• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 597 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.

Original question:

ADS131M08: What's the common mode input range of ADS131M08?

ADS131M08: ADS131M08

Paulo Pecegueiro
Prodigy 120 points
Part Number: ADS131M08
Other Parts Discussed in Thread: ADS131E08, , OPA330


Hi, I have hardware that reads straing gauges powered by 3.3V, uses an ADS131E08 with PGA = 12 and VREF = 2.4V.

I am reading voltages between (1.65V + 0.2V) to (0.2V - 1.65V) with full 24 bits resolution.

Range: 1.85V to 1.45V  (referenced to  1.65V).

 


I am planning to change the A / D converter to ADS131M08, as it is smaller and cheaper.

The strain gauges will be powered with 3V on the new hardware, so they will generate a voltage referenced to 1.5V.

My idea was to amplify the strain gauge signal to vary between (1.5V + 1.2V to 1.5V - 1.2V) and use all 24-bit resolution when PGA = 1.

The new input voltage range must be: 2.7V to 0.3V (referenced to 1.5V)

However after reading some questions on the forum, I was unsure if I could use it in the same way.

From what I understand in the forum topics, I could use it only to read voltages between 1.2V to -1.2V referenced to AGND (0V).

Range: 1.2V to -1.2V (reference to AGND (0V) )

So if I were to use ADS131M08 with single supply I would need to condition the signal to vary between 0V and 1.2V and the resolution would be only 23bits.

I would like to understand this better, the datasheet does not make it clear.

  • 0
    Prodigy 120 points

    ERRATA (Second line)

    I am reading voltages between (1.65V + 0.2V) to (1.65V - 0.2V) with full 24 bits resolution.

  • 0 in reply to Paulo Pecegueiro
    TI__Guru* 92655 points

    Paolo,


    The ADC is a differential input, so the ADC converts inputs based on the measurement between the analog inputs of AINP-AINN. These analog inputs must be within the valid input range for the ADC, where the absolute voltage for both inputs can be measured. The input range for the ADC can be best described with these lines from the Recommended Operating Conditions table in the datasheet (page 5).

    Based on this table, if the PGA gain is 1, both AINP and AINN must be between AGND-1.3V and AVDD. If you're using an AVDD of 3.3V, then the absolute voltages of AINP and AINN must be between -1.3V to 3.3V.

    Then, measuring the voltage must be within the range of the ADC in magnitude. For this measurement, the voltage of (AINP-AINN) must be between +VREF and -VREF. This is presumably +1.25V and -1.25V.

    As an example, set AINP=-0.1 and AINN=-1.3V, and VREF=1.25V. Both values are between -1.3V and 3.3V. Also AINP-AINN=1.2V, so this is a measurable value for the ADC.

    As a second example, set AINP=2.1V and AINN=3.3V, and VREF=1.25V. Both values are between -1.3V and 3.3V. Also AINP-AINN=-1.2V, so this is also a measurable value for the ADC.

    Going back to the first range in your original post, if your inputs are between 1.85V to 1.45V referenced to 1.65V, this is within the range of the ADC. Both inputs are between -1.3V and 3.3V. Your maximum signal would be AINP-AINN, or 200mV. You could even use a PGA gain of 4, and make a valid measurement. This way, you wouldn't need an external amplifier which could add offset and gain error. Note that gains higher than 4 limit the input range further.

    Hopefully this answers your input range questions. Out of curiosity, did you feel that one of my previous posts described the input range incorrectly? Can you post a link to that thread? 


    Joseph Wu

  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Here (https://e2e.ti.com/support/data-converters/f/73/t/887404?tisearch=e2e-sitesearch&keymatch=ADS131M08) you said: 

    "Kazuya-san,

    Yes, the ADS131M08 can be used for a single-ended input measurement with the negative input pin tied to ground. For this type of information, you would look at the Recommended Operating Conditions table on page 5 of the datasheet. Here is the section for analog inputs:

    This device has a negative charge pump that allows for input signals that go as low as 1.3V below ground. However, you can operate this device with the inputs going only from 0V to VREF. I would note that the ADCs are fully differential, so if the inputs are only positive values, then you would only get output data from 000000h to 7FFFFFh, which would be equivalent to 23 bits.

    Joseph Wu"

    I will use the ADS131M08 as a single-ended input, but in my case I need to read values from 0.3V to 2.7V.

    Using an OPA330 I will amplify a small signal of ± 6mV from a straing gauge bridge to 0.3V to 2.7V (referenced to 1.5V). This single-ended signal will be sent to the acquisition pcb via a 30cm shielded cable. On the acquisition board, the negatives input pins of ADS131M08 will be tied to 1.5V (So the ADS131M08 will read from 1.5V to 0.3V or 1.5V to 2.7V), I know this mechanism will generate a extra offset, however for my application it will not be a problem. I just want to certify me that I will get the 24bits. What do you think about this?

  • 0 in reply to Paulo Pecegueiro
    TI__Guru* 92655 points

    Paulo,


    For your configuration, with AINN connected to 1.5V and AINP going from 0.3V to 2.7V is what I would call pseudo-differential. You have a negative input that is stationary, but the positive input goes above and below the negative input, so that the ADC achieves both positive and negative values.

    For the other post, I was pointing out that if the AINN is set to ground and AINP only goes above ground, the output will only have positive values. With only half of the ADC range used, this is equivalent to 23 bits.

    With your setup as a pseudo-differential measurement, the ADC should span then entire range of 24 bits. However, I would note that you will have noise that will limit the resolution. Table 1 in the datasheet shows the noise performance of the device, and Table 2 shows the effective resolution that you would get with the noise.

    By adding in another op-amp, and setting it in gain, you add in some offset and some gain error. Another important consideration is that the op-amp adds in noise. If the op-amp is doing the amplifying, then the op-amp noise will be amplified as well. If your measurement is ±6mV and the input noise of the op-amp is 1.1uV (from 0.1 to 10Hz), then the noise is 1 part in 11000 or so. This is about 13 to 14 bits of effective resolution.


    Joseph Wu