• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 45 subscribers
  • Views 1661 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.

Offset Errors and Resolution for Delta-Sigma ADC (16-bit ADS1110)

Leon Rasberry
Prodigy 30 points
Other Parts Discussed in Thread: ADS1110, ADS1115, ADS1114

I have several questions on the TI ADS1110 ADC:

(1) I am trying to determine if negative offset readings for a single channel are typical for the ADS1110 ADC device.

The data in Decimal for 5 channels (starting with 1 in our numbering) is given below:

Chan1 Chan2 Chan3 Chan4 Chan5

14968 00015 00036 -00021 00015

Chan1 is loaded with 200kOhm resistor. All other channels are Open. All Open channels seem to have a positive offset except channel 4. Are negative offset possible?  The signal conditioning circuitry is configured such that GND (0V) should be the minimum and all other voltages should be positive. Is this typical for this ADC, or should I be looking for leakage problems in the surrounding circuitry?

(2) Am I reading the ADC spec correctly? Is the ADC Resolution reduced from 16-bits to 12-bits for 240 Samples Per Second (or 240 SPS in the document)? If the Resolutions is reduced to 12-bit, is there a comparable ADC device which is 16-bit in the 250 SPS to 1000 SPS range? Preferably this would be a device which is still relatively low power consumption.

  • 0
    TI__Guru** 113765 points

    Leon,

    First of all it is never a good idea to leave inputs floating.  I assume that is what you mean by open.  To actually check offset you need to short the input pins together, and then analyze the output code.  Again, you don't want to leave the inputs floating but tie them to ground, or some voltage up to 2.048V.

    Is it possible to have a negative offset?  Yes, the inputs to the ADS1110 are bipolar.  This means that the output code is based on the + input relative to the - input.

    And yes, the resolution does drop for the higher data rates.  You might want to take a look at the ADS1115 which is a four channel device, and/or the ADS1114 which is a single channel device.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 30 points

    I had to use Two's Complement calculations to determine that there was a Negative Offset Error ( I was told this was not valid).    How should I be adjusting the ADC counts (or codes) for negative offset error versus positive offset error?  i want to adjust ( or Zero) each of the ADC's channels readings by Adding counts for negative offset error and Subtracting counts for positive. 

  • 0 in reply to Leon Rasberry
    TI__Guru** 113765 points

    Leon,

    The output code is binary two's complement.  See Output Code Calculation on page 6 of the datasheet.  Also view Table 2 on page 7.  At 240sps, the offset within the device should be minimal (only a few codes).  Obviously at 12 bits this value would be smaller by virtue of one LSB, but if you are actually measuring at 16 bits the offset is the same, but at this point you might want to adjust by subtracting/adding the offset.  For example, at 12 bits you may only see 1/2 code of offset, where at 16 bits you might see maybe 20 codes.

    So, if you do desire to adjust for the offset then you add counts for negative and subtract counts for positive offset.  This is assuming that you have measured the offset with the inputs shorted together.

    Perhaps you have some confusion as you are measuring single ended.  If one of your inputs is grounded, then you apply voltages in the range of ground to +2.048V, the output code may be positive or negative depending on which input is grounded.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 30 points

    All the channels should be connected the same, and only channel 4 has a negative offset error.  Therefore, the others working with me insist that all channels should have positive offset error.   However, I saw on page 3 of the datasheet in the ELECTRICAL CHARACTERISTICS table that there is component of Offset Error called "Integral Nonlinearity," which is intrinsic to the Delta-Sigma conversion process that can be +/- in value.  I believe this is what accounts for a significant part of the negative offset error.