ADS131A04: Unused Analog Input Wiring

Hello, I wanted to follow up with one more quick question regarding voltage follower buffering for the analog inputs. Should I increase the output resistance and capacitance of the voltage follower buffers (OPA4192) if my analog signal has a maximum bandwidth of 50Hz?

Hello Dale,

Thank you for clarifying that and my apologies, I should have looked further to see that the AVSS pin was already tied to ground!

I have attempted to follow the FAQ guide link and here are my results/approach thus far:

• Fmod = 83.591kHz. Using a 16.384MHz XTAL and setting CLK_DIV and ICLK_DIV to the max values (divide the xtal frequency by 14 twice).
• I'm setting the OSR to 4096, which creates a sample rate of 20.408Hz.
• Fc-diff = 835.92 Hz (100 times less than Fmod). This is higher than the output sample rate, so it seems to be a good goal based on the guide.
• Vov = 2.5V, Vesd = 5.3V, and Imax=10mA. With an added 10% tolerance to the voltage and 30% tolerance to the current. I think Rfilter must be greater than 440 ohms. 442 Ohms is the next standard size up I see on Mouser so Rfilter = 442 ohms.
• Cdiff = 1/(2*pi*835.92Hz*(2*442ohms)) = 215 nanofarads.
• Ccm = Cdiff / 10 = 21.5nF

For my signal GREEN_BUF, it is the reference voltage for the three different differential measurements, which makes me wonder if I should augment the resistor and capacitor values for that one signal in particular? The first thing that comes to mind for that signal is possibly dividing it's Rfilter resistance by three and multiplying it's Ccm capacitance by 3? Here is the updated schematic to include the anti-aliasing filter recommended in the FAQ:

Hi Tanner,

You updated design looks good to me. f_C-DIFF is recommended 10x to 100x lower compared to  f_MOD, you can select a slightly higher f_C-DIFF frequency to have a small capacitor value. Large capacitive load can lead to an oscillation on your amplifiers.

Is your TMP2 (GREEN_BUF's input) a DC or real temperature signal?

BR,

Dale

Hi Dale,

Awesome! f_C-DIFF was chosen through f_MOD being 83.591kHz after diving the XTAL 16.384MHz by 14 twice. I wonder if I should modify my approach due to the increased differential input impedance with this somewhat slower modulator frequency when compared to the usual 4.096MHz specified in the datasheet?

Yellow and TMP2 are both DC temperature measurements produced by off board LM20BIM7 sensors. Sometimes TMP2 may be left unconnected depending on the sensor the ADC is interfacing with, however the "Yellow" LM20BIM7 temperature signal will always be present. Should I add a jumper to ground the TMP2 signal when it is not present or would it be okay to just leave it floating when it is not present?

Edit: I just realized that I had the TMP2 and green inputs swapped around. Here is the fixed schematic:

Hi Tanner,

No, that should not be a concern for you, you do not need to match with 4.096MHz modulator frequency.

My question was actually about GREEN_BUF because you used this signal for all negative input pins AIN1N/AIN2N/AIN3N. I did not quite understand about this design. For example, you are measuring the difference between YELLOW_BUF and GREEN_BUF on the first channel, is GREEN_BUF a DC reference signal?

If TMP2 signal is not present or floating, the differential voltage between TEMP2_BUF and GREEN_BUF can vary a lot, but I guess you do not need to measure anything on channel 2 when TMP2 is floating, so it is not a problem.

BR,

Dale

Hello Dale,

Thank you for explaining that.

You are correct, GREEN_BUF is the buffered output from an offboard DC reference signal produced by an external sensor (GREEN). They recommend buffering the signals before the ADC in this attached diagram.

• Rate signal (off board signal) = White signal (in my schematic)
• Ref = GREEN
• TMP1 = YELLOW
• TMP2 = TMP2

Hello Dale,

That makes sense and will do. Thank you for the support and I look forward to testing this ADC!

Thanks,
Tanner