Maker Pro
Maker Pro

Ceramic capacitors substitute

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
I have been reading about the global shortage of MLCCs and also how the properties of ceramic capacitors can vary if the PCB is flexed.

Some manufacturers are recommended the use of newer polymer (electrolytic) capacitors as an alternative. Can polymer capacitors be substituted where a ceramic capacitor would normally be used e.g. supply bypassing and in an oscillator circuit for the timing element?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Aluminium Polymer capacitors have high leakage compared with normal aluminium electrolytic caps and waaaaay more than ceramic. As such, I wouldn't use a polymer cap as a replacement for a ceramic in an oscillator. I'd choose a film capacitor of some type.

As for bypassing, tantalum or plain old ceramic (not mlcc) should be fine.

Mlcc caps don't tend to vary in capacitance when the board flexes, they tend to snap and fail either open or short. They do vary in capacitance with applied voltage though.
 

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
I was reading some of the manufacturer's white papers and they claim that polymer technology is superior to MLCCs. For example the Panasonic POSCAP or OS-CON:

https://www.newark.com/wcsstore/Ext...ic/PAN_18072_whitepaper_MLCCvsPolymer_web.pdf

https://pdfs.semanticscholar.org/presentation/1f64/365df3f3d486a2e34d8e660b21765d9b5892.pdfhttps://pdfs.semanticscholar.org/presentation/1f64/365df3f3d486a2e34d8e660b21765d9b5892.pdf

I know this may be biased as it's written by Panasonic, but I have seen similar recommendations from other capacitor manufacturers.

If a PCB may be subject to slight flexing, would normal ceramic capacitors be fine rather than MLCC - in particular when used as the timing element in an oscillator?

I always thought film capacitors were really only good for high-voltage applications as the capacitance can drift significantly over time and aren't as stable with temperature fluctuations?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
You will note that those documents have almost no mention of leakage. The leakage characteristics of polymer caps is still an issue. In applications like power supplies it can be ignored, but definitely not in timing circuits.

I'm not saying that polymer caps aren't good, they have some amazing characteristics, but there are still cases where you would choose a ceramic or a tantalum, it even a regular electrolytic over them.

Will they improve? Sure. Nobody uses waxed paper capacitors any more; once upon a time they were the go-to capacitor. As polymer caps improve, maybe regular electrolytics will be the next capacitor to go the way of the dodo.

And you seem to have film capacitors wrong in your head too. There are many types of film capacitor, if you want the best in stability, leakage, any temperature performance, you'll probably end up with a film capacitor. Years ago you would have chosen mica. These days we have plastics which perform similarly and which are far easier to handle.

I just happen to have been looking at a design where polymer caps might be a very good option, so I've been reading datasheets and doing comparasons :)

Oh, one other thing, at the moment the polymer caps can also be more expensive than some alternatives, although it looks like that difference is reducing.

Another really weird thing is that surface mount polymer capacitors have significantly worse ESR characteristics than through hole equivalents. I'd love to know why that is the case. (It's especially weird given that the metal can type SMD devices are not that much different to their through hole cousins.)
 
Last edited:

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
Thanks for your reply (*steve*), I've been refreshing my knowledge and some types of film capacitors are indeed preferred in applications that need frequency stability e.g.Polypropylene.

So would this Vishay PolyTant be suitable for use in an oscillator circuit:

T55P475M010C0200 (0.2 Ohms ESR)

Also, would this be a suitable substitute as a bypass capacitor for a microprocessor supply:

ECHU1C104GX5 Polyphenylene Sulphide

I'm trying to avoid ceramic capacitors as the PCB is likely to flex.
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
My initial feeling is that I would go the other way, but I'll look them up later.
 

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
The oscillator frequency is <20Hz and I seem to recall reading that the slightly higher ESR in tantalums can help damp any ringing?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
The datasheet for the T55P475M010C0200 mentions that it's suitable for supply bypassing and does not have any information about leakage or temperature stability of its capacitance. I wouldn't recommend it for critical timing purposes.

The ECHU1C104GX5 on the other hand has a datasheet that is all about capacitance stability with temperature and frequency and all the other things you want in a timing circuit or oscillator.

So is go with my gut feeling and use them exactly the opposite way you thought.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Oh, the are ceramic capacitors that have flexible mountings designed for use in applications where there might be large temperature changes, unleaded solder, and/or board flexing.

Incidentally, lead free solder is apparently more rigid than tin/lead solder, making breakages of ceramic capacitors more likely.

You can also consider adding slots in your board to limit the amount of flex.

I believe Dave Jones posted a video about all of this that you might like to look up. (Look for eevblog #1037)
 

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
The datasheet for the T55P475M010C0200 mentions that it's suitable for supply bypassing and does not have any information about leakage or temperature stability of its capacitance. I wouldn't recommend it for critical timing purposes.

The ECHU1C104GX5 on the other hand has a datasheet that is all about capacitance stability with temperature and frequency and all the other things you want in a timing circuit or oscillator.

So is go with my gut feeling and use them exactly the opposite way you thought.

Hmm, I can't seem to find SMD film capacitors readily available in 4.7µF.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
It all depends on how much stability you require. At that frequency leakage isn't likely to be a significant issue.

Have you watched that video yet? It's almost half an hour long, so I don't think you have.
 

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
The video was really useful, thanks!

I think I'll probably opt for some automotive-qualified ceramic capacitors that can tolerate board flex such as the Vishay VJ-series or Murata VCJ-series.

Edit: Just my luck, 4.7µF seems to be on back-order in both of the above series! However, there is a higher voltage part available in the TDK CGA-series.

Would X7R dielectric be suitable for use in the oscillator circuit?
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Would X7R dielectric be suitable for use in the oscillator circuit?

I can't tell because I don't know anything about the oscollator other than its frequency.

Perhaps you could post the circuit and give the parameters under which it should operate. Things like frequency tolerance, temperature range, supply voltage range etc. If you're using the voltage across the capacitor to generate a ramp, does it need to be linear, etc

You could also try looking up the characteristics of X7R dielectric and try answering the question yourself.
 

Cirkit

Oct 28, 2015
155
Joined
Oct 28, 2015
Messages
155
The oscillator is used for a hardware PWM circuit to drive a motor. I have attached a schematic. The arrow points to where the output is taken.
 

Attachments

  • Op-Amp oscillator schematic.png
    Op-Amp oscillator schematic.png
    11 KB · Views: 25

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
OK. it's nothing really critical. Use an aluminium electrolytic or a tantalum.
 
Top