• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 1428 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Will Connecting a SIngle-Ended Device into a Differential PGA be beneficial?

Roger Dallas
Prodigy 160 points
Other Parts Discussed in Thread: TINA-TI

Hello,

I'm hoping someone could clarify this for me - and I apologize that there are some basics missing on my part.

I have a single-ended sensor - let's call it an antenna. I'm pickup up random noise in addition to the signals of interest. Yes, I realize DSP processing etc, can help, but I want the design to be good from the input and forward.

The question: can connecting a device that's inherently a single-ended device, where one end currently connects to ground, be this connected in differential to help reject spurious noise signals that enter that sensor? I am hoping so.

For this task, I am looking at one of the INA series TI amplifiers. However, there are 2 constraints. One, the power supply must be single, +/-3.3 to 5V, and the other is, the sensor requires input tuning with a capacitor parallel with it, to ground.

Can someone provide some guidance on my questions?

Thanks so much!

  • 0
    TI__Mastermind 27715 points

    Hello Roger,

    Differential transmission can be hugely advantageous for noisy signals, if the noise is common to both components of the signal. In that case, the differential receiver will reject the majority of the common-mode noise. However, if the noise is differential, it will be gained up by the receiving amplifier.

    Single-ended to differential signal conversion is most useful before noise sources are present, because if the noise is already part of the signal, it will get converted to differential with the desired signal and will therefore not be rejected down the chain. In that case, smart filtering might be the better approach to reduce the noise at the system input.

    The app note linked below describes converting single-ended to differential signals in single-supply systems (for video applications). Without knowing more about your application and its respective parameters, this could be a good starting point: http://www.ti.com/lit/pdf/slyt427

    We offer a wide variety of single-supply INAs for receiving the differential signal. You can browse them here. 

    What may be useful is to build up a test circuit in TINA-TI which includes your sensor and noise sources. You can then try the SE-to-Diff conversion using our device macromodels and see the results before spending any time or money building up a real evaluation circuit.

    Best regards,

    Ian Williams
    Linear Applications Engineer
    Precision Analog - Op Amps

  • 0 in reply to Ian Williams
    Prodigy 160 points

    Hello Ian,

    Thank you for a concise and clear response, appreciated.

    These frequencies are lower, just up to 1K, and involve NFMI (near-field magnetic induction) transmission and pickup. Short distances are involved, perhaps 40 feet. Presently, the antenna is single-ended grounded, and is subject to picking up any near-field radiated EMI, such as an engine starting, even if 50 ft away. The front end has significant gain, in the order of 85 to 90dB. I am taking advantage of good filtering methods - using a 15th order LPF with a Bessel-ish response. It helps significantly, but appears that when a noise burst is present in the frequency zone of one you are trying to detect it gets in to the antenna, and makes it hard to differentiate between noise and signal, even with DSP. When binning energy values, the energy doesn't seem to care where it cam from. So even though the present system works well, noise is the issue.

    I believe I understand your comment on noise present at the source. I think that with an sensor such as an antenna, would this always be the case? (or, maybe not if the radiated noise is further away) So in essence you are saying that in my case, using any type of differential input wouldn't really help? I was thinking that a noise burst, even though at the same frequency, wouldn't be "in phase" like the signal I am trying to detect, making differential detection more likely to work.

    I realize this is wandering off the track a bit, but would differential ADC (with or without differential input) be of benefit here under the same conditions?

    What are your thoughts on this?

    Thanks again Ian -


    NB - I see your profile says you're into audio, that's great. I'm also an "Audioholic" and enjoy "eargasms" whenever possible, on my home-brew vacuum tube and FET amplifier system.

  • 0 in reply to Roger Dallas
    TI__Mastermind 27715 points

    Hi Roger,

    Unfortunately, if the noise is coupled into your sensor output and is within your signal bandwidth, there's not much that I can think of which can be done to eliminate it. Converting to differential signaling would be beneficial for noise which couples in after that conversion stage, but not before.

    Is it possible to place the sensor in some kind of enclosure or shield that would block noise while still allowing desired signal? Something of that nature which prevents the noise coupling in the first place might be your best bet.

    Best regards,

    Ian Williams

  • 0 in reply to Ian Williams
    Prodigy 160 points

    What you are saying makes perfect sense. If the noise is present at the source, whether the medium is air or wires,

    it's part of the signal and easy to see that differential won't do much for that. To answer your question, since the signal is (monopole) magnetic near-field, it's not possible to shield the detector effectively in terms of physical constraints.

    Thanks for taking the time to answer my questions!

  • 0 in reply to Roger Dallas
    TI__Mastermind 27715 points

    Hello Roger,

    You're very welcome, sorry I wasn't able to provide more assistance. Best of luck to you as move forward and overcome your design challenges!

    Best regards,

    Ian Williams