Maker Pro
Maker Pro

Jim Thompson's gyrator at VLF

Darren Holdstock

Jan 20, 2009
5
Joined
Jan 20, 2009
Messages
5
Hello helpful fellow designers, might I draw upon your expertise?

I'm designing a low-noise VLF bandpass filter right now, and the 4 Hz centre frequency necessitates the use of some kind of gyrator-type circuit. After much shopping around the clear leader so far is Jim Thompson's classic architecture, but I'd like to check out some others just to validate this. Has anyone got links to the array of gyrators from Winfield Hill, and/or some design equations for the Riordan architecture? I would be most grateful. I'm unlikely to use the Riordan as the output noise is too high (unlike JT's which is many orders of magnitude quiter), but it would be nice to have a handle on it as information is sparse, and when it's not sparse it's misleading or just plain wrong.

I did come across something that may be of use to those designing JT's gyrator (analog-innovations.com/SED/GyratorFilterBP.pdf): As the last stage of the PSPICE sim I thought to check the signal swing at the op-amp outputs, and a good job I did as the values I'd calculated necessitated an op-amp with 700 V rails. Some hacking later and I'd found the problem - the signal magnitude of the op-amp outputs is given by the equation

Vopamp_output = Vin * R5 / R3

If R5 = R3, then the outputs won't saturate if the input doesn't exceed the op-amps headroom.

Anyhow, this got me thinking about the importance of keeping an eye on the gain-bandwidth product of the amplifier, and how the relationship between R5 and R3 might affect the gain peaking characteristic. Gain peaking shouldn't be a problem for me on this occasion, centred around 4 Hz as I am, but if R5/R3 affects this then it might be useful information.

Any links to gyrator circuits gratefully received - the information out there really is quite patchy.

Much thanks, Darren.
 
Top