When measuring AC rms volt of TMP36 temperature sensor by TDS Tektronix 2002C digital oscilloscope, I got value of 5.97mV.What's the absolute error in my measurement (absolute error from oscilloscope )? I went through the manual but could not figure it out. It would be a great help.enter image description here]

  • 1
    \$\begingroup\$ looks like a signal to noise ratio<1 , mostly >=5MHz .. usena differential ferrite CM choke around twisted pairs or shield it and ensure scope gnd leads are very short. Try A-B mode on cal test signal to get a flat line on 1mV/div then measure across a load R or add 100pF to filter RF \$\endgroup\$ – Sunnyskyguy EE75 Nov 17 '16 at 23:09
  • \$\begingroup\$ Thanks Tony, I was trying to find sensor noise of TMP36 sensor. I isolated the temperature sensor by putting it in a plastic shield (wrapper) and placed in normal water for 3/4 minutes. TMP36 gives 10mV/degree Celsius, so I put 1mV for the vertical scale in the Oscilloscope. I was just trying to find 5.97 +/- some quantity that is the error in reading from the oscilloscope \$\endgroup\$ – Dinesh Nov 18 '16 at 15:04
  • 1
    \$\begingroup\$ there error is not in the scope but your filter, shield design or probe method \$\endgroup\$ – Sunnyskyguy EE75 Nov 18 '16 at 15:29
  • \$\begingroup\$ Add decoupling caps , use twisted pair, add low pass filter 100Hz to get noise down to 10 uV \$\endgroup\$ – Sunnyskyguy EE75 Nov 18 '16 at 15:30
  • \$\begingroup\$ Thanks so much Tony, I appreciate it a lot.. I am using analogRead(pin) of Arduino to read the temperature from TMP36, analogRead(pin) takes about 100 microseconds (0.0001 s) to read an analog input, so the maximum reading rate is about 10,000 times a second (upto 10 kHz). Is there anyway from where I can find the operating frequency range of TMP36 from its datasheet? I want to know it because I want to to check if I can use other sensors that can read values with the analogRead(pin) of arduino. I believe arduino's analogRead works only for certain frequency range of the sensors... \$\endgroup\$ – Dinesh Nov 18 '16 at 16:17

Consider temperature as a DC voltage that ramps slowly.

  • add a low pass RC filter such as 100k to 1uF near ADC input, so the fastest change full scale possible is 1V/100ms or 1mV/100us
  • consider averaging 100 readings to reduce noise / sq.rt.(100)= 1/10th noise. There is no signal frequency from sensor, only DC at fixed temp.
  • you are oversampling temperature at 1000x rate necessary. you could accumulate N results and divide by N to reduce jitter by \$\ N^{\frac{1}{2}}\$ and/or add RC

This should give you much better results for TMP36 with 10mV per 'C with 0.5V offset at 0'C

  • \$\begingroup\$ Thanks Tony..For my last question, will I be able to measure any sensors using the analogRead (pin) of Arduino or does the frequency range of the sensor also play some role whether it could be read or not by the arduino anaogRead function \$\endgroup\$ – Dinesh Nov 18 '16 at 18:12
  • \$\begingroup\$ I need to see detailed schematic to be sure , but sensor bandwidth of >100kHz is for rapid turn on off to save uA power, if not doing this a LPF reduces noise \$\endgroup\$ – Sunnyskyguy EE75 Nov 18 '16 at 20:25
  • \$\begingroup\$ rms measure (ac coupling mode) gives sensor noise experimentally, is there any theoretical explanation for it? \$\endgroup\$ – Dinesh Nov 18 '16 at 23:52
  • \$\begingroup\$ of course but aren't going to try to suppress the noise as I said. \$\endgroup\$ – Sunnyskyguy EE75 Nov 18 '16 at 23:54
  • \$\begingroup\$ Thanks Tony..but still I am looking for some theoretical explanation how rms (ac mode) represents sensor noise. \$\endgroup\$ – Dinesh Nov 18 '16 at 23:59

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.