EMRFD Message Archive 14447

Message Date From Subject
14447 2017-11-26 13:12:24 boblarkin02 High value Chip Capacitors for Bypass and Filtering
For general info, I posted notes relative to measurements I made on the multi-layer chip caps, such as 0.47 uF in an 0805 package.  It is at
http://www.proaxis.com/~boblar k/MLCC/MLCC_Meas1.html

This is much along the lines of the material in section 2.8 of EMRFD, but also reflects the way these chip caps have gotten smaller and higher in capacity.  It also has some measurements on power filters made with these caps.

Comments are welcome!

Bob  W7PUA
14448 2017-11-26 17:44:47 Bill Carver Re: High value Chip Capacitors for Bypass and Filtering
Those are exciting measurements.
Not even close to being on the same level as your measurements, but I filtered 13.6V to a 100W switcher INSIDE an SDR package with three grounds of vie 0.33uF SMD caps in parallel on a scrap of PCB bolted to the chassis, with copper foil rolled under the board edge, and ferrite rod 12uH coils, and could find no switcher spurs at the -135 dB noise floor of the receiver.
W7AAZ


14449 2017-11-26 17:53:53 Bill Carver Re: High value Chip Capacitors for Bypass and Filtering
TYPO: each of the the three capacitors to ground were composed of "FIVE" 0.33 uF SMD caps in parallel. Not "vie" 0.33 uF
Bill

14450 2017-11-27 08:32:02 jgaffke Re: High value Chip Capacitors for Bypass and Filtering
One weird characteristic of X5R/X7R MLCC caps that I was recently made aware of is how capacitance changes with voltage.  When using a bleeding edge cap that packs the maximum capacitance into the smallest physical size, the capacitance as the DC voltage approaches the rated maximum voltage for the capacitor might be half the rated capacitance.  The rated capacitance is only seen near zero volts DC.

Hmm, might be a useful trick for tuning a preselector or oscillator?

Jerry, KE7ER
    


---In emrfd@yahoogroups.com, wrote :

For general info, I posted notes relative to measurements I made on the multi-layer chip caps, such as 0.47 uF in an 0805 package.  It is at
http://www.proaxis.com/~boblar k/MLCC/MLCC_Meas1.html

This is much along the lines of the material in section 2.8 of EMRFD, but also reflects the way these chip caps have gotten smaller and higher in capacity.  It also has some measurements on power filters made with these caps.

Comments are welcome!

Bob  W7PUA
14451 2017-11-27 09:32:53 Graham / KE9H Re: High value Chip Capacitors for Bypass and Filtering
They are also piezoelectric and therefore microphonic.
They can "sing" when used in power supply circuits.
Be careful using them in low level audio or microphone input circuits.
Or, if mechanically stressed, they can add Voltages to your input.
Tap on the board, and hear the tapping in your output.
--- Graham / KE9H

==

14452 2017-11-27 12:06:19 jgaffke Re: High value Chip Capacitors for Bypass and Filtering
Speaking of "mechanically stressed", those bleeding edge caps have extremely thin
dielectric layers, and thus are mechanically fragile.  A bit of board flex and they can
become internally shorted with no apparent damage when viewed externally under
a microscope.  I've seen it happen, smoked a bunch of boards is spectacular fashion.

It can help to place them near other parts that are mechanically stiff.
And/or orient them such that most board flex would try to bend them across the width, not length.
And/or go to a bunch of 0603's or 0805's instead of one large 1206.

Mostly, avoid the biggest cap/voltage combination for any particular size in any particular catalog,
select something a factor of two or more down from the best they can cram in there.
Especially if going with a second tier manufacturer.
That should help with the capacitance change per applied voltage and also with mechanical strength 
on those really dense X5R and X7R surface mount caps.  

14506 2018-01-10 08:53:14 ipetroianu Re: High value Chip Capacitors for Bypass and Filtering
And talking about board flex and possible failure there are models made to take board flex and the failure is fail safe, meaning they fail open. In either case you can mitigate the failure by using two caps instead of one.
If they fail short then use two in series, if they fail open place two in parallel. Also, use the smallest foot print and place them perpendicular on the edge of the board.
One more thing: board flex tolerant models are used in automotive, so search for automotive qualified, or AECQ qualified devices.
Ion
VA3NOI