calibration

Discussion in 'General Electronics Chat' started by Robert123, Apr 17, 2010.

  1. Robert123

    Thread Starter New Member

    Feb 12, 2010
    29
    0
    I have a friend that always ask me (with joke)
    Why always we need to calibrate every electronic device ?
    If anybody here has an answer for this I will glad to know (The answer can be like a joke but I want to know also the real answer)
    Thanks
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    To ensure that the readings given by the instrument(s) are reasonably accurate.

    The further you get away from the standards, the more error you will have introduced into the readings.

    Electronic devices "drift" over time due to a number of reasons. If it is desired that they be very accurate, they must be re-calibrated at periodic intervals to ensure continued accuracy. The instruments/devices used for calibrating other instruments are very expensive, and must themselves be periodically calibrated against higher-level standards.
     
  3. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Because electronic devices drift. For most home electronics it doesn't matter, but if the use is a life support or somesuch, it becomes critical. A lot of devices, such as resistors, have gotten much better. I've seen older styles drift over 100% of their value with normal use.

    I worked a year as a metrology tech, mostly with oscopes, frequency synthesizers, and optical devices. Most cases it was just a matter of keeping the circuits in balance, tweaked to best performance. With something like DVMs (especially cheap ones) there isn't much you can do. If they are far out at all you throw them away.

    Like I said though, for home and many commercial uses it doesn't matter.

    ************

    Ya beat me this time Wookie!
     
  4. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    If you think about the effect temperature can have on electronics, youll see the need.

    If you were using a 12v battery as the power supply for a scale, the accuracy is dependent upon the electronics and mechanical devices know what the other is doing.

    Heat can cause a 12v battery to change output. Heat can also cause the metal in a scales sensor to expand. This expansion and power increase can cause the scale to read an incorrect value. So calibrate will say to the processor, "Whatever voltage and mechanical stresses are being acted on us now will be known as zero."

    The flexed steel top of a scale can push against the load sensor causing the scale to read 5g when there is nothing on the scale. But if it were twenty degrees cooler, the steel wouldn't flex as far so the sensor would see a different reading as zero.
     
  5. Robert123

    Thread Starter New Member

    Feb 12, 2010
    29
    0
    Can you give examples of open loop control that need or don't need to calibrate ?
    **************************************
    Ya beat me this time Bill Marsden, retched and Wookie!
     
  6. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Using a pot on the neg feedback loop of an opamp can be used to calibrate (or set) the gain of the amp to adjust for environmental variables.
     
  7. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    If you cut off some random stick called it a metre and used it to measure something with, that would be OK and could be very consistent.

    If you now sent those measurements to me they would be useless as my metre stick would be different from yours.

    There is no magic in the length of a metre stick, just so long as everyone use the same.

    Volts, amps etc are just as arbitrary.

    The process of everyone using the same is called standardisation and requires calibration of metre sticks or voltmeters.
     
    Last edited: Apr 18, 2010
  8. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    To add to what Studiot said, minor errors accumulate. Electronics used math, very closely. If an amp is supposed to X10, and is X10.1 it may make enough difference to kill something (or just throw it way off) down the line. If you have 4 amps in a row, each at X10.1, the final result would be X10.4K instead of X10.0K.

    One of the many lessons students learn is parts tolerance. A typical resistor is rated for ± 5%. Statistically the errors average out, but also statistically sometimes the errors can stack up in a bad sort of way. Calibration brings everything back into control. If a pot is added the amp mentioned by way of example can be tweaked back to X10 ± 0.0001%, if that was the precision needed.

    Even modern parts still drift. If not the resistors then the capacitors, if not the capacitors then the solid state components themselves. Things change.
     
Loading...