in one of his last posts. That's the goal and we have to see how we can achieve it in every concrete case.

[Omnibus]

The problem I have, it seems, is that I'm considering the slope of E-t curve while I should be considering that of dE/dt-t. That's regarding the last studies. As far as the 1X probe results, however, all is right and it's confirmed also by the dE/dt-t slope there. So, actually, having already results showing OU (those with the 1X probe, not the last ones with the offset) the final picking at straws by the critics would be, again, is the voltage measured by the 1X probe (current and resistance being the true ones) the real voltage applied to the RC circuit? To answer this question I did some studies such as those to recover the known value of a resistor by using voltage measured across a shunt with 1X compared to 10X probe and dividing it by the known current. It was seen that the 1X probe recovers the resistance value better throughout the period. Wonder what other studies can there be to ascertain we're measuring the true voltage across the RC circuit with the 1X voltage probe?

[teslaalset]

I did some simulations on the measurement with the 5 V offset.
I used the 'trendline' option in my Excel.
As you can see, it's kind of tricky to use only one cycle and determine the slope.
The right one in below graph represents 15 full cycles, while the left one represents one full cycle

[Omnibus]

I did some simulations on the measurement with the 5 V offset.
I used the 'trendline' option in my Excel.
As you can see, it's kind of tricky to use only one cycle and determine the slope.
The right one in below graph represents 15 full cycles, while the left one represents one full cycle

I agree. The one cycle experiments were done to match the theoretical derivation (the derivation I already did and the one that is to be done with an offset). However, again, the problem is that I'm trying to judge for the energy spent per unit time from the E-t curve while, because of non-linearity, the correct way is to judge from the dE/dt-t curve. That was the way the derivation was carried out, as a matter of fact. So, thus far, the conclusion is that the 10X probe loads unfavorably the pulse generator while the 1X probe fortunately loads it in a manner to have the I displaced with respect to V and I and V to be of such magnitude as to yield an easily reproducible OU. Would be nice to model that loading in the case of the 10X probe, if that's at all possible.

[Omnibus]

Now, here's what it looks like for two periods. That's for 1X probe. From these two periods it may appear there's no OU. However, take a look at the results for many more periods I'm gonna post next.