Simple gyrator circuit

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

peter purpose

Well-known member
Joined
Jun 3, 2004
Messages
2,082
Location
London
Gentlemen,
I'm thinking of simulating the 1H inductor in an API 553 eq type circuit and came across this.

gyrator.gif


Would you even bother replacing a real inductor with this, or am I pissing in the wind?

Cheers
peter
 
With no return path for V+, it's a do-nothing circuit.

If you connect the low side of R3 to an external circuit, it's going to want to draw its emitter current through that circuit, which is probably not going to be welcome.

The schematic seems incomplete.
 
Peter

I was messing with a gyrator circuit from some website to design a graphic for a college assignment. Never managed to get it working.

It was based on tis website I think

http://www.geofex.com/Article_Folders/EQs/paramet.htm

Don`t know what I was doing wrong. It`s probably cos I`m thick as you know !!
 
Thanks Dave,
I was stumped as to where to inject the signal. Would this make sense?

gyrator.gif


Rob,
I'd be surprised that anything you make would actually work.. :green:
 
No---none of the circuits so far work.

The standard lossy synthetic inductor uses a unity-gain buffer, which in some cases can be a simple emitter follower if you are not too concerned about performance. That is, the finite input and output impedances of the e-follower will degrade things. If you use a complementary feedback pair matters improve dramatically.

There are two common forms of UGB-based synthetic inductors, and the information can be found in a wikipedia ref and associated links:

http://en.wikipedia.org/wiki/Gyrator, which shows the commonest form,

and from Fred Forssell's site:

http://www.forsselltech.com/schematics/Gyrator1.htm,

for the equal-valued R version. Usually this version is more useful for relatively small L values and associated higher frequency resonators.

One of the drawbacks of the first circuit is that the equivalent inductance you get has not only an R in series but also an R in parallel. If you work to make the latter R very large the noise gain gets high.

Note also that these circuits will only simulate inductors with one end connected to common. Synthesizing good floating inductors is much more complex!
 
[quote author="NewYorkDave"]With no return path for V+, it's a do-nothing circuit.

If you connect the low side of R3 to an external circuit, it's going to want to draw its emitter current through that circuit, which is probably not going to be welcome.

The schematic seems incomplete.[/quote]

It's not a complete circuit, but it's a complete (simulated) inductor.

It's a typical circuit to emulate a (quite lossy) inductor for filtering
the supply voltage. In that case, the circuit to be supplied with voltage
is the current return path to GND.

Same thing, only using a FET: (page 3 + 4)

JH.
 
[quote author="bcarso"]And the damn thing is floating too![/quote]

Yes. You could connect the capacitor to GND instead - that would give lowpass filtering of PSU ripple (mostly) independent of the load. Or, they way it is drawn here, capacitor connected to the emitter / source side,
emulates an inductor, i.e. the filtering is done by the interaction of the gyrator circuit and the load.
I don't know which method is better in an absolute sense, but that gyrator is great when you want to emulate a huge piece of iron in a traditional power supply without changing the general behaviour of the original.

I don't think this is any good for signal processing, though.

JH.
 
[quote author="jhaible"][quote author="bcarso"]And the damn thing is floating too![/quote]

Yes. You could connect the capacitor to GND instead - that would give lowpass filtering of PSU ripple (mostly) independent of the load. Or, they way it is drawn here, capacitor connected to the emitter / source side,
emulates an inductor, i.e. the filtering is done by the interaction of the gyrator circuit and the load.
I don't know which method is better in an absolute sense, but that gyrator is great when you want to emulate a huge piece of iron in a traditional power supply without changing the general behaviour of the original.

I don't think this is any good for signal processing, though.

JH.[/quote]

But---yes not any good in its present form for signal processing but if "suitably" refined...might be fun as they say.

In an old automotive amp at H*rman, Brad Plunkett did something a bit similar and it came to be called colloquially the Plunkductor, but I don't think he ever analyzed it in terms of being a gyrator of sorts. It "got out of the way" when there was some moderate signal present and became more or less a short circuit.
 
It has appealing simplicity, but note that it requires some unipolar current in addition to any signal current---that is, the current can not go to zero or reverse or the circuit doesn't work at all.

What is fun about it is that it uses some intrinsic properties of transistors---high collector/drain impedance, for example---to achieve its purpose. Jurgen's MOSFET version has the advantage of a hgih gate resistance as well so it doesn't load down the RC network so much.

One could implant it in the middle of a diode bridge for bipolar operation although the crossover distortion would be terrible.

BTW, while looking for an online schematic of synthetic floating inductors I ran across this great ref by Dennis Bohn, which didn't give me what I wanted but is wonderful in its own right:

http://www.rane.com/note122.html

EDIT: A really good, albeit necessarily terse, slide show tutorial on active filters including a bunch of OTA topologies I ran across:

http://msic.ee.ncku.edu.tw/course/AdvIC/slides/ch8.pdf

This guy (?) knows his stuff.

EDIT 2: Knock me down with a feather---this is from Taiwan, as I found backing up a bit with the web addy: "Welcome to the National Cheng Kung University Electrical Engineering Mixed Signal Integrated Circuit Lab!!!
 
Marvellous stuff Brad, thank you very much.

I think it high time I bought myself a publication explaining the difference between diodes and doughnuts.

peter :thumb:
 
[quote author="peter purpose"]Marvellous stuff Brad, thank you very much.

I think it high time I bought myself a publication explaining the difference between diodes and doughnuts.

peter :thumb:[/quote]

I think there was a collaborative effort to produce a document underway among Sara Lee, Betty Crocker, and the Keebler elves, but it deteriorated when the elves got into Sara's cooking sherry and began leering at Sara while discussing the distinction between doughnut holes and electron-hole pairings.

Duncan Hines arrived late and was evidently unable to salvage much from the conference, save the baked goods.
 
That will get you going for a grounded synthetic inductor. The active portion is just a unity-gain amp with high input Z, and could be higher performance but should work. For that matter you could use an IC op amp. The two gate R's are not needed---the d.c. ref is through the 10k's to ground, and the other gate is directly tied to the output.

A variant circuit which also uses a unity-gain amp is more often seen, but I was told once that this one has advantages at high frequencies sometimes.
 
Cheers Brad.
When you say 'higher performance', what would you like to be seeing?
More current? More stages of amplification? Racing stripes?

peter
 

Latest posts

Back
Top