• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 709 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.

ADS1248 internal (ambient) temperature sensor - spikes in the data - 20sps

craig christensen
Prodigy 40 points
Other Parts Discussed in Thread: ADS1248

Hello all,

I am measuring the ADS1248 internal temperature sensor (ambient temperature monitor) and I am seeing a lot of noise spikes when using the 20 samples per second. I do not see the same problem when using the 160 samples per second range.

Any ideas?

  • 0
    TI__Guru* 92715 points
    Craig,


    Offhand, I can't think of anything that would cause the problem to be at 20SPS, but not at 160SPS. Usually the problem is the other way around where the problem is at 160SPS but not at 20SPS. In that case, it is usually 50 or 60Hz noise coupling into the device somewhere. At 20SPS, the digital filter does a better job in rejecting line noise.

    Regardless, I'd guess that it is some sort of EMI noise coupling in. I'd check to see if there is noise in the power supply or coupling in through the reference.

    What frequency is the spiking and does it relate to anything on your board? Can you post a long run of data with this measurement? It often helps to get the raw data and then try to figure out what external problems might be causing this disturbance. After that, perhaps you could post a schematic with the ADC?


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 40 points

    Joseph,


    Thank you for your reply. I am only seeing the spikes on the internal temperature measurement at 20sps. It does not show up on the external measurements, this was tested feeding in a 1mV signal. I ran an FFT on 2048 cycles and there was no dominant frequency. It looks like random noise. Maybe we are doing something in the processing. I don't know. Whatever it is it is specific to the internal temperature measurement.

  • 0 in reply to craig christensen
    TI__Guru* 92715 points
    Craig,


    In that case, how large is the standard deviation of the measurement? For the internal temperature sensor, it translates to a 118mV measurement at room temperature, and I'd be curious how much noise is seen from a voltage point of view.

    It's been a while since I've looked at the internal temperature sensor. If I have a chance, I'll use an EVM to see if I can duplicate your noise. I've got a couple of things going, so may not be able to look at it until tomorrow.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 40 points

    More information.

    4 - measurement runs.

    ADS1248 - 20SPS, internal temperature
    samples Average Std dev Max Min
      C C C C
    65 24.93663 0.140197 25.25571 24.28702
    63 24.94005 0.078059 25.04202 24.47599
    1024 25.04576 0.124154 25.58596 24.3294
    2048 25.12458 0.140412 25.63982 24.24817

  • 0 in reply to craig christensen
    TI__Guru* 92715 points

    Craig,

    I've taken a few measurements, and I don't see the noise that you are seeing. However I did want to point out a few things. When I first powered up the board to make some measurements, it did take quite a while to settle out the temperature of the board. In the end, I did have to wait a bit to get something that is close to a steady-state value. The result below shows the first 20SPS measurement. It's still settling. I accidentally took a 5SPS measurement before this, so I'd already had 4 minutes of settling.

    Also, a series of 1024 measurements at 20SPS takes almost a full minute to settle. This means that any transient effects, like a breeze from the air conditioner, or even movement over the board could change the measurement. In the end, I just put the board in a static bag to make sure I didn't have to worry about any air flow.

    I took measurements for 20SPS and 160SPS. The standard deviation runs at about 5 to 7 uV. You can see the plots below.

    After that I took data at 2kSPS and the noise got a bit worse. If I had to guess, it looks like it's about 30uV standard deviation in noise. At a change of 405uV/C, this is still less than 0.1C/ I believe this is still better than what you're getting which I think is closer to 50 or 60uV of noise in standard deviation. Regardless, the 2kSPS plot is below.

    What are you using as the reference? I'm using the internal reference, which should have a pretty good noise performance.

    Joseph Wu