Photoelectric gyrator

Not really, but get a better look, and explanation at...

Thanks for the link - it's nice to be able to get the documents in PDF format.

I built a little voltage controlled floating gyrator circuit with an LM13700 that I was pretty proud of until I realized it had already been done! The circuit is flawed, though, in the fact that trying to divide down the input voltage to fit in the linear range of the LM13700 appears as a resistor in series with the simulated inductor. Coupled with the low transconductance of the LM13700 this destroys the Q of the simulated inductor except at the highest frequencies the 13700 is capable of, unless the input signal is restricted to about 20 mV peak to peak.

This IC:

looks more promising for building a voltage controlled inductor with larger input voltages, and for higher frequencies.

I've looked at at that one for a few jobs, but it would never quite work. It's faster and beefier than an LM13700, it's shy one input and is _horribly_ nonlinear.

Cheers

Phil Hobbs

Arwe gyrators really useful? For making filters, other active filter forms seem more practical.

John

Are you saying state variable filters are more practical? Since you can spin LC implementations with gyrators, you end up with lower component sensitivities than state variable filters.

Of course. I settled into gyrator implementations by ~1980.

Larkin is clueless, as usual, but just has to keep posting, hoping for attention ;-) ...Jim Thompson

Kind of seems that way. Or sallen-key, which take less parts.

Since you

I just don't see many gyrator-based filter designs. Simple gyrators synthesize grounded inductors, which limits things some.

John

I'm generally underwhelmed by fancy analogue active filters--their dynamic range is the pits and they take a lot of parts for what they do. I'm sure there are applications where those aren't horrible problems, but I rarely encounter them. Chip design folks live in a different space, of course, where matching is easier and parts are nearly free by comparison.

Once or twice in my career I've used things like the LMF100 Swiss Army knife switched cap filter, despite its even worse dynamic range, because I could get the selectivity I wanted, and the dynamic range requirement had been reduced by the superhet front end.

Cheers

Phil Hobbs

The place they are useful is at low frequencies, where inductors aren't practical. Above roughly 5 MHz, real inductors win.

Yup, switched cap filters are terrible. Noisy in their own right, plus they alias everything in sight, including their own power supply noise. You generally have to pre- and post-filter with RCs at least. Still, for high-level stuff, the size and tunability are nice.

I did a cool double-conversion superhet FSK modem, for the Reuters news wire system, just before the internet blew the technology away. Full of MF10s and 4000-series CMOS. It worked pretty well, if only because the phone lines were pretty noisy to start.

John

My prejudice is to lower the circuit impedance and use inductors wherever possible. They do make pretty good SMPS antennas, unfortunately. For fancy stuff, I'd way rather use some fairly crude filter for antialiasing and do the filtering in software, or use a superhet.

Cheers

Phil Hobbs

Properly designed gyrator-based filters don't have nearly the dynamic range problems that state-variable and other similar forms have. ...Jim Thompson

So I'm told--much more like LC filters, as one might expect. However, building floating gyrators is famously difficult, so people use 'double capacitors', i.e. multiply all the admittances by j omega, so inductors turn into resistors and capacitors turn into 'double capacitors'.

Never been that excited by it myself.

Cheers

Phil Hobbs

There _are_ tricks to fake floating gyrators ;-)

There are also ways to do state-variable and like-kind filters that don't have internal gains >1. Some of us know those tricks ;-) ...Jim Thompson