• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 141 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.

ADS125H02: Input setup for single-ended and differential measurements

nameduser
Prodigy 30 points
Part Number: ADS125H02

Tool/software:

Hello everyone,

I'm struggling to understand the input setup of the ADS125H02 ADC. I want it to be usable for single-ended and differential measurements. Differential measurements should be done through AIN0 and AIN1 and single-ended between AINCOM and AIN0/1. Is a setup like this feasible/recommended or not? Specifically, I'm wondering about the components before the input. The PCB Layout Example (fig. 103) places a 10 nF cap between AIN0 and AIN1. What is the purpose of this capacitor? I would assume than when changing to single-ended measurements, this cap will cause coupling of both signals. And should AINCOM be tied to ground in this case or does it make sense to have it floating for any measurements (symmetrical supply, so that GND is in the middle would allow for positive and negative voltages)?

Furthermore, I have some questions unrelated to the title: The example design (ADS125H02EVM) places 47 Ohm resistors in the CS, start and reset signals whereas they are not included in the PCB Layout Example (fig. 103). Are these needed or are they implemented "just in case"? Additionally, fig 103 recommends to tie CLKIN to GND behind a 47 Ohm resistor to use the internal oscillator. Can't the connection to GND be made without the resistor?

Thank you very much!

  • 0
    TI__Mastermind 47091 points

    Hi nameduser,

    The ADS125H02 always measures differentially between the inputs selected via the MUX bits in the MODE4 register. Connecting AINCOM, AIN0, or AIN1 to ground would enable a single-ended measurement

    The 10nF cap between AIN1 and AIN0 is a differential filter capacitor, to help eliminate unwanted differential noise/signals. The cap on each pin to ground is the common-mode capacitor, that helps remove unwanted common-mode signals/noise. You can reference this FAQ for more information: https://e2e.ti.com/support/data-converters-group/data-converters/f/data-converters-forum/955466/faq-delta-sigma-adc-anti-aliasing-filter-component-selection

    The 47 ohm resistors are for impedance matching to reduce reflections. However the signal speeds are not that fast for the ADS125H02 so these are probably not critical. Of course you can include them as 0 ohm resistors on your design, and then populate them if necessary as you begin prototyping

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 30 points

    Thanks for your response, that answers some questions!

    But some still remain: The link you provided does show C_DIFF in combination with R_FILTER, creating a differential RC low pass filter. But I don't see a resistor in figure 103. I can't find the caps to ground in fig. 103 in the manual (but the ADC has an LPF to GND for the input as shown in fig 57 and the AA-filter is created by the external 1 nF if I understand the datasheet correctly). 

    I think one part of my initial question wasn't stated clearly enough. I want to expose AINCOM, AIN0 and AIN1 and leave it up to user if they want to measure a differential signal between AIN0 and AIN1 or up to two single-ended signals between AINCOM and AIN0/1. Does placing the C_DIFF between AIN0 and AIN1 influence the measurement performance when measuring single-ended inputs between AINCOM and AIN0 and AINCOM and AIN1? If it does, what is the better fix: Leave out the differential cap or allow only 1-channel single-ended measurements and leave either AIN0 or AIN1 floating (or would it be better to put the signal onto AIN0 and AIN1 at the same time)?

    And finally, must AINCOM be connected to GND for single-ended measurements or can/should it be left floating if the reference is provided externally? (I'd assume its better to tie it to GND because otherwise a voltage difference between the external reference and the GND might lead to a too high common-mode voltage, is this assumption correct?)

    I hope this clarifies my initial questions enough. If it doesn't, please let me know so I can attempt to explain it better.

    Thank you! 

  • 0 in reply to nameduser
    TI__Mastermind 47091 points

    Hi nameduser,

    nameduser said:
    The link you provided does show C_DIFF in combination with R_FILTER, creating a differential RC low pass filter. But I don't see a resistor in figure 103. I can't find the caps to ground in fig. 103 in the manual (but the ADC has an LPF to GND for the input as shown in fig 57 and the AA-filter is created by the external 1 nF if I understand the datasheet correctly). 

    The layout guidelines only show the differential cap because typically this should be placed as close to the input pins as possible. There is a little more leeway on the resistors + common-mode caps, which is why they are not shown specifically in the layout guidelines. You can see they are there in Fig 100 and Fig 98 for example

    nameduser said:
    I think one part of my initial question wasn't stated clearly enough. I want to expose AINCOM, AIN0 and AIN1 and leave it up to user if they want to measure a differential signal between AIN0 and AIN1 or up to two single-ended signals between AINCOM and AIN0/1. Does placing the C_DIFF between AIN0 and AIN1 influence the measurement performance when measuring single-ended inputs between AINCOM and AIN0 and AINCOM and AIN1? If it does, what is the better fix: Leave out the differential cap or allow only 1-channel single-ended measurements and leave either AIN0 or AIN1 floating (or would it be better to put the signal onto AIN0 and AIN1 at the same time)?

    I don't have a good answer to this question, as it likely depends on the signals you are measuring and the performance you expect. If you are measuring DC signals, then you can probably assume very little coupling between channels when the diff cap is installed. If your system is subject to higher frequency noise, then this might couple across the diff cap. But since this is an error case, and not specified operating conditions, maybe that is okay?

    My suggestion would be to include footprints for both caps, and then actually measure to see if there is a performance difference in your specific system

    You can also try this on our EVM, as shown below. At least you can directly connect AINCOM to ground by replacing C10 with a 0ohm resistor, then remove/install C3 to see if there is a performance difference

    nameduser said:
    And finally, must AINCOM be connected to GND for single-ended measurements or can/should it be left floating if the reference is provided externally? (I'd assume its better to tie it to GND because otherwise a voltage difference between the external reference and the GND might lead to a too high common-mode voltage, is this assumption correct?)

    I don't understand this question. As I mentioned, the ADC is always measuring differentially between the various options in the MUX bits in the MODE4 register. What would you be measuring your single-ended signal with respect to if AINCOM is floating? And what does this have to do with the reference?

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 30 points

    Thanks, that's answering up a lot of my questions!

    To answer the last part: Maybe if have a false understanding of single-ended measurements. I'd assume it's similar to a stereo headphone jack with 3 pins: Signal 1 (Left), Signal 2 (Right) and the zero-reference for both lines (GND; the shield of a cable). The difference to a differential input is that both signals share the reference (GND). So n+1 wires for n single ended measurements and 2n wires for differential measurement. Or does single-ended mean that the GND takes some random path (= is implied) and really only n wires are needed (and AINCOM is not an exposed port and only tied to internal GND)?

    And input RC filters are recommended for every application is what I infer from your reply, right? So for a general-purpose measurement, I should just copy the reference design/EVM cutoff frequencies.

  • +1 in reply to nameduser
    TI__Mastermind 47091 points

    Hi nameduser,

    nameduser said:
    To answer the last part: Maybe if have a false understanding of single-ended measurements. I'd assume it's similar to a stereo headphone jack with 3 pins: Signal 1 (Left), Signal 2 (Right) and the zero-reference for both lines (GND; the shield of a cable). The difference to a differential input is that both signals share the reference (GND). So n+1 wires for n single ended measurements and 2n wires for differential measurement. Or does single-ended mean that the GND takes some random path (= is implied) and really only n wires are needed (and AINCOM is not an exposed port and only tied to internal GND)?

    Single-ended is a measurement with respect to ground. That would be DGND in the case of the ADS125H02, since there are multiple ground / power nets. So you would need to ground AINCOM, and then measure AIN0 and AIN1 with respect to ground (AINCOM)

    The differential measurement is (AIN0 - AIN1), or (AIN1 - AIN0). But note for your single-ended measurements where the ADC is measuring with respect to ground (AINCOM), the ADC is still taking a differential measurement [(AIN0 - AINCOM) or (AIN1 - AINCOM)]. AINCOM could be any voltage, the ADC doesn't really care. It just so happens in your case that AINCOM = ground

    I was confusing the term "reference" as voltage reference, not ground reference. The ADC uses the voltage reference to determine the output code, and I assumed this was what you were referring to

    nameduser said:
    And input RC filters are recommended for every application is what I infer from your reply, right? So for a general-purpose measurement, I should just copy the reference design/EVM cutoff frequencies.

    The RC filters are there for anti-aliasing and input protection, as described in the FAQ I linked to previously. So I would use that information to make sure your design needs are met by the filter, and adjust if necessary.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 30 points

    Thank you so much for your fast and incredibly helpful support! I think that answered all of my questions.