enter image description here

I am using an ADS1115 as an analog to digital converter.

R2 and C2 are used as a low pass filter. I have used two schottky diodes as voltage clippers to protect against over voltage. To avoid floating condition I have put a pull down resistor of 10K(R6.) I have modelled a resistive voltage sensor as a voltage divider. I have used a 100K resistor to model the input impedance of the ADC .

The issue is because of the pull down resistor I am getting wrong analog voltage readings.

Any suggestions for modifications in circuit so that I can read proper analog values?

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$ – Voltage Spike Aug 21 at 15:47

R6 is too low an input impedance as seen by the divider's output impedance and combines with R5 to distort the divider ratio. R2/C2 also distort the divider ratio for the same reason at frequencies where C2's impedance is low. This results in a filter with a time constant that changes as the the divider ratio changes since R2/R4/R5/R6 all combine to make the R in the RC filter.

Add a buffer (opamp voltage follower) to isolate the RC filter from the divider so the time constant stays the same. Then put the protection circuit on the buffer input to protect the buffer and ADC behind it.

The buffer also lets you add a low pull resistor to the output of the buffer without distorting the divider ratio but this won't stop an open input from floating since a buffer still drives and overpowers the pulldown even if its input is floating.

Therefore, to prevent floating inputs, add a pulldown resistor to the buffer input that is large relative to the divider resistance (at least 100x larger than your total divider resistance R4+R5) so it does not distort the divider ratio by much.

| improve this answer | |

The ADC1115 is not 100K input but depends on FSR voltage and is likely closer to 10M.

  • sampling rate is low so you need a Nyquist noise filter not what you have, for 16bit noise free and use the differential input with balanced shielded STP cable.

  • remove 10k load and insert 10k series to diodes which may be redundant already inside but useful for ESD/EOS enhancement.

  • reconsider your noise environment and go differential if not sure.

Nyquist filter would block frequencies at 1/2 sampling rate down to resolution of ADC from interference glitches. So choose your worst case noise environment specs and then filter as needed.

If no-connect noise still too high add RF cap across input with RC=sample time interval. If worried about offset replace 10k shunt with 1M


Replace :

  • R6 with 1M.
  • R2 with 10k.
  • Change R7 to ≈ 10M.
  • consider Diff. Input.
  • Add RF cap to input ~ RC = 1/SPS rate. Or T >=3/SPS if high random noise.
  • Estimate worst case SNR(f) and include Nyquest Filter as required. Rather than C2

Next time include electrical noise ambient specs, SNR, V+ and more specs before asking again

| improve this answer | |
  • 1
    \$\begingroup\$ thanks for help tony stewart \$\endgroup\$ – power machines Aug 22 at 4:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.