• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 49 subscribers
  • Views 172 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TLV170: Is adjustable Fc with fixed Q possible?

Robert Valentine
Prodigy 205 points
Part Number: TLV170
Other Parts Discussed in Thread: TLV9101, TLV9301,

Hello,

I'm working on an upgrade to an analog front end in a product and I would like to take a data acquisition input channel from fixed Fc to a configurable Fc via MCU.  I've been reviewing the sallen-key and multi-feedback topologies (I'm using the sallen key now) and have run into a problem with both topologies.

Using the following two articles (sallen key and multi-feedback App notes):

https://www.ti.com/lit/an/sloa024b/sloa024b.pdf?ts=1659502637496&ref_url=https%253A%252F%252Fsearch.brave.com%252F

https://www.ti.com/lit/an/sboa114/sboa114.pdf?ts=1659503288804&ref_url=https%253A%252F%252Fsearch.brave.com%252F

They both indicate that the Q value of the filter is tied to the Fc by way of the resistor and caps.  It seems that the corner frequency is tuned by some of the same passive components the Q values are tuned with.  This means I can't adjust the corner frequency of the filter without affecting the Q value.  This is a problem because I want the Q value to stay at 0.707 so I don't have excessive attenuation or peaking which could cause overshoot on fast edge rates.

Is it possible to use one of these topologies or a different topology to design a filter supporting a configurable Fc while keeping the Q the same value?  I was planning on using a digital POT to modify the resistors for different Fc values, but now I see doing this will cause the Q to change every time I adjust the Fc of the filter.  At this point I think I've gone down a fundamentally incorrect path.

Thanks for any input!

  • 0
    Guru 140200 points

    Hi Robert,

    in a unity-gain Sallen Key low pass filter the Q is programmed by the ratio of the two capacitors while the two resistors are kept equal. You can easily tune the corner frequency without changing the Q by tuning both the resistors. But keep them always equal.

    This technique can also be seen in some active crossover networks where the same is done with the unity-gain Sallen Key high pass filter. By tuning all four resistors at the same time the crossover frequency can easily be tuned.

    If you want to additionally tune the Q by the help of a resistor tuning, you should look at the "state variable filter".

    Kai 

  • 0 in reply to kai klaas69
    TI__Mastermind 22240 points

    Hi Robert,

    Kai is correct. We also have this other application note called the Active Low-Pass Filter Design that might provide more information on how to design Sallen-Key and MFB. 

    We also have newer devices that can be used rather than the TLV170. I recommend taking a look at the TLV9301 and the TLV9101. 


    Best Regards,

    Robert Clifton 

  • 0 in reply to kai klaas69
    Prodigy 205 points

    Hi guys,

    Thank you so much for the reply and your time!  I see now that setting the resistors equal and capacitors to different values allows for the tuning of Q and fc.  

    So just as a sanity check, this means I could used digital pots as the fc tune option and set my Q to ~0.707 with the two caps?  I'm seeing that when modifying the equations around equation 9 on page 4 in the app note setting r1=r2.

    This is great news but does however mean I will have to have 2 digitally controlled pots per sallen key stage.  Do either of you know if there is a smarter way customers approach this type of thing that could use less pots?  It would be nice if I could increase the stage count to two sk stages yielding 4 poles but this makes component count quite high.

    The tricky part in my situation is the 0-3.3v single supply analog front end power as opposed to a lot of analog front end using negative supply rails like +/-2.5 or +/-5.0 etc...

    I will also look at changing my bom and doing some simulation to incorporate the newer op amp you recommend.

    Thanks again for the help!

  • 0 in reply to Robert Valentine
    TI__Mastermind 22240 points

    Hi Robert, 

    It depends! Is the customer needing to have the circuit dynamically change the fc or are there only a few specific fc that they care about? 

    If they need it to be dynamic then I think the digital pots might be the best option for them. 

    Best Regards,

    Robert Clifton 

  • 0 in reply to Robert Clifton56
    Prodigy 205 points

    Hi Robert,

    I really only need some preset corner frequencies.  So, giving the user say 8 or 10 different corner frequencies to select from.  I thought about passives and a switch/mux but that approach drives the component and board realest up so much and my best way to control this feature turns out to be i2c due to my MCU having no free discrete io.

  • 0 in reply to Robert Valentine
    TI__Mastermind 22240 points

    Hi Robert, 

    You already see where my mind was heading! Yeah then I think the digital pot might be one of the better options, though there might be some alternatives that might work. This starts falling outside of my expertise in op-amps though! 

    If you no longer have any more questions regarding the op-amp, I'll go ahead and close this thread. 

    Let us know if you have any further questions! 

    Best Regards,

    Robert Clifton