I'm working on a system with a distance sensor. As the distance increase, the output voltage decrease. ( Fig. 2 Page 5 : http://www.farnell.com/datasheets/1657845.pdf?_ga=1.97138143.1773643066.1472611965)

Is it better to let the ADC "read" the raw voltage or should I amplify it when the output voltage is too low? (and consider the amplification in software)

Edit :

How is it possible to know the distance when the output is 1v? (Both 2cm and 27cm match with this voltage)

Any idea? Or should I avoid the scenario of a under 10cm distance object?

  • \$\begingroup\$ I think everything is about ADC's precision. If that precision isn't good enough, I should amplify it. Am I right? \$\endgroup\$ – M.Ferru Oct 12 '16 at 17:14

You have to tell us how the device is to be used before anyone can answer this meaningfully.

You've got 0.5V of change over the range of the device. This is a significant amount for a 10-bit A/D, so you wouldn't need to amplify if you were trying to tell the difference between the low end and the high end of the distance range.

Assuming the output is linear with distance, if you were trying to resolve 1mm of motion, the voltage change would be 0.5/700, or 0.7mV. This is a bit much to expect a 10/bit A/D to handle, and you would need amplification.

Assuming a 5V reference, 1 LSB on a 10bit A/d is a bit under 5mV. A 1cm change in your device would cause a change of about 7mV. So if you're trying to resolve 1cm, this is still pushing it. I'd recommend amplification of about a factor of 4 (more than that risks saturation). You'd need to remove offsets, obviously, to put your signal in the 0-5V range.

If you're trying to resolve 10cm, you probably don't need amplification

  • \$\begingroup\$ Yes, I'll use the 10-bits A/D from a MSP430. I will measure distance from 10cm to 80cm. But I don't really need a 1mm precision \$\endgroup\$ – M.Ferru Oct 12 '16 at 17:17
  • \$\begingroup\$ @M.Ferru -- answer modified to convey more info. \$\endgroup\$ – Scott Seidman Oct 12 '16 at 17:29
  • \$\begingroup\$ Thanks ! I'll amplify the signal if it's under 1V (25cm) for far distance. \$\endgroup\$ – M.Ferru Oct 12 '16 at 17:43
  • \$\begingroup\$ @M.Ferru You must experiment with device before design. Consider only large white paper or same. 90% ref. as the target and nothing else. Also expect OpAmp is driving this signal \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Oct 12 '16 at 17:45

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.