Component Selection Question

[Andy Peters]

Yes - have to consider what were other options at the time and compare to now. Not just copy for 'historical' accuracy like it's a listed building  ;D

Yes - they can 'go with a bang' . Playing Devil's advocate I'd probably say that with an expensive board I'd get the build checked then checked again 

I wish I could post a photo of the board, because then I could say, "OK, _you_ find the reversed cap!" (It's packed full of tiny stuff.)

Mmmm...100uF/25V/X7R/1206. That's impressive,
Do you have a link or part number for that - I couldn't find it and it could be of real interest.

A TDK part at Digikey. Expensive. Note the metal frame leads. 

Those organic polymer caps look interesting - well in as much as a cap' can be  interesting ;D - but they haven't lined up with anything I've needed to date.

They're basically a substitute for tantalum caps. When you need bulk capacitance in a reasonable package but you the ESR of a standard aluminum electrolytic will burn you, the organic polymer caps are a compelling replacement.

 

[Newmarket]

I wish I could post a photo of the board, because then I could say, "OK, _you_ find the reversed cap!" (It's packed full of tiny stuff.)

A TDK part at Digikey. Expensive. Note the metal frame leads. 

They're basically a substitute for tantalum caps. When you need bulk capacitance in a reasonable package but you the ESR of a standard aluminum electrolytic will burn you, the organic polymer caps are a compelling replacement.

Fair point. I don't have that degree of component density / inspection challenge to deal with atm.

I had a feeling that the 100u / 25V Ceramic would be a bit pricey 

On the organic polymer caps I could have been a bit clearer - I'm aware of them and have considered using them a few times but the numbers (uF / Vdc) haven't worked out for my particular applications. They do look good specwise though.

 

[aomahana]

Hi there,

With regard to tantalum capacitors ....

I have experienced two sudden explosive and smoking failures.
Both were in commercial designs, no reverse or over voltage.
They had been in use for some years though.

The first was on a RS Components designed low voltage power supply board.
And the other, a bypass capacitor, on the rails of the output IC of a Lexicon PCM90 reverb.

I am now cautious with the use of tantalum, and would usually choose an alternative if possible.

 

[CurtZHP]

Do not use tantalum caps: they explode when reverse-biased or when subject to working voltages above their ratings.

I had an electrolytic cap explode in my face when I inadvertently reversed it.  The casing hit me right in the nose! 

And it was no fun cleaning that electrolyte snot off the rest of the board.

 

[JohnRoberts]

I had an electrolytic cap explode in my face when I inadvertently reversed it.  The casing hit me right in the nose! 

And it was no fun cleaning that electrolyte snot off the rest of the board.

Modern electrolytic caps are designed with a pressure relief (rubber bung) in the bottom so the metal cans do not literally explode creating shrapnel, but the electrolyte being forcibly ejected can make a mess. 

I've shared this story before but back in the 60's working on an early switching power supply I blew up a large film capacitor... luckily I was away from my bench the moment it exploded but it rained down small particles of metallized film all over the bench, looking like it snowed.

====

WRT tantalum, modern parts are probably better than the old grenades, but the modern low Z aluminum are so good that there is no compelling need to use tantalums.

Last tantalum I used on purpose was back in the 80s for a decode RC time constant, to match the encode that also used a tantalum. While very subtle in a time constant circuit with different attack and decay impedances, the tantalum's dielectric absorption could make a side chain tracking difference vs aluminum (very very subtle, and arguably over engineered.)

JR 

 

[rackmonkey]

For small value (picofarad) generally you want a C0G or NP0 dielectric ceramic cap. Some might argue for film caps for best stability. For standard 0.1 uF decoupling the X7R ceramics are fine. 

Just be careful with where you decide to put C0G/NP0 ceramics, depending on package. Ceramics in general suffer from piezoelectric effects which create microphonics. While the newer C0G/NP0s don't exhibit this property in SMD packages, the leaded ones can. Note that the statement "no piezoelectric behavior" doesn't appear above the leaded package products here:

http://www.tscgroup.com/acrobat/sm01001b.pdf

Blog post of a TI app engineer upon his realization of the potential effects in circuit. X7R's are not good in this regard:

https://e2e.ti.com/blogs_/archives/b/precisionhub/archive/2014/12/23/stress-induced-outbursts-microphonics-in-ceramic-capacitors-part-2

So use some caution and common sense here. Keep leaded ceramics of any type out of the signal path, at least. I just use film everywhere it's practical (size being the primary limitation).

BT

 

[JohnRoberts]

Note that
there are no measurable piezoelectric effects in Class 1 capacitors, such as C0G or NP0 - neither of which is
considered ferroelectric

COG or NPO are good capacitors for audio paths.

JR

 

[rackmonkey]

Even the leaded ones? There's a lot of mixed information on that point, a lot of which is from seemingly credible sources. Why the lack of statements regarding piezoelectric susceptibility in the data sheet?

BT

 

[Andy Peters]

Even the leaded ones?

Who uses leaded capacitors in production any more?

 

[rackmonkey]

The OP doesn't sound like he's a boutique manufacturer or otherwise. Straight up DIYer, as I read it. I don't know too many like him that do SMD. I've been building/designing for years and only mess with that stuff when I have to. 

BT

 

[JohnRoberts]

Even the leaded ones? There's a lot of mixed information on that point, a lot of which is from seemingly credible sources. Why the lack of statements regarding piezoelectric susceptibility in the data sheet?

BT

I just cut and pasted  from a kemet document about ceramic capacitors.  I consider KEMET credible.

Cheap ceramic caps are crapo and should not be used for audio... cog and npo are neutral and linear...

JR

 

[Newmarket]

Modern electrolytic caps are designed with a pressure relief (rubber bung) in the bottom so the metal cans do not literally explode creating shrapnel, but the electrolyte being forcibly ejected can make a mess. 

Or they have a scored cross on the top that will rupture to let the pressure out.
I happened to see one go a few years ago. I just happened to be at the bench when it went - quite a visual event with the 'steam' coming out in a swirl. Seemed to be a random failure.

I've shared this story before but back in the 60's working on an early switching power supply I blew up a large film capacitor... luckily I was away from my bench the moment it exploded but it rained down small particles of metallized film all over the bench, looking like it snowed.

JR

Oh yes - that sounds familiar from back in the mid/late 80s for me .  Linear Power Supply circuit set up on bench. Idling. No one near it but we saw the space fill with the 'snow'. Again very visual but bit of a pain to clear up 

 

[Newmarket]

Who uses leaded capacitors in production any more?

I do - but these are high voltage ceramics (>1kV) rated devices. And they may be going out of production 
Due to their size and shape they seem less susceptible than smaller capacitors with thinner legs.
And I don't think microphony would be a significant issue in my non-audio applications.

 

[Monte McGuire]

Even the leaded ones? There's a lot of mixed information on that point, a lot of which is from seemingly credible sources. Why the lack of statements regarding piezoelectric susceptibility in the data sheet?

Leaded caps are MLCCs with leads soldered on, dipped in epoxy. There should be no difference. The problem may be poorly specified caps that are not really NP0/C0G sold as NP0/C0G.