PSC Motor cap loosing value due to voltage spikes 1

[jbgs1]

I was recently told by a capacitor manufacturer that my self healing polypropylene dry film capacitor was gradually loosing value due to high voltage spikes generated by the motor each time the motor is disconnected from line power (turned off). This is a fractional HP, 120 V. motor with a 5 Mfd cap. With an oscillascope connected to the cap lead, I have seen an occasional 1.5 kV spike that dissipates to 0 in about 2 microseconds when I switch the power off to the motor. Can anyone shead any light on what may be causing this voltage spike? Any help is appreciated.

 

[electricpete]

inductive kick? (v = L di/dt)

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[biggeorgetechnical]

I agree with electricpete, pretty clear.
(Though a "good" start/run capacitor ought to withtand normal operation spikes for a long, long, long time.)

I have seen capacitors fail open like this in HVAC compressor applications (Start/Run capacitors) When measured with a capacitance bridge they have failed open. i.e. no capacitance left. I believe these capacitors are designed to fail open, providing some short circuit protection in this fashion.

I have witnessed a similar phenomena with mylar film capacitors over the years. (In a somewhat different application) The net effect was that the capacitor loses it's capcitance. I have seen 1 micro-farad capacitors drop down to less than 90 nano-farad. The net effect is that the capacitor can no longer properly provide the parameter requested. i.e. if a start/run cap, there is very little phase shift provided by almost "zero capacitance" If a ripple capacitor in a DC power supply, the DC output becomes very choppy, if a pulse of current is required in a pulse application, it only coughs up a tiny trickle of energy.

 

[ScottyUK]

Self-healing capacitors 'heal' themselves by locally vapourising the metalised film when a dielectric film breakdown occurs, rather than forming a short circuit and tripping a breaker or blowing the capacitor up. Breakdown is usually caused by excessive voltage such as those which exist on the public mains supply. The 'healing' mechanism causes gradual loss of surface area of the capacitor plates, hence the progressive loss of capacitance.

I think Evox-Rifa published a few tech notes about this a while ago. There is a brief description in pages 8-9 of

http://www.evoxrifa.com/technote_pdf/rfi_fact.pdf

. Their app notes are generally worth a read if you have time.



----------------------------------

Sometimes I only open my mouth to swap feet...

 

[itsmoked]

So they aren't 'self healing' they are just not 'self destructive'

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com