Micro Switch Testing

[pjr1121]

This circuit, repeated in parallel for 20 switches is prone to some problems.  Since the resistor values vary, the load resistance of your circuit varies as well.  Your supplied current to each of the 20 circuits will vary based on this load resistance.  A current source/sink will force the current flow through the switch to be more stable.  This will prevent issues such as 1 switch dying sooner since it has 3.5A while another is only getting 2.8, and another 2.7 (these variations are probably exaggerated).  If current is to flow through 10 switches at a time from a single source, you should ensure that they are all the same to ensure the vailidity of your test results.  The problem with this, however, is you will no longer be able to detect changes in the switch resistance since you will be using a Current Source instead of a Voltage Source (Current is ALWAYS 3A)

Also, make sure that the resistor you are measuring, R3, is 2-3 orders of magnitude less than the imput impeedence of the DAQ to prevent parasitic effects from the DAQ itself.

[johnsold]

If you take one digital sample just before the switch changes state or somewhere in the middle of the stable parts of the cam rotation, you should be able to avoid the bounce issues and only have two data points per cycle for each switch.  The voltage divider method should work fine for that method. As one of the other responders mentioned, you need to choose the divider resistors to be compatible with the input requirements of the DAQ device.

If the accuracy of the current drawn by the resistors (when you factor in tolerance and temperature effects) is good enough you do not need the extra complexity and cost of an electronic current regulator. For a life test I would think resistors would be adequate.

Lynn