E21 Lecture #13

Project Planning

Assignment for this week:

More on measurement

Direct measurement

Indirect measurement

Null Measurement

Measurement quality

Measurement Errors

(Different from measurement noise)

Solutions

  • Use high-quality instruments and good designs
  • Use the instrument that least disturbs what is being measured
  • Use several different instruments to measure the quantity
  • Use multiple measurements and average the results
  • Will remove zero-mean Gaussian noise
  • Will not remove salt & pepper noise
  • Use multiple measurements and take the median value
  • Will remove salt & pepper noise

Error Analysis

  • Identify all contributing sources of noise in the system
  • Quantify the noise sources
  • Do a worst-case analysis where you assume all of the noise sources are biased in the same direction
  • Look at both the positive and negative cases
  • Compare the resulting error to the signal to determine its effect
  • If all of the errors are random and of the same order of magnitude, then you can approximate the size of the error as the square root of the sum of the squared errors

Quantifying error in a measurement

  • Take N measurements and find their average
  • Calculate the standard deviation of your population (from your measurements)
  • Calculate the "standard error of the mean" by dividing the standard deviation by the square root of N
  • Report your measurement as the mean +/- the standard error of the mean

Sensing with a computer

External stimulus -> sensor/transducer -> amplifier -> analog processing -> A/D converter

-> computer port -> software processing -> sensor output/decision-making

Transducer: converts one form of energy (information) to another (usually electrical)

Amplifier circuit:

Analog processing

A/D converter

Computer Port

Software procesing

Decision-making