Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
05/18/10 19:28
Read: times


 
#176008 - Drift is any change from the intial value
Responding to: ???'s previous message
We know temperature drift, drift caused by electric or magnetic fields, caused by mechanical tension, caused by pressure, caused by moisture, caused by radiation, etc.

We distinguish reversible and irreversible changes. A reversible change disappears when the parameters are set to the inital parameters. An irreversible change does not disappear, but tells us, that the component has changed its "shape", that it shows aging.

Aging can be tremendously accelerated by heat. Arrhenius had made some remarkable analysis on that.

A very special and highly undesired drift phenomen is hysteresis. Hysteresis makes, that a certain component value depends on what has happened to the component in the past. An example: You stand on a scale and see 77kg on the display. Standing again on the same scale might show a reading of 77,2kg.

Hysteresis is typical for pressure sensors. Some hysteresis will disappear after waiting some time, others will not, if the limit of elasticity was exceeded, for instance.

To keep an resistor as precise as possible, you must forbid all extremes. Limit the surface temperature of resistor to less than 40°C, avoid moisture, avoid temperature cycling and keep the ambient conditions stable. The soldering is critical either: Overheating and contaminating with agressive chemicals during and after the soldering will increase the drift. Also, avoid mechanical tension during the mounting.

But all this does not help, if the resistor isn't manufactured by using the most precise and stable materials and optimum manufacturing processes.

Kai Klaas



List of 16 messages in thread
TopicAuthorDate
Drifting in electronic components            01/01/70 00:00      
   100mV is 2%            01/01/70 00:00      
      Educative            01/01/70 00:00      
         Component modelling            01/01/70 00:00      
            Modelling            01/01/70 00:00      
               Quantitive modelling            01/01/70 00:00      
            For resistors?            01/01/70 00:00      
               I think you are getting the wrong idea            01/01/70 00:00      
               Similar but different            01/01/70 00:00      
                  Distinguish "drift" from short-term changes            01/01/70 00:00      
                     Trying not to be pedantic            01/01/70 00:00      
                        Drift is any change from the intial value            01/01/70 00:00      
   How actually measured?            01/01/70 00:00      
      An example only            01/01/70 00:00      
         No general answer...            01/01/70 00:00      
   Ratiometric techniques; Calibration            01/01/70 00:00      

Back to Subject List