Maker Pro
Maker Pro

twin T circuit

L

lerameur

Jan 1, 1970
0
Hello,

I would liek to knwo if someone has the equations for a Twin tee RC
active filter. I am doing the circuit from analof filter from Les
Thede. I do the circuit they have in the book , It looks very good on
my simulation, but when I try ti implement my circuit, I am way off.
My stop band is wider then his example, Does Twin tee only take
narrow band ??
thanks

ken
 
E

ehsjr

Jan 1, 1970
0
lerameur said:
Hello,

I would liek to knwo if someone has the equations for a Twin tee RC
active filter. I am doing the circuit from analof filter from Les
Thede. I do the circuit they have in the book , It looks very good on
my simulation, but when I try ti implement my circuit, I am way off.
My stop band is wider then his example, Does Twin tee only take
narrow band ??
thanks

ken

Lots of web sites. Here's one:
http://www.drp.fmph.uniba.sk/ESM/twin.pdf

Ed
 
P

Phil Allison

Jan 1, 1970
0
** Groper Alert !
I would liek to knwo if someone has the equations for a Twin tee RC
active filter. I am doing the circuit from analof filter from Les
Thede. I do the circuit they have in the book , It looks very good on
my simulation, but when I try ti implement my circuit, I am way off.
My stop band is wider then his example,


** Got a low drive impedance and high load impedance ??



......... Phil
 
J

john jardine

Jan 1, 1970
0
lerameur said:
Hello,

I would liek to knwo if someone has the equations for a Twin tee RC
active filter. I am doing the circuit from analof filter from Les
Thede. I do the circuit they have in the book , It looks very good on
my simulation, but when I try ti implement my circuit, I am way off.
My stop band is wider then his example, Does Twin tee only take
narrow band ??
thanks

ken

Few people will have the expensive Theade book. Best if you could post the
circuit somewhere, as there are any number of possible TwinT arrangements.
Would suspect the wide stopband is down to not enough gain from a feedback
opamp, leading to a low Q notch. Use a pot' for the feedback resistor. Q's >
say 5 are usually quite critical as to the gain setting resistor value.
Fractionally too much gain and the whole lot oscillates.
 
L

lerameur

Jan 1, 1970
0
** For high Q factors - ie a sharp notch use this circuit.

http://www.national.com/ms/LB/LB-5.pdf

....... Phil

Thanks for the replies.
I am looking for a circuit with active element not passive.
The Lm102 looks good but there is not enough tuning.
My bandstop extend from 1k to 10Khz
at 2Khz i want it to be at -30dB

here is a link to my circuit wi th the results, Its almost there, But
I think this is the maximum i can obtain right ?
http://users.encs.concordia.ca/~jk_ouell/

the %7E is a tilt, works either way :)

ken
 
J

john jardine

Jan 1, 1970
0
lerameur said:
Thanks for the replies.
I am looking for a circuit with active element not passive.
The Lm102 looks good but there is not enough tuning.
My bandstop extend from 1k to 10Khz
at 2Khz i want it to be at -30dB

here is a link to my circuit wi th the results, Its almost there, But
I think this is the maximum i can obtain right ?
http://users.encs.concordia.ca/~jk_ouell/

the %7E is a tilt, works either way :)

ken

Circuit seems massive overkill.
Just one stage could give better results than what you're showing. (also
need a lot more data points in your sim).
Is that overwritten "RU4841" gain resistor on the left really "90" ohms and
not 30k?.
Why R160 and C90?. Seems no reason for 'em.
 
P

Phil Allison

Jan 1, 1970
0
"lerameur"


** Groper Idiot Alert !

Thanks for the replies.
I am looking for a circuit with active element not passive.


** Then fucking say so - you moron.

The Lm102 looks good but there is not enough tuning.
My bandstop extend from 1k to 10Khz
at 2Khz i want it to be at -30dB


** Fucking Say What YOU want in the FIRST POST -


YOU ASININE , ASD FUCKED , WOG IDIOT !!!!!!!!!!!!






.......... Phil
 
L

lerameur

Jan 1, 1970
0
Circuit seems massive overkill.
Just one stage could give better results than what you're showing. (also
need a lot more data points in your sim).
Is that overwritten "RU4841" gain resistor on the left really "90" ohms and
not 30k?.
Why R160 and C90?. Seems no reason for 'em.

actually they are pretty usefull,
AS for the 90 Ohm resistor, It do not matter to what I chosse, it is
again. From what I calculated, it did make any sens, I obtained a
negative value

I have joined on the web site the steps that I used to obtain my
circuit (in a pdf).

ken
 
J

john jardine

Jan 1, 1970
0
lerameur said:
actually they are pretty usefull,
AS for the 90 Ohm resistor, It do not matter to what I chosse, it is
again. From what I calculated, it did make any sens, I obtained a
negative value

I have joined on the web site the steps that I used to obtain my
circuit (in a pdf).

ken
The TwinT acts like a series resonant LC circuit and rejects only -one-
particular (designed) frequency. Rolloff slopes dependant on designed Q. You
will not achieve the 1k-10k bandstop using it.
You should be designing something like a third order 10kHz low pass followed
by a third order 1kHz high pass.
 
L

lerameur

Jan 1, 1970
0
The TwinT acts like a series resonant LC circuit and rejects only -one-
particular (designed) frequency. Rolloff slopes dependant on designed Q. You
will not achieve the 1k-10k bandstop using it.
You should be designing something like a third order 10kHz low pass followed
by a third order 1kHz high pass.

Would that create a bandstop or bandpass ??

k
 
J

john jardine

Jan 1, 1970
0
lerameur said:
Would that create a bandstop or bandpass ??

k
Oops, yes, bandpass. (my bad!)
Having reread your wants, it looks you're after a helluva bandstop, needing
at least 5th order filters plus some kind of (probably inadequate)
summation.
Maybe yes, that 3 section TwinT is looking the way to go.
Fired up the filter prog and poked around a bit. There's lots of
arrangements that will do the job nicely but they all use thousands of bits.
For minimum components and in line with what you've already built, I looked
at the TwinT arrangements.
As you calculated, your arrangement needs a negative resistor in the first
stage, so that's out.
Using inverting biquads gets rid of the negative element but then there's a
problem with poor rejection.
Further poking around turned up a nice 3rd order Elliptic design that will
nicely do the -30dB at 2kHz (1% ripple).
Total time steering the prog' was about 15 minutes. Wouldn't have dreamt of
doing the same manually!.
If of interest, I could put the circuit diagram on A.B.S.E. or email it.
 
C

Charles Schuler

Jan 1, 1970
0
lerameur said:
Hello,

I would liek to knwo if someone has the equations for a Twin tee RC
active filter. I am doing the circuit from analof filter from Les
Thede. I do the circuit they have in the book , It looks very good on
my simulation, but when I try ti implement my circuit, I am way off.
My stop band is wider then his example, Does Twin tee only take
narrow band ??
thanks

Component tolerance is a killer with Twin-T circuits ... that's most likely
why your simulation works and your breadboard does not.
 
E

ehsjr

Jan 1, 1970
0
lerameur said:
Thanks for the replies.
I am looking for a circuit with active element not passive.

The passives determine the frequency in the Twin-T. Mismatched
passives will minimize the notch depth. The last time I built
one of these, the math predicted great results, but real world
components yielded disappointing results. By analogy, you may
be focussing on improving the transmission in your car, when
the real problem is that the engine doesn't work.

Ed
 
Top