Capacitor Leakage current
Interesting article on AL capacitors from Epcos
Regarding leakage current and other phenomena
Whenever I check electrolytic leakage current and compare to the manufacturer's specification for leakage, I get results that are not similar.
What are the units used in the standard leakage formula?
Leakage = factor * V * C
usually I see spec quoting IL=0.02VC
using Amperes and Volts and Farads, I often get the factor coming out better than F<=0.00001. This seems so much less leakage that I always suspect I am doing something wrong.
Nippon Chemicon LXZ ALUMINUM ELECTROLYTIC CAPACITORS
I=0.01CV or 3uA, whichever is greater.
Where, I : Max. leakage current (uA),
C : Nominal capacitance (uF),
V : Rated voltage (V)
(at 20C after 2 minutes)
Measuring temperature and voltage influence the leakage
current. The leakage current shows higher values as the
temperature and voltage increase.
In general, the leakage current is measured at 20C by applying
the rated voltage to capacitor through a resistor of 1000 Ohms
in series. The leakage current is the value several minutes later
after the capacitor has reached the rated voltage. The catalog
prescribes the measuring temperature and time.
The text above is from Nippon Chemicon publications.
The last thing capacitor manufacturers want is part
rejection based on leakage. Their parts typically
perform 100 times better for leakage than the
Example 1000uF 16V LXZ Capacitor Leakage
A couple of recent examples using cheap sub min electrolytics.
470uF 16V 0.085uA leakage.
100uF 10V 0.2uA leakage.
Since leakage is proportional to rated voltage, I have a question about leakage during actual use.
Take a capacitor rated at 50V. Lets assume that maximum leakage is specified as 0.01CV.
I understand that if this capacitor is used up to its rated voltage, leakage should not be greater than 0.01CV. But then, you have taken V to its maximum value.
So comes the question: does leakage depend on working voltage?
For example, take a 100uF, 50V capacitor rated for maximum leakage of 0.01CV. Then, it should be around 50uA. But if you use this cap under voltages that don't exceed 10V, would maximum leakage be 10uA, or still 50uA? :rolleyes:
I think it may depend on whether the cap has been reformed to full maximum voltage and how long since it was reformed.
I am pretty sure that a freshly reformed to 50V capacitor will have a leakage <<<<0.01CV @ 10V
Whereas a long time stored and not reformed since manufacture electrolytic will, maybe, eventually meet the <=0.01CV specification after many hours/days/weeks/months of in circuit use at it's very low operating voltage.
eg. the the DC blocking cap in the NFB may be rated for 16V, or 63Vdc, but in circuit it sees ~20mVdc. What is the leakage current? How long does it take for the leakage current to stabilise?
What is the equivalent resistance for that leakage current?
How much will the output offset vary during the "in circuit" reforming due to the NOT equal to unity, DC gain of the amplifier?
I think Andrew has it right. Leakage current in an electrolytic is a function both of current voltage, and the past history. It increases with current voltage, but reduces with past voltage. The oxide film is always slowly degrading, and the leakage current rebuilds it. Therefore it might be concluded that given a fixed voltage the current would eventually settle at about the same level whatever that voltage is. The current might depend more on temperature than voltage, as increased temperature will probably cause faster oxide degradation and so need more current to maintain it.
|All times are GMT. The time now is 01:37 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright ©1999-2015 diyAudio