Re: Finally Heisenberg's is dead!

It's tiresome, and hard to get designs done, when everything gets revolutionized every few days.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Jan Panteltje's post did claim that the work was revolutionary, but he's dim enough not to have realised that it was Schroedinger's cat that might or might not have been dead, not Heisenberg's.

In fact the work is a spin-off of some work on quantum computing, and since you don't go in for that, you can indulgence in what you fondly imagine to be design without any anxiety.

As you can not-design without anxiety.

Imagine if you can, a series R+L+voltage-source circuit where all three elements can be independently arbitrarily modulated at MHz bandwidth. Well, conservation of energy isn't everything.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

I do design stuff from time to time, and most of it gets posted here. It's not serious, in the sense that it usually isn't properly toleranced or documented, but in the sense of arranging components to get a desired result, it is serious enough.

It's kind of hard to modulate resistance at all - at least in a resistor.

Inductance is similarly inflexible. If you add a second inductor coupled to the first, you can do all sorts of stuff, but the only way you can modulate the inductance at MHz rates is by saturating a magnetic core in the return path, and that's above your pay grade.

Voltage sources can be modulated without difficulty.

I don't need much imagination at all to see you typing out word salad that you don't actually understand.

That's not the only way. Injecting a variable current at 90-degrees to the main current will simulate a variable inductor or capacitor. I have used this system to 'pull' a 16.5 Mc/s crystal oscilator using an EF91 as the gain-controlled phase-shift amplifier.

Clue: The 90-degree phase shift is being injected into the cathode of the first EF91 (modulator) from the cathode of the second EF91 (oscillator) with the choke between the cathodes and the 680-ohm cathode resistor forming the phase-shift network.

Of course it's hard. That's why they sell well.

Oh it works fine. We'll be shipping the R+L version soon. Adding V is obviously easy.

My dilemma is what to call it. "Memristor" was invented to describe a resistor whose value was a function of its current history. Memristors have been revolutionizing memory technology regularly for decades now.

A series RLV with all three values arbitrarily modulated by anything needs a suitably grand name. I was thinking Frankimpedance.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Omnipedance?

Gyropedance is sort of a play on gyrator. That sounds a little scientific.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

I've used that approach (in LTSpice only) to make a Wien bridge where both amplitude and frequency could be controlled in parallel by in-phase and quadrature feedback paths. It used two good (and - as it turned out

- hideously expensive), AD four-quadrant-multiplier chips.

I've still got the chips.

For small deviations is frequency and amplitude it generated a pretty low distortion sine wave, at least in simulation.

But you aren't modulating the inductance or the coupling.

Fantasy electronics?

It might, to people with little grasp of science.

In the case of "Gyrator", Tellegen named it to describe how it rotates the behaviour of the conventional elements of L, C & R. As far as I can see your product doesn't involve any form of rotation, so the "Gyr" bit is misleading.

Well, we're synthesizing arbitrary impedances, arguably a bit more than what a classic gyrator does. But advertising slogans don't need to pass peer review in academic journals.

One problem with faking inductances by gyrating caps is energy storage density.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Fleximpedance?

That's not bad. Could be a newsletter title.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

But they do need to be tailored to the target audience.

Falls a bit short of suggesting arbitrary impedances (which might need to be negative).

We call them "customers"

Sure, once you have a general impedance synthesis system you can just program things negative. That has stability issues, but we program both resistances and inductances close to zero, below the actual values inherent in the circuits.

We can stably simulate negative values, provided the customer is presenting an external impedance that makes the parallel combination net positive.

Of course, if he presents a resonant load and we make a negative impedance, it oscillates.

As an EE student, I did a project of making a 2-terminal negative resistor and plugged negative values into the usual voltage divider and resonant circuit equations, and demonstrated that they worked in real life.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Ok. It can be called the Larkinator - a programmable modpedance generator :-)

Ed