Decoupler test tool

[Murphyv310]

Bigger values tend to increase in value and I have seen motor run caps go wildly up in value, OK as cap droppers though as they are under less stress.

Any idea for a mechanism for capacitance rising? The usual reason for apparent rise in C is a side effect of instruments that use charge/discharge to measure C. Such as the Peak ones and anything else that isn't a bridge. These will interpret leakage as increased C.

Possible theories are that the "plates" of the capacitor when they corrode have a larger surface area, also any moisture held in a paper dielectric gives the effect of it being thinner.
If you add a parallel resistor to cheap testers when measuring a cap it doesn't alter the C value significantly.

Another test can be in frame stages. Often frame speed is well out as can linearity, if you replace them for the correct value the fault is repaired but try measuring the cap and substitute it with a new component of the value and lo and behold you are often still faced with the same fault. Frame stages are a very good example of requiring the correct value of components so show up this phenomenon very well.