cap?

mercrazy

Newly Enlightened
Joined
Jul 28, 2014
Messages
160
my LED driver calls for 10uf, x7r ceramic cap. what happens if i use a 25volt cap and it gets 30volts?
thanks
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
the manufacturers tend to build the caps to tolerate more voltage than the nominal rating. This helps them tolerate some of the usual manufacturing variation, and probably reduces the number of angry phone calls that they get. :)

To use a cap beyond its rated voltage is not a great idea, but probably means a shorter lifetime.

Don't forget that some families of ceramic caps show a change in capacitance as the applied voltage changes. This usually means that the effective capacitance decreases with increasing voltage. IIRC, the X7R wasn't too bad in this regard. You might want to check the datasheet, though.
 

ssanasisredna

Enlightened
Joined
Oct 19, 2016
Messages
457
All the high density dielectrics behave pretty similarly w.r.t. DC bias voltage and capacitance. If it's a 1210 cap, it will be about 3uF at 30V, a 1206, 2ish uF, 0805 as low as 1uF.
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
All the high density dielectrics behave pretty similarly w.r.t. DC bias voltage and capacitance. If it's a 1210 cap, it will be about 3uF at 30V, a 1206, 2ish uF, 0805 as low as 1uF.

I disagree. I wanted 10 uF at 25V for a board I was working on recently. I was looking exclusively at 1206 caps, and found that they ranged from around 3uF to barely 1 uF at 25V. I also found that most data sheets do not include this information.

Keep in mind that X7R does not denote a type of ceramic, nor any characteristic of a ceramic except capacitance vs. temperature. Any ceramic that meets the X7R temperature characteristics can be called X7R. And they do vary substantially from one manufacturer to another, maybe even from one line to another.

I ended up using tantalum polymer, as they exhibit far less variation with bias. I ended up with substantially less board real estate for 10 uF at 25V than I could have with ceramic.

If you find this topic interesting, Maxim's app note 5527 is very informative.

To the OP's question, if the driver specs a 25V part, then you probably shouldn't put 30V on the driver. If the driver is designed for 30V, it should have a 35 or 50V part on it. If you are talking about a brief transient at powerup or something like that, it's probably fine.
 

ssanasisredna

Enlightened
Joined
Oct 19, 2016
Messages
457
You could disagree but you would be predominantly wrong .... as I should have clarified all high density ceramic dielectrics. You will note I said all high density dielectrics and did not specify whether X7R, X5S, etc. as for the most part the temperature tolerance of the capacitor has far less impact on capacitance than the size of the capacitor. The actual dielectric also has less impact versus the field strength which is impacted by size/layer thickness, though the dielectric has an impact on these design variables. We would be pretty much talking classII ceramics here. I am sure you could find a 10uF/25V 1206 that was 1uF at 25V, but it would be uncommon these days as it would not be economically viable (maybe some tier 2/3 Chinese vendors). From the major MLCC vendors, Yageo, Murata, TDK, Kemet, etc. the case size (for a given voltage/capacitance) is going to be the dominant variable in how much capacitance drop you will have at rated voltage and not the temp rating. It can be hard to see that in the Murata 5527 tutorial, but it is buried in the graph.



I disagree. I wanted 10 uF at 25V for a board I was working on recently. I was looking exclusively at 1206 caps, and found that they ranged from around 3uF to barely 1 uF at 25V. I also found that most data sheets do not include this information.

Keep in mind that X7R does not denote a type of ceramic, nor any characteristic of a ceramic except capacitance vs. temperature. Any ceramic that meets the X7R temperature characteristics can be called X7R. And they do vary substantially from one manufacturer to another, maybe even from one line to another.

I ended up using tantalum polymer, as they exhibit far less variation with bias. I ended up with substantially less board real estate for 10 uF at 25V than I could have with ceramic.

If you find this topic interesting, Maxim's app note 5527 is very informative.

To the OP's question, if the driver specs a 25V part, then you probably shouldn't put 30V on the driver. If the driver is designed for 30V, it should have a 35 or 50V part on it. If you are talking about a brief transient at powerup or something like that, it's probably fine.
 

mercrazy

Newly Enlightened
Joined
Jul 28, 2014
Messages
160
thanks all. this has been very informative.
the driver chip is rated for 40v. it will be powered by 2- 12volt batteries in series so voltage could go up to 28v when alternator is charging. the chip says to use low esr cap so i thought i was limited to x7r or x5r caps?
i'm only concerned with reliability and not so much with light flicker or whatever problems a sub par cap would cause. datasheet said it would smooth current ripple. also says high esr lowers efficiency.
thanks again.
 

ssanasisredna

Enlightened
Joined
Oct 19, 2016
Messages
457
Given it is battery powered, ultra long life is probably not a concern. Depending on your switching frequency, an aluminum electrolytic or tantalum capacitor could be just fine. Can't comment on the impact of ESR in your circuit without knowing output voltage, current, etc.
 
Top