Josef, can one get more details ? such as what is CCII+ in the figure? any description or statements regarding this circuit. In theory oscillators are made with wien bridge with equal R's. If series  R tends to infinity, the gain required for oscillations will also increase tremendously. I tried to do some simulation but could not get anywhere. I had assumed that the box shown is a multiplier, but that could be wrong. CC sounds it has something to do with current! If it is a current operated device, the output side R has a significant role and could be the missing R. And how is Iz related to the other currents, Ia and Ix? interesting circuit, though.

Yes, the cloven hoof is hidden in the implementation of the amplifier.

CCII+ is a current controlled current source.

Input current (control) is Ix, output current is Iz and pays Iz = Ix. Voltage Uy defines (ideally) voltage Ux = Uy (sometimes named as diamond transistor). That should be enough to solve the problem.

Yes, indeed.

The "cloven-hoof" is concealed inside the chip.

This method of schematics is obtuse

and requires a side-bar note for clarification.

Was Just browsing,

and came across your astute observation.

Hi Glen,

The "cloven-hoof" there are not inside the chip. It is due the realization of the amplifier with CCII. Exact description You can find in:

Research WIEN BRIDGE OSCILLATOR WITH REAL AMPLIFIERS

Josef

I remember these things from college days. They drift like crazy with any change in temperature, so we had to replace one of the resistors with a thermistor.

And do you remember how large was the R?

It could be maybe the effect of the non-ideal output resistance of CCII (no infinite value - and temperature dependence). If the R is small enough then all can be OK.

Here's a crude attempt to reproduce something like it. As I recall, Rx was the thermistor, and the thing ran at 1MHz

Hi Mark.

I think your picture shows an oscillator with a classic OPA (voltage-controlled voltage source) - not with a CCII (current controlled current source).

Hi Josef - do you consider the problem (your question) to be solved?

Hi Lutz,

your question is proper - to tell the truth. "Wien divider" in the circuit "contains" output resistance (R) of the amplifying structure - it is truth. At the same time the output resistance of the oscillator is non zero too. So we must "subscribe" current flowing via "output R" or we must use a voltage follower with high input impedance - to no change circumstances in the circuit. From this point of view the circuit is not klasic Wien oscillator. But it is used "Wien divider" there.

Some "thinking" to this circuit we can see in: Research WIEN BRIDGE OSCILLATOR WITH REAL AMPLIFIERS

1.) Well - for my opinion, the "heart" of the WIEN oscillator is the "WIEN bridge" which consits of the well-known RC bandpass (WIEN divider) and the two resistors with a 2:1 ratio. The error voltage across the midpoints (if tuned) is zero and requires an amplifier with infinite gain (resp. with 100 dB gain for a real and very small error voltage). So the question is: Can we detect and describe a WIEN bridge within the given CCII circuit? I don`t think so, even if we consider the CCII as ideal.

2.) By the way and in this context: Thinking of the classical opamp-based WIEN oscillator - there is an alternative circuit with exactly the same properties: Replace the WIEN divider (bandpass) with the classical passive RC-CR bandpass. Surprisingly, this circuit is never (or very seldom) mentioned in textbooks, although the oscillator has IDENTICAL properties if compared with the classical WIEN oscillator (same parts values, same divider ratios, same gain).

Or - do you know some pros or cons of this circuit?

Joseph,

Your answer is properly worded , my wording was obtuse. Thank you.

Indeed, the 'hoof' is in the "realization" of the circuit, as a whole unit.

...

This is an interesting , short study. Lutz has a good classical perspective. IMHO.

...

I was involved in an active filter group several years back, ,

wherein the Designer Utilized the Internal Output Impedance of an OpAmp as part of the tuned circuit.

When the chip was changed , then the circuit failed. Analysis was difficult in that group.

[ Barrie is still busy as ever. Amazing engineer. ]

Hi Glen,

in this case we do not use the output resistance of OPA. For the CCII it is ideally infinite. It is property of the "connection". If the output resistance is "finite", it is parallel connection of R ant this output resistance.

Josef