diyAudio (
-   Equipment & Tools (
-   -   Build -- Active Twin-T notch filter for distortion analysis (

richiem 9th May 2011 08:04 PM

Build -- Active Twin-T notch filter for distortion analysis
I've built an active Twin-T filter that works pretty well in conjunction with a spectrum analyzer -- I use the software analyzer ARTA under CrossOver Mac on my hackintosh.

The details are here:
Active Twin-T filter

Nulls as low as -90dB are pretty easy to achieve, and -100dB or more can be obtained for a short time. Deep nulls are not needed, however, when used with a spectrum analyzer -- you're just getting the fundamental level down low enough that the analyzer's or PC's input circuits are comfortable with the level and its own distortion doesn't intrude; -60dB to -80dB works great.

hesener 9th May 2011 09:25 PM

very nice!

jez 11th June 2011 02:30 PM

Just been reading your web site... very nice work! Did you get any further with that HP based oscillator?
I fear you may have missed out on something very nice in the form of the Morrey modifications to the IG18 though ;)
I don't have an IG18 but built the Morrey version from scratch, with a few changes of my own such as the regulated PSU and using a home-made "LDR and LED in a tube" instead of the specified Vactrol part. This required some changes to other feedback resistor values etc.
It performs very well indeed. My homebrewed distortion meter has a noise floor of 0.0014% (on the lowest range of 0.01% FSD) and after very careful nulling of the meter this noise floor is unchanged by the output of the oscillator!
One can surmise then that the actual distortion must be considerably lower than the 0.0014% mark.
Tests using PC based analysis merely showed the specified 0.002% of the USB audio interface.

richiem 11th June 2011 05:58 PM

@jez -- I was attracted to the Morrey mods when they first appeared, then the problem reports started to show up. I was a bit put off by the complexity, though, and so never built one. The mod of the IG-18 to the HP 239 circuit is on hold for the moment, waiting for further small mods to the discrete circuit in IG-18 #1.

Try using a notch filter between your Morrey and the PC spectrum analyzer -- even a passive one can give you a lot of info if you adjust the displayed levels of the harmonics for the attenuation factors of the filter. A passive Twin-T will attenuate the 2nd H by roughly 9dB and the 3rd by roughly 5dB.

jez 11th June 2011 06:35 PM

As a matter of interest what exactly are the "problem reports" with these mods?
I know I had quite a bit of trouble with loop stability causing amplitude bounce to build up in an oscillatory way...

klewis 13th September 2011 11:01 PM


I've been building your twin t notch filter. I'm almost done. I was wondering about the user of the three 1k trimmers. How are they used in setting up the filter? Should I initially set them at mid point values?

I'll post build photos next week (going out of town for the weekend). Hopefully it starts up and runs right off the bat. I enjoy reading your web entries and your re-greening projects.



richiem 14th September 2011 01:26 AM

This may turn out to be a duplicate post -- if so, sorry

@jez -- Sorry I didn't see your note -- I don't remember well what I read about the problems in AA, but I think they mainly had to do with stability issues.

@klewis -- I look forward to seeing your build. I initially set the three 1k trimmers to midpoint, then I set the filter to 1kHz, which is what I use most. I balanced as best I could with the coarse and medium switches, with the 10-turn pots at mid-point, then got the best null I could with the trimmers. For me, that worked OK at the ends of the range, keeping the ten-turns away from their end points at all frequencies. It takes a lot of fiddling around to get it all to work on all frequencies of all ranges. Enjoy.

klewis 15th September 2011 04:38 PM

3 Attachment(s)

Attached are progress phone photos. I found the same tin at Michael's and added a 1/16th inch aluminum to the inside of the top to stiffen it up a bit. I etched a small board for the opamp section and used the quad opamp in lieu of singles. The resistors are on the bottom of the board. I also included the out bnc in the board. The stepped switches are done, but, not in the photos.


richiem 15th September 2011 07:20 PM

Adding the stiffening plate is a great idea -- wish now I'd taken the time to do that.

klewis 19th September 2011 02:01 AM


Two more questions for you.

1. Gain SW-4: What does pin one (0db) connect to? or does it just float?

2. Resistor R148 value is 1N? (on leg between C4 and OPA134 U1)

Thanks again.


All times are GMT. The time now is 10:35 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2