• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 46 subscribers
  • Views 576 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.

ADS1278 - noise and failure at low temperature

Jean-Philippe VERSCHUEREN
Intellectual 310 points
Other Parts Discussed in Thread: ADS1278

Hello,

We are currently testing an acquisition system equipped with ADS1278

At room temperature some channels of some boards exhibit a higher noise level (spectrum density increasing linearly with frequency).

When decreasing the temperature, the noise level is increasing drastically starting at about -5°C.

When decreasing the temperature further, the output of the ADC1278 simply disappears and all channels are then read as zeroes.

When increasing the temperature, the situation is back to normal and when increasing further above 30°C, the noise dissapears on all channels.

I have checked all the power supplies and timings and those seem in line with the datasheet requirements.

Operating conditions:

Analog supply: 5V

Digital Core: 1.8V

I/O supply: 3.3V

Frame Sync, TDM data outputs, fixed position TDM, normal operating mode

CLK = SCLK = 640 kHz, frame period 2560 tCLK, low speed, 250 Hz data output rate, 3 ADS1278 in daisy chain (i.e. 24 channels total).

Mode[1:0] = [1,1], CLKDIV = 1

I also did a test with a cold spray to cool down the ADS1278 only; just to make sure the problem is really located on the ADC and not on other IC.

  • 0
    TI__Guru 57336 points

    Hello Jean-Philippe,

    Thanks for your post.

    When you cool the system below -5°C, are the device power and CLK signals still valid?

    How are you heating and cooling the system? You mentioned using the cold spray on the IC, but what devices are nearby on the PCB and does this technique produce the same results?

    It would be helpful to see the noise data in a table / graph over temperature. Also, how are the inputs configured? The noise specifications in the datasheet are measured with the inputs shorted and biased to mid-supply.

    Best Regards,

  • 0 in reply to Ryan Andrews
    TI__Guru 57336 points

    I saw in your other post that you mentioned using a thermal chamber and that you no longer see anything on DOUT < -10°C.

    I will bring this up with the designers. Meanwhile, please provide any details about your setup and noise data that you can share.

    Thanks,

  • 0 in reply to Ryan Andrews
    Intellectual 310 points

    I will make some snapshots of the noise vs. time and frequency and post them.

    In the meantime:

    We have about 15 boards assembled with 3 ADS1278 on each of them. All of them exhibit the problem but at slighly different temperature levels (statistical dispersion) and not simultaneously on all channels. Some ADC1278 already exhibit a noise problem on some channels at room temperature; statistically more often on their first channel.

    The ADS1278 on the edge of the board; i.e. the one away from the voltage regulators where the board is warmer is affected first, then the second one, then the third one close to the warmest part of the board (about 10°C above ambient as measured with a thermal camera).

    But the behavior is always the following:

    When you cool down noise appears, increases and finally DOUT remains at low level.

    When going back to room temperature, DOUT is OK again with high noise level on the readings, then noise decreases while temperature rises.

    On the channels that already exhibit noise at room temperature, increasing the temperature further (i.e. up to 30°C and above) solves the noise issue.

    I was first thinking at power voltages, CLK, or FSYNC issues as testing was done in a thermal chamber on the entire board, but everything seems in line with the datasheet requirements. I then decided to purchase a cold spray to be able to cool down locally. With the narrow nozzle I am able to cool down the ADS1278 only, i.e. without affecting surrounding components. Even with a light cool air flow, the effect is immediate (noise increase on the readings, then lost of DOUT).

    I have asked for ADS1278EP samples from you (my guess is that they would come from another manufacturing batch). They should get here soon. I also had to procure a rework station to be able to remove the ADS1278 from the board because of the thermal pad underneath. If we succeed in replacing the chip it should give us more clues.

  • 0 in reply to Jean-Philippe VERSCHUEREN
    Intellectual 310 points

    We finally found the problem. It is due to a printed circuit board hidden problem. Some of the ground connections to the ADS1278 were missing.

    One device is now operating perfectly at -40°C.

    Thanks for your support.

  • 0 in reply to Jean-Philippe VERSCHUEREN
    TI__Guru 57336 points

    Hi Jean-Philippe,

    I'm glad you found the source of the problem. :)

    Best of luck,