Depth of discharge of electrolytic capacitors.

[mrinalmani]

What should be a safe depth of discharge for aluminium electrolytic capacitors undergoing charge/discharge cycle at around 100Hz?
Ripple current is not the issue here, it is well within limits.

I need to place a capacitor across a 230V RMS bridge rectifier with 0.5A average load current. With each cycle the voltage across the capacitor will swing. Is 100V swing ok from capacitor's point of view?

 

[KlausST]

Hi,

A schematic would be helpful.

The description is not clear to me.

Klaus

 

[mrinalmani]

I am attaching the schematic with the oscilloscope output.
The question is how much ripple in voltage is acceptable.
Since electrolytic capacitors are not meant to be frequently charged and discharged, each ripple wave is equivalent to partial charge/discharge. To ensure long term reliability, what should be a rough accepted value of discharge percentage?

 

[KlausST]

Hi,

It's much clearer now...

Since electrolytic capacitors are not meant to be frequently charged and discharged, each ripple wave is equivalent to partial charge/discharge. To ensure long term reliability...

I think the capacitors are made for charge - discharge.
And the amount of voltage ripple is not the point to reduce lifetime.
It is more important to keep the current ripple (which causes power loss = heating) within it's specification.

How much voltage ripple is acceptable depends on your application. For a lamp or a heating application there is no need for a capacitor at all.

Klaus

 

[mrinalmani]

Capacitors are indeed meant for regular charge/discharge, but not electrolytic ones. The electrolytic version have their life limited to generally 10K cycles.

 

[FvM]

The electrolytic version have their life limited to generally 10K cycles.

Where do you see this specification? Sounds like an erroneous believe to me.

 

[KlausST]

Hi,

The electrolytic version have their life limited to generally 10K cycles.

If you have a lifetime limitation, then you should find the information in the relating datasheets.

In the past I have read a lot of electrolytic capacitor datasheets, but can't remember such a specification.

In case your electrolytic capacitors have this limitation, then I recommend to choose capacitors without this limitation.

Klaus

 

[mrinalmani]

As far as I can recapitulate, every note that I have read related to power storage devices say the same thing.
Electrolytic capacitors are meant to act as a voltage reservoir and just like batteries, they have a limited cycle life.
I am attaching a few clips that mention this.

 

[KlausST]

Hi,

Really some interesting informations. Thanks for sharing.
You did your job right by reading datasheets and application notes directly from the manufacturer. There you get the most reliable information.

Now that there is the limitation, the only thing you can do is to use capacitors without this limitations.
The last picture of your post#8 shows a diagram. In the original PDF just below the chart there is the following paragraph :

lf the application is a circuit that has large fluctuations in voltage, such as a power supply for an AC servo amplifier or an inverter, select a QS,QR series capacitor that allows rapid charging and discharging. QS,QR series capacitors empIoy a special structure to increase their durability against rapid charging and discharging.

Best is to rely on manufacturer information..

Klaus

 

[D.A.(Tony)Stewart]

This article gives a measurable relationship of lifespan of electrolytics and stress factors like temperature rise and voltage margin. T rise is due to circuit impedance relative to internal ESR.

 

[FvM]

The Epcos General technical informations for aluminium electrolytic capacitors give a different view on superimposed AC voltage:

3.1.2 Operating voltage Vop
The capacitors can be operated continuously at full rated voltage (including superimposed AC voltage) within the entire operating temperature range. The permissible voltage range for continuous operation lies between the rated voltage and 0 V.

3.1.5 Superimposed AC, ripple voltage
A superimposed alternating voltage, or ripple voltage, may be applied to aluminum electrolytic capacitors, provided that:
- the sum of the direct voltage and superimposed alternating voltage does not exceed the rated voltage, and
- the rated ripple current is not exceeded (refer to chapter "General technical information, 4 Ripple current considerations") and that no polarity reversal will occur.

There seems to be an unclarity about "charge/discharge operation". As far as I understand, the term refers to high curren discharge with low or zero series resistance, in other words doesn't apply to the present discussion.

3.2.6 Charge-discharge proof
Frequent charging/discharging cycles may lead to a decrease in capacitance. Due to their special design aluminum electrolytic capacitors produced by EPCOS are charge-discharge proof. This means that 10e6 switching cycles will cause a capacitance reduction of less than 10% (Charge-discharge test to IEC 60384-4).

 

[D.A.(Tony)Stewart]

I am attaching the schematic with the oscilloscope output.
The question is how much ripple in voltage is acceptable.
Since electrolytic capacitors are not meant to be frequently charged and discharged, each ripple wave is equivalent to partial charge/discharge. To ensure long term reliability, what should be a rough accepted value of discharge percentage?View attachment 118941

The aging factor in capacitors is not the same as batteries which depends on product of full cycles * depth of discharge.
Capacitors life depends on levels of stress caused by margin to rated voltage and thermal rise from ripple current.

Choosing low ESR Caps , much lower than load will result in lower T rise and less aging.
ESR =1 to 5% Rload, depending on load would be reasonable.

Motor start Caps have to compare with ESR of motor coil on startup, and are rated for ripple current but are always higher ESR caps of similar material, size, voltage and general purpose low ESR use. But Motor start caps have lower duty factor use only at startup. They are also much cheaper than similar Polyester Vac rated caps with low ESR.

here is more stress with less ripple as less ripple also corresponds with a lower duty cycle for charging and thus a much higher ratio of peak power to load power.

In your thread #3, The discharge droop is about 20% of the peak voltage and the charge time is about 20% of the cycle, thus the charge current must be about 5 times discharge current and thus power dissipation 25x.during charge compared to discharge. As a check RC=>>10ms which is inverse to ripple factor.

 

[mrinalmani]

I think Mr. FVM is right about the rapid nature of charge and discharge cycle which the datasheets are mentioning.
Last night after posting this thread I tested an 82uF 400V capacitor.
I applied 230V RMS through one diode, ie. a half-wave rectifier.
I adjusted the load such that the ripple voltage was 160V pk-pk
This leads to...
Vp = 325V
V_rip = 160V
I_RMS = 0.7A
Frequency = 50Hz

Although this is not a complete discharge, it still is a 50% discharge during each cycle.
I let the setup operate for two hours. But nothing happened.
Temperature rise was nearly 45 C
I also checked the vent area. There was no bulging and no convex appearance due to internal pressure.

Similar sized capacitors are averagely rated at around 0.6A to 1.5A. I think the temperature rise was justifiable.
I will test the circuit again with complete charge/discharge, but for that that a lower frequency source will be needed to ensure that the ripple current limit is not crossed.

 

[KlausST]

Hi,

But nothing happened

The datasheets say that there is a decrease in capacitance. Did you check capacitance before and after the test?

Your increase in temperature tells me it is a standard capacitor with about 2 Ohms ESR. Am I wrong here?
With a high quality high frequency, low ESR capacitor i'd expect less heating.

Klaus

 

[FvM]

I didn't notice before, but the first datasheet excerpt in post #8 is referring to polypropylene film capacitors. It's irrelevant for the present discussion. To anticipate possibly questions, the 20% ripple voltage rating isn't a general limitation of poylpropylene capacitors, it only exists for specific types.

Returning to the thread topic, besides the said high current charge/discharge and wellknown ripple current ratings, there are apparently no indications in datasheets about problems with higher ripple voltage.

So the test results are just expectable.

 

[mrinalmani]

Oh yes! #8 image 1 is for film capacitors. It went unnoticed.

I don't know much about the capacitor rating. But after googling a bit, it appears to be rated below 500mA RMS.
And of course I checked only the physical appearance and the temperature rise and not the depreciation in capacitance.