• State TI Thinks Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 7 answers
  • Subscribers 16 subscribers
  • Views 4068 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.

Choosing switch for opamp gain manipulation (simple PGA circuit).

Mateusz Kaluzny17
Expert 1065 points
Other Parts Discussed in Thread: SN74CB3Q3251, TS12A12511, TS5A3359, TS5A3159, TMUX6104, TPS7A91, TPS7A39, TS5A22364

Hello dear Ti community.

Today I'm looking for some good IC for building simple PGA amplifier.

I'm not looking for many gain options, two/three/four will be more than enough :)
But to be honest two will be enough and easy to control (two state low or high).

There is so many switches that I don't know which one should I pick.
I know that Ron matter in some cases.

I'm going to implement it that way :



PS: should I use feedback capacitor to make the transition pop-less?
Or maybe there is another way to achieve stable and pop-less transition between positions?

  • 0
    TI__Mastermind 45440 points

    Mateusz,

    The signal switch parameter that will likely be most interest to you will be the voltage range of your signal and the Ron flatness because you would like to have the same resistance through the switch as the input signal changes.  Unfortunately, there isn't a way to sort on  Ron flatness on Ti.com but you can sort for devices with lower on state resistance which will also give you Ron flatness. 

    I don't know what voltage range you would like to support so here are some different options TS5A3159, TS12A12511, TS5A3359, or SN74CB3Q3251. 

    Let us know how else we can help with your selection.

    Adam

  • 0 in reply to Adam Torma
    Expert 1065 points
    Hello Adam :)

    So here's the short description of my project and my needs.

    My power supply range is symetrical +/- 18V for opamps.
    I'm going to use switch to achieve gain of 2 and gain of 4 with typical audio input section (consumer audio).
    I don't need anything more, this is why I don't use any PGA chip to control volume/gain transition.
    My volume control is a liitle modified TIDU034 design and I want to add gain transition to U1A stage.
    And input stage is non-inverting one
    ATM I've got Rf 1k and Rg 1k.

    If you need to know anything more just feel free to ask :)
  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points

    Mateusz,

    The +/-18V supply requirement will quickly filter the number of options in the current TI switch portfolio. What do you think about the new TMUX6104? 

    Thank you,

    Adam

  • 0 in reply to Adam Torma
    Expert 1065 points
    That chip look promising :)
    Buf I've got some other question.

    If I will use another regulator (5V like TPS7A91 or TPS7A39 for bipolar power supply with lower voltage than main).
    Will it change anything?
    Typical dac output is in range of 2-3VRMS (sometimes even lower).
  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points

    Mateusz,

    If you can power the switch with a different supply rail you will be able to have more switches available to you.  You can filter for different power supply combinations for switches on the ti.com switch parametric search tool.

    Thank you,

    Adam

  • 0 in reply to Adam Torma
    Expert 1065 points
    I saw many switches for audio are working from 5V.
    But will it work fine? Im afraid that lower voltage switch will introduce some problems?

    Would you be so kindly and tell me which things should I avoid and how to build everything properly?

    I'm looking for best possible solution to provide gain in my design.
    I can afford additional regulator for switching IC or even lower voltage bipolar supply.

    If you need more information about my design, just feel free to ask :)
    I just want to avoid relays because they're unpredictible, same for mechanical DPDT swtich (long feedback traces etc.).
  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points
    Mateusz,

    All the switches on that table are passive FET switches and can be used for any signal even if they say they are for audio, I2C, SPI, LAN, etc.

    What is your concern with lower voltages? If this is a concern for you then you can use the TMUX6104.

    You should avoid passing signals through signal switches that are outside the recommended operating conditions of the device. As mentioned above, you should also avoid devices with high Ron flatness because you would not want the resistance in your feedback network to vary across the voltage range.

    Thank you,
    Adam
  • 0 in reply to Adam Torma
    Expert 1065 points
    This answer is really helpful :)
    How about other things?

    Should I care about off isolation, crosstalk?
    I want to achieve pop-less transition between both positions.
    So should I care about transition time?

    Which chip is best for my app? I will use only two setting anyway (easier control with with mechanical switch without logic IC).
    Excluding voltage, because I can use virtually any voltage to power up those switch via LDO, zener diode etc.

    Should I care about THD of those ICs? Or NFB will take care of it?
    I saw so many switches from other companies which was designed for audio with negative signal handling capability.
  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points

    Mateusz,

    Are you amplifying an audio signal?  If you are, you will need to consider audio specs like THD, crosstalk, isolation, etc.  Please think about using the TS5A2236X devices.  

    Thank you,

    Adam

  • 0 in reply to Adam Torma
    Expert 1065 points

    TS5A22364 look very promissing :)

    And yes, Im going to use it in my audio amplifier input stage to manipulate gain.
    Maybe second chip will be used to manipulate Rsource.

    According to datasheet spec, I saw that ON state will add very low resistance to signal path, matching is very good too.

    So I just want to ask, how about switching application using only Rgain switch (Rfeedback will be fixed), I read that application will be less precise than those from 1st post.
    But after all if I take typical resistor tolerance and deviation from 1K value (my Rfeedback) it's clear for me that nothing bad should happen.
    Worst case scenario will add 1.8 ohm to signal path + missmatch between channel will be aprox. 0.15 ohm.

    1K 1% resistor tolerance will be in range of 10 ohm (so it could be 1010 ohm or 990 ohm).
    So adding another ~ 2ohm will not change anything.

    Even when using 0.5//0.25% tolerance resistor everything should be fine :)
    Taking typical performance values, it will be still less than 1 ohm, which will be less than 0.1% deviation from declared value.
    Even with very high gain level (with fixed Rfeedback) like 5 or 10 it still not an issue.

    Am I right about it? Or should I stay with circuit from 1st post?
    I think that in my circuit both will give good performance, because I'm not looking for high gain or ultra low values resistor for feedback path.

    Big advantage of those MUX family is that I've got to IC in one package with independent control.
    Superb choice for audio (stereo).


  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points

    Mateusz,

    I don't understand the circuit you are referencing of Rgain switch but your reasoning appears logical in the above post.  I'm glad you liked the TS5A226X devices since they are optimized for audio applications. 

    Adam

  • 0 in reply to Adam Torma
    Expert 1065 points
    I was refering to this site :

    www.ecircuitcenter.com/.../op_pga.htm

    They described two PGA design aproaches.
    I'm still thinking which one should I pick for my audio device.

    Could you help me with that?
    1st approach will give me more flexibility in term of gain setting, but 2nd type is more often recommended due to IC switch imperfections like Ron, Ron Flat etc.
  • 0 in reply to Mateusz Kaluzny17
    TI__Mastermind 45440 points
    Mateusz,

    I think you are well versed in the trade offs in the different approaches. You would also consider the accuracy needed in your design as resistor tolerances are another consideration. I can show the trade offs that exist with different designs using signal switches but the design decisions are ultimately made by our customers.

    Let me know if you have additional questions.

    Thank you,
    Adam