I'm using few electrolytic caps as a “battery” for my circuit when there are power outages. As I have a very unique dimension requeriment, I found the sweet spot between size/capacity in 2x 2200uF 26V and I have come to the results that charging them up to 17V is sufficient for the requirements, including a safety margin, However the more the better.
I have not been able to obtain clear confirmation of the benefit of charging a capacitor below its rated voltage . As this circuit will be working continuously my main concern is the cap to life more than you and me and secondly, to have the maximum stored energy
I would like to hear your experiences about the keys for keep the caps last as long as possible and whether it is worthwhile to undercharge them or even replace them with a higher or lower voltage.
Electrolytic capacitors tend to last a long time. I'd expect more than 10 years. I have some electronics a lot older than that! The capacitance may drop over time but electrolytics are not usually used where capacitance value is critical.
They might go bad sooner if left unused and not occasionally charged. Otherwise, as far as I know applied doesn't affect life unless you go over-voltage, and as you say, some "safety margin" is a "good practice".
...As you may know, capacitors make lousy batteries (by themselves) because the discharge curve is the opposite of an ideal battery... The capacitor discharges fast at first and then the voltage levels-off as it becomes more discharged. But it you just need to keep the circuit alive for a short time, it's OK. The last time I changed batteries in my home thermostat I put a capacitor across the battery holder so I didn't have to re-program everything after replacing the batteries. But I have a different plan for next time - I bought an extra battery holder and I'll connect in parallel with clip leads while changing batteries.
Be sure to consider the tolerance of the capacitor value and how it may affect your project over time. I typically use two electrolytic capacitors in parallel to ensure continued operation in case one fails.
There is no benefit nor problem. You just have much fewer electrons stored in the chemical reaction of the electrolytic compound. For a capacitor rated at 26 volts, you are storing about 1/2 the number of electrons as you could be storing.
The other side of your design is how rapidly you need to remove the electrons from your capacitors. That is limited by the internal resistance of the capacitors, aka ISR.
The failure mode I generally see is they lose value. A lot depends on the quality of the CAP. Over time, the electrolyte inside the capacitor can dry out, leading to a reduction in capacitance and an increase in Equivalent Series Resistance (ESR).
It's okay, I have some in a box I won't be opening for maybe months or longer. But the OP might check out the kind used on PC mobos to keep CMOS alive. There are some that hold a Farad or more, the cheaper ones I have are maybe .1 or .2 F which is pretty incredible to me!
Electrolytics are very reliable and you can still find surplus equipment from many decades ago where they still work. EG I have old laptops about 20 yo still working, and a power supply I built about 50 years ago with no issues.
Ripple current kills electrolytics so you should avoid that.
You should know Q = Cv = iT
which will show you how the voltage will drop as you draw current.
and 4mF * 17V will supply 4mA for 17 seconds (to complete discharge)
If you are using a 20V supply and dropping it to 5V for a circuit you are wasting a lot of energy unless you use a buck converter.
Can you please point What could be the keys to a good cap? or just go for a good japanese brand?
Yeah, I was thinking on it at the beginning, but I was worried about the charge circuit design and the life cycles so I desisted from that path
That gives me confidence, but many of these devices do not run all day and night continuously, which would probably have caused them to fail by now (or not)
Thanks for that suggest, the power supply is definitely not stable (some pulses) and I will look into it
I'm doing that with a buck regulator but what do you mean exactly? the waste from 5v to 0v?
That's still not a unit of charge.
Again, 50-100mW for how long? Not an hour, I assume. 50mWh = 180J
18V on a 2200uF capacitor is approx. 0.36J Not to mention leakage of a capacitor (virtually all regular caps will pretty much be self-discharged after an hour).
Anyway, feel free to ignore this; I assume you've somehow worked out that the 360mJ you're storing is sufficient for your application and that you only need to bridge brownouts/blackouts lasting a couple of milliseconds.
Yes, even more because the regulator still works at 5.1v
Not sure if I understand your formula, cause seems that contemplates to attach a fixed load of 10mA to the cap or just emulate a LDO, in my case using a switch regulator the load will be lower at 17V,16,15.. and son on..
This is the formula I'm using and almost match my real results
if I finally confirm that undervoltage has no benefit I can raise the charging voltage to 20V which would give 4 extra seconds
I take this post to clarify that my concerns are about maximising capacitor life, and not about knowing how much energy capacitors will get or how many seconds will last my circuit powered .
My experience of replacing faulty capacitors in power supplies is that to expect 10 years is reasonable, to expect 20 years is possible but might be asking for trouble. Read the datasheet from the manufacturer, capacitors are not all created equal. If you want really long life get hold of ones specified for spacecraft. I imagine they are not cheap.
I can't be more specific than others have been, but agree with stay below the rated voltage.
If you are looking for an answer something like: For 10 years life don't exceed 80% of the rated voltage, for 20 years life don't exceed 50% then I think you will be out of luck, I don't think such an answer exists.
Have you done the study to identify the various manufacturing procedures used in making the capacitor you are using? Have you looked at the various different manufacturers and picked the ones making long-lived capacitors? Then I suspect you have no way of knowing which capacitor will have the longer life.
.