Lowest rail voltage that can be used with 50V electrolytic capacitor

[cupoftea]

Hi,
Say, you have a design with 100 Electrolytic caps on a 35V rail for which you spec 50V El caps. You need one similar value cap for the 5V rail….obviously you re-use another of the 50V El caps…..you don’t buy a single 6v3 El cap for that….

….however…

Maybe you do, because Electrolytic capacitors are not capacitors below 10% of their rated voltage? Also, they have severe loss of capacitance below 20% of their rated voltage?
What is the lowest voltage rail that a 50V El cap could be used on?

I’ve checked , and there is no capacitor manufacturer that discusses this.

 

[wwfeldman]

after being in storage for a long time, or being used at 1/4 to 1/3 their rated voltage,
electrolytic caps need to be re-formed to restore/repair the oxide layer

you can use the 50V cap in a 5V circuit, but after a while, it will be a 5V cap,
and if used as a 50V later, will need to be reformed first.

 

[FvM]

Maybe you do, because Electrolytic capacitors are not capacitors below 10% of their rated voltage? Also, they have severe loss of capacitance below 20% of their rated voltage?

Neither of the assumptions is correct. Are you rewriting theory of electrolytic capacitors?

Epcos/TDK general information states:

The permissible voltage range for continuous operation lies between the rated voltage and 0 V.

 

[cupoftea]

Thanks, somewhere i definetely have a Babani book on 555 timers....it clearly says that El caps cannot be used for timing in 555 circuits, as they tend to have too high voltage ratings, and dont work as caps.

Also, an extremely experienced Electronics manager told me that el caps dont form below 10% of rated voltage.

El caps have a similarity to batteries, and this is the origins of this.

.......***Quoting***.....
...quoting page 11 of the book "IC555 projects" by E.A.Parr (babani publishers) "An electrolytic capacitor does not become a capacitor until about 0.1 of its voltage rating....."
book first published in 1978
book revised and reprinted n 1981
Book printed last in 2006.

The permissible voltage range for continuous operation lies between the rated voltage and 0 V.

...thanks, but i believe that just means that its "permissible"...ie not going to damage the cap...but it wont have any capacitance below 10% of rated v,

 

[FvM]

Neither experience nor manufacturer specifications and application notes indicate that the claimed behaviour exists.

 

[The Electrician]

...thanks, but i believe that just means that its "permissible"...ie not going to damage the cap...but it wont have any capacitance below 10% of rated v,

If this were true, then LCR meters wouldn't be able to measure the capacitance of electrolytic caps when there is zero applied DC voltage. But, in fact, LCR meters have no problem measuring them, and the measured value is usually within the large tolerance band that electrolytics have. It's just not true that electrolytics don't have any capacitance below 10% of rated voltage.

Get an LCR meter and measure some electrolytics yourself.

 

[Easy peasy]

The lowest rail it can be used on is 0.00000001 volt rail.

 

[cupoftea]

Get an LCR meter and measure some electrolytics yourself.

Thanks, i think they use high frequency to get round the problem in LCR's

There is a difference in capacitancee seen from charge_time testing, and impedance testing with high frequency......and at low voltages, i think the high freq way works, but not the charge_time test.

 

[FvM]

Sounds like you are trying to save somehow the speculative claims of your previous posts instead of thinking it over. It's well known that AC and DC (charge/discharge) capacitance are different, with factor 1.1 to 1.5 higher DC value. The behaviour can be seen in principle by using a RC ladder model that represents frequency dependent losses and capacitance. There's however no significant voltage dependency.

A different question is if continuous operation at lower or zero DC bias will cause a long term (over years) reduction of oxide layer thickness. I didn't yet find respective manufacturer information. If such an effect exists, capacitance is expected to increase rather than decrease.