"A Continuous-Time Active Filter", AudioXpress Magazine, Aug 2010

[Dr. Ram]

Thought of sharing this just-published article with friends on DIYAudio. If you happen to get hold of this issue and read the article, please feel free to ask questions or shoot feedback. The first para of the article appears below. The article itself is only a summary of 2 pages, but does have plenty of schematics.

"Active (crossover) filters are not a new idea, having been around for decades. However, if you are on the lookout for an all-analog filter that offers the ability to continuously vary the crossover frequency via a simple front panel potentiometer, the choices are limited. Furthermore, if you also want the flexibility to play around with different transfer functions—again, by simply flipping a front panel rotary switch—your choices narrow down still further. These twin requirements led me to design the AF-108 continuous-time active filter."

[jan.didden]

I gather that you are the designer of the unit and the author of the article?

jd

[Dr. Ram]

Yes - designer, builder and author. I should probably have said "...sharing my just-published article..." Thanks for asking :-).

Originally Posted by janneman

I gather that you are the designer of the unit and the author of the article?
jd

[CharlieLaub]

What you have built looks very nice. I have not yet read your article in aX, but I will look for it in my local bookshop.

As for how to implement a second order HP/LP transfer function, there is an IC that will do this, is time continuous, and is readily available: the UAF42 from Burr-Brown. I use this in various filters used for active speakers and other things. Q and crossover frequency are continuously adjustable via external potentiometers (one dual channel, one mono). Is this what you use in your project/product?

-Charlie

[Dr. Ram]

Originally Posted by CharlieLaub

...there is an IC that will do this, is time continuous, and is readily available: the UAF42 from Burr-Brown... Is this what you use in your project/product?

-Charlie

Thank you Charlie. I hope you will enjoy the article. Indeed the UAF42 is a great choice for active filters, and I gave it a hard look for this project. The UAF42's primary advantage is the fact that you have low-tolerance, stable caps built-in. I decided against it mainly because of the risk of working with delicate phase margins, since I wanted a unit whose frequency could be varied upto 100kHz. The VCA in the circuit (SSM2164) introduces its own share of phase shift as you approach that frequency, and I found that the op amp's phase margin spec started to matter tremendously for stability. To retain the flexibility of trying different op amps, I went for a design that would allow different ones to be plugged in rather than tie myself to the UAF42.

-Ram