Twin T notch filter design quick help

[acehobojoe]

I have calculated the way I want the notch filter to work using this site: http://earmark.net/gesr/opamp/twint_notch.htm

however I know there is also a formula. fc     =     1 / (2 pi R C)

I think I'm doing it correctly. Based on the site, in order for me to get a notch at around 200hz with a Q of around 2, I will need :

C = 180nF
Ro = 4420
Rq low = 287
Rq high = 34000

The only way to see if this works is to test it out, I'll try it. If this is completely off, let me know! I also probably don't want to have complete attenuation, so I'll probably need a way to make it less deep of a notch.

I also heard that there is an easy way to achieve the reverse of this with the op amp.

Anyway. thanks.

[GibsonGM]

I agree, the only way to know if YOU like how it sounds, is to test it out!   
 
I do have to ask, why did you pick 200Hz for your center frequency, though?  That is quite low - you may find the low end of the filter to be unuseable - the notch is relatively wide, and that center freq shifts it left quite a bit.  1K may be a better place to set it initially, IMHO.

Did you know you can do much the same thing with the Big Muff Pi tone stack?  AND you can set it up to make the mids adjustable!  You can control the amount of cut, so no worrying about the depth of the notch.   There are a few easier ways to do it than with the op amp - with fewer parts - guitar is not that sensitive as to need so much precision.   But, try it out, by all means!   Even if you don't use it for this, you might find something you want to use it for later.  Breadboards are great that way....

An example:  http://www.muzique.com/lab/tone3.htm

[Kipper4]

Check htis recent AMZ addition too Joe

http://www.diystompboxes.com/smfforum/index.php?topic=109705.0

[acehobojoe]

Thank you guys! these are plenty help with what I'm trying to do. I just want to get new ways to  change tone in my circuits, the 200hz bump was there because I find many players don't like the 200-300 "mud" in there signal, though I'm not sure what people would prefer.

these two tone things are helpful!

[acehobojoe]

So, I've been testing a few things on ltspice and I cannot get a very good notch at the 200hz where I would like.

I am getting a very steep notch! and I'm going to keep trying to figure out how to make it more subtle. It would  be nice for it to be only passive components, but if I can make it a bit better with an op amp that would be fine.

[antonis]

I am getting a very steep notch!

Quite expected if your design has a relative high Q ...

[GibsonGM]

Maybe just use a PASSIVE notch filter, then?  It has a much lower Q, with wider bandwidth as a result.   In other words - many already-built TONE STACKS have just this built in!   Use your opamp to recover from insertion loss.

Or, use an online calculator and set the Q lower, which will make the bandwidth wider... 

[acehobojoe]

I guess I should have asked how to make it less "deep" not steep. steepness is fine, I just don't want 90db of reduction.. haha

[antonis]

I just don't want 90db of reduction.. haha

Just don't deal with 100th order filters.. haha

[R.G.]

The Twin T is a special case of RC filters. It generates its notch by phase shifting the signal two directions, one in each "T" path, and then mixing them. The phases are arranged to cancel out at the resonance of the filter. Because it's a phase cancellation, not an attenuation, the Twin T is one of the very few passive RC filters that can generate an infinitely deep notch - that is, complete cancellation.

It can only do that when the components are values at the ideal values. If the cap or resistor values are not what the equations say they should be, but instead are off a bit, the notch will not be infinitely deep.

The "Q" of a Twin T is also a bit unusual, in that it is limited to about 0.5. It's not very sharp, even if it's perfectly tuned to be very deep. There are tricks with opamps to sharpen up the Q to any reasonable level. Some of these are recounted on geofex.   

If you want a less-deep notch, but sharper (that is, not as wide a dip) then you're either going to have to tinker with offsetting the component values, or adding opamps, or both.

[PRR]

> using this site: http://earmark.net/gesr/opamp/twint_notch.htm
> need a way to make it less deep of a notch.

I've also added two resistors that allow some control over the Q, but they also affect notch depth.--- the trim resistors just make the notch a little wider

Also consider not using a "calculator" which ASSumes and FORCES exact-equal values. You don't get exact-equal in real life. Not-equal is sometimes a good trick for "reducing perfection" (spoiling an infinite null).

[acehobojoe]

Got it a little less than 24db. still going to tweak.

[R.G.]

For some reason, my browser covers up your posting with the circuit and some traces.

However, if that's the ciruit, it's not a twin T. It's a bridged-T, which is a different animal.

[acehobojoe]

I'm guessing the bridged t is better for complete attenuation, I'll try the twin t and see what I get.

[PRR]

> a little less than 24db

180K to 10K is 25dB. The leverage is in the right place and the ratio is almost right. I'd jack the 10K to 20K, 40K, 100K, etc.

[acehobojoe]

Thank you,
Now we're talking:

It's much better, just around 1db of reduction, I'll work on the Q.

I was just confused because of the graph.
The graph on this program fluctuates with different values.

[acehobojoe]

An updated pic: http://i.imgur.com/3XjXIMw.png

My BRIDGED T filter seems to be doing well, I would still like to keep trying to decrease the q, but that about as much as it will go without completely going crazy.

about 1.5 db of attenuation at the 220-240 range is what I was aiming for.