questions about two-thomas biquad filter

[lily1981216]

tow-thomas filter

Hi, everyone, I want to design a two-thomas biquad filter with 1.2MHz cutoff frequency. But how to compute the value of the passive components(R, C)?And by the way, who have the two papers about filters?
L. C. Thomas, The Biquad: Part I - Some Practical Design Considerations
sALLEN KEY , A practical method of designing Rc active filter

I can't download the papers and if someone have these, please help me. Thanks very much!

 

[LvW]

l. c. thomas biquad

Both papers (Thomas from 1971, Sallen/Key from 1955) are rather old, but are the basis of classical RC filter structures. Thus, their contents is reflected in several books on filter design and, perhaps, you don´t need the original publications.

Concerning the Tow-Thomas structure please see the enclosed pdf sheet.
But be aware that for a pole frequency of 1.2 MHz you need amplifiers which have a unity gain bandwidth of at least 100 MHz. These integrator topologies are rather suszeptible (sensitive) to parasitic opamp phase shifts.
Regards and good luck.

 

[LvW]

tow-thomas 2

Explanation to the pdf paper: Ao is the low pass gain and Am is band pass midband gain.

 

[mengcy]

tow thomas biquad

You could download a filter solution(verision 10.0),it could calculate the passive components parameters for you!

Good luck!

 

[lily1981216]

biquad filter design

Both papers (Thomas from 1971, Sallen/Key from 1955) are rather old, but are the basis of classical RC filter structures. Thus, their contents is reflected in several books on filter design and, perhaps, you don´t need the original publications.

Concerning the Tow-Thomas structure please see the enclosed pdf sheet.
But be aware that for a pole frequency of 1.2 MHz you need amplifiers which have a unity gain bandwidth of at least 100 MHz. These integrator topologies are rather suszeptible (sensitive) to parasitic opamp phase shifts.
Regards and good luck.

Thank you very much for your paper and it's quite useful. But there is a little confusion to ask you. Does the wp in the paper mean the cutoff frequncy?
And if I use a differential structure to design the tow-thomas filter, should the values of the passive components be modified? Thanks again!

 

[LvW]

pdf biquad

................. Does the wp in the paper mean the cutoff frequncy?
And if I use a differential structure to design the tow-thomas filter, should the values of the passive components be modified? Thanks again!

1.) As outlined in th paper - wp is the pole frequency. This is NOT the cut-off frequency, but in the vicinity of it. These pole data are tabulated for different lowpass approximations in textbooks (or can be found elsewhere). They are related to the wanted cut-off - depending on the filter degree and the approximation.
2.) In differential topologies, the component values can remain unchanged.
--------------------
Do you really need this specific filter structure with three opamps ? It´s not the best, unless you need a lowpass and a bandpass output at the same time.

LvW

 

[lily1981216]

the biquad: part pdf

LvW, I greatly aprecciate all your replying! It's quite useful! Thank you!
I select the structure, because I want to design a frequency tunned filter. And I saw some papers about frequency tuned filter select the tow-thomas filter and introduce the structure that the frequency could be tuned just by modifying the the cap value. In fact, I don't know some other structures more suitable to the frequency tuned filter with the maximum cutoff frequency above 1MHz.

 

[ttomm]

analog biquad filter topology

I'll second mengcy's suggestion: download FilterFree from nuhertz.com, and it'll give you the component values in a tow-thomas circuit based on your needs. Select butterworth, implementation of"active", and "thomas 1" or "thomas 2" for the topology.

 

[LvW]

midband gain in tow thomas

I select the structure, because I want to design a frequency tunned filter. And I saw some papers about frequency tuned filter select the tow-thomas filter and introduce the structure that the frequency could be tuned just by modifying the the cap value. In fact, I don't know some other structures more suitable to the frequency tuned filter with the maximum cutoff frequency above 1MHz.

Hi lily19811216,

may I start with a general remark: To select one filter circuitry (out of perhaps 20...30 alternatives) is a very difficult and tricky task which always results in a compromize. This compromize depends strongly on the requirements and the operating conditions. Therefore, it would have been better from the beginning to say something about your basic requirements:
Tuning capability and frequency "above" (above???) 1 MHz. (By the way: 1 MHz or even more may be a problem for Tow-Thomas, which is rather sensitive to phase errors which accumulate due to 3 opamps in series !)

But that´s not sufficient.
Therefore my questions:
1.) Low pass of which order ? (I hope: n=2, otherwise tuning is a problem).
2.) Tuning across which range ? (1% for final adjustments or 20....50 % ?)
3.) What about pole Q ? Constant across this range ? Or do you require the bandwidth to be constant ? Or does it not matter at all?
4.) What about the filter gain ? Constant during tuning ?
5.) Which part to be changed during tuning ? R or C ? Any preference ?
Only one part ? (Because in many cases two parts in parallel are to be tuned) .

I think, you now can agree that filter design is a real "design task" and cannot be done by a filter program. Such a software can only calculate parts values - at the end of the design process ! The first 98% of the job are up to you.
Regards
LvW

Added after some minutes: Sorry, cancel "bandwidth" in point 3.) (I had a bandpass in my mind)