A Cautionary Capacitor Tale

The title of this "tutorial" is Temperature and Voltage Variation of Ceramic Capacitors, or Why Your 4.7µF Capacitor Becomes a 0.33µF Capacitor. Quite an eye opener.

http://www.maximintegrated.com/app-notes/index.mvp/id/5527

I found some nice reasonable 10µF 0805 MLCCs, have been using them and thinking I'm fat dumb and happy. Things are in fact working fine. But now I need to RTFDS and I'm quite apprehensive about what I might find!

Temperature / voltage driven variations of capacitance have been known for decades, particularly in the RF world.

dhenry:
Temperature / voltage driven variations of capacitance have been known for decades, particularly in the RF world.

It was the voltage-driven variation that really surprised me.

Surprised me also, however I've never been a user of SMD mounted capacitors in my projects, so haven't been bitten by this. It appears the smaller the SMD package selected the more drastic cap Vs running voltage variations seen:

Note, first, that as the package size increases, the capacitance variation with applied DC voltage decreases, and substantially.

We see a steady improvement as we progress to larger capacitor sizes, until we reach the 1210 size. Going beyond that size yields no improvement.

As the capacitor vendors have made smaller and smaller components, they have had to compromise on the materials used. To get the needed volumetric efficiencies in the smaller sizes, they have had to accept worse voltage coefficients.

Remember too that the ceramic capacitor type designations, such as X7R X5R, and Y5V, imply nothing about voltage coefficients. Engineers must check the data to know, really know, how a specific capacitor will perform under voltage.
Finally, keep in mind that, as we continue to drive madly to smaller and smaller sizes, this is becoming more of an issue every day.

Lefty

I checked the datasheet for the 10µF 0805 MLCC I'm using, and there is nothing about capacitance vs. voltage :frowning:

That why the article stated you must seek out the 'detailed information' that would contain information on voltage variation.

This is a very important point, so I will repeat it. A vendor can call a capacitor X7R (or X5R or any other type) as long as it meets the temperature coefficient specs, regardless of how bad the voltage coefficient is.

The larger lesson here, as you may have surmised, is "read the data sheet," every time, no exceptions. Ask for detailed data when the data sheet does not contain sufficient information. Remember too that the ceramic capacitor type designations, such as X7R X5R, and Y5V, imply nothing about voltage coefficients. Engineers must check the data to know, really know, how a specific capacitor will perform under voltage.

Lefty

retrolefty:

[quote author=Jack Christensen link=topic=137624.msg1034156#msg1034156 date=1355619703]
I checked the datasheet for the 10µF 0805 MLCC I'm using, and there is nothing about capacitance vs. voltage :frowning:

That why the article stated you must seek out the 'detailed information' that would contain information on voltage variation.
[/quote]

Yeah, saw that, thought I'd check a few. Didn't find any capacitance vs. voltage info in datasheets from several manufacturers. Sounds like information that is only divulged on request. I'd say this is a dirty little secret of the MLCC manufacturers.

Yeah, saw that, thought I'd check a few. Didn't find any capacitance vs. voltage info in datasheets from several manufacturers. Sounds like information that is only divulged on request. I'd say this is a dirty little secret of the MLCC manufacturers.

That because it's not something they probably want to brag about. :wink:

I always felt in my gut you whimper snappers using all ya fancy schmancy SMD sized stuff were headed down a path towards evil. Excuse me, my lawn is calling. :wink:

Lefty

I'd say this is a dirty little secret of the MLCC manufacturers.

You probably want to approach this "problem" with an engineering perspective: just because something is less than ideal doesn't mean it has to be fixed at any cost, or at all.

For example, ask yourself where those capacitors are typically used and see if valuation variations make a material difference.

I would also submit that anyone using a 22uf ceramic is most likely a nut job (or doing a nut job).

Yes or "The Sky is Falling," from whence comes this report?. MLCC's were just being advertised in 1220 cases when I retired and I've seen no other information to support this.
I 'm not trying to say it isn't so, just that it is New to me as I've not seen prior discussion. It certainly sounds reasonable. I use them but only for GP bypassing, I'd not trust them to measure as they are marked.
Any bypassing I do is a 220Uf Electrolytic, A 47Uf Tant and a .1 Uf for each IC or inch of Vcc trace/conductor/wire used. A small investment for peace of mind, cheap too @ about $.40 per 4 Sq in PCB.
The Tantalum costs $0.25 US...

Bob

retrolefty:
That because it's not something they probably want to brag about. :wink:

No doubt!

I always felt in my gut you whimper snappers using all ya fancy schmancy SMD sized stuff were headed down a path towards evil. Excuse me, my lawn is calling. :wink:

Baha! Well I am a relative newcomer to that game, but have been getting along quite well. I work under one of those magnifying lamps with the circular fluorescent tube. Haven't tried anything smaller than 0805, I might be able to do 0603, will have to try it one of these days. The stuff is actually kind of fun to work with, and can certainly save some real estate. My current project has all SMD resistors and capacitors, except for the electrolytics. Also an SOIC (very easy), and a SOT-23-6. The SOT package is tiny but surprisingly easy to solder. Then there are a couple big honking DIPs so it's definitely a scatterbrained sort of thing! I do believe I cooked one of the SOTs the other day soldering it. Pretty unusual for me, I'll have to be more careful and maybe dial the iron down a bit.

Lawn??!! Waaay past that season here!

Lol this is quite interesting, thanks for bringing it up, I just designed something and I had to calculate a capacitor for a boostrapping circuit, and now I'm hoping it works since I used a 0805 2.2uf cap, atleast now if I have problems with that ill know where to look, and maybe ill just automatially stack an extra cap on there, because it seems I will definetly have some sort of drop

Are you all using metric or imperial measurement? I am seeing references to standard sizes that are used in both measurement methods.