So I took the advice from Scott and checked out the Bill of Material for a Standford Research Utra Low Jitter Clock Generator (forgot the model number) :
http://www.diyaudio.com/forums/showthread.php?postid=1801262#post1801262
It is, not unexpectedly, full of 1206 SMD caps, mostly X7R's, with less than 10% NPO's.
Just an observation.
Patrick
http://www.diyaudio.com/forums/showthread.php?postid=1801262#post1801262
It is, not unexpectedly, full of 1206 SMD caps, mostly X7R's, with less than 10% NPO's.
Just an observation.
Patrick
EUVL said:
That means this eqpm was designed to work in a non vibration environment and will do so.
But, it seems, you guys have a great pleasure to generalize things.
I mentioned several times, that ceramics are not good for high-end audio due vibrations and their piezoelectric effect.
There are, for sure, other applications where ceramics will work fine, but not audio.
Regards,
Tibi
My audio setup is vibrations free: music stored on computer, external DAC linked through USB and headphones... does it mean I can use ceramic caps ?
And you (Tibi) are again generalizing the term "ceramics".
That is just like stating all films are the same. In other words, that polystyrenes are bad because PET distorts.
NP0s are a non-polar dielectric, UNLIKE class 2 and 3 ceramics.
While there has been posted an assertion that even NP0s have some piezolectric character, there has been NO evidence regarding how much; or whether it is EVEN significant at all, especially in any specific applications NORMALLY encountered.
This is like arguing in a vacuum that polystrenes should NEVER be used in audio because they can't stand high temperatures.
PUT IT IN PERSPECTIVE!
That is just like stating all films are the same. In other words, that polystyrenes are bad because PET distorts.
NP0s are a non-polar dielectric, UNLIKE class 2 and 3 ceramics.
While there has been posted an assertion that even NP0s have some piezolectric character, there has been NO evidence regarding how much; or whether it is EVEN significant at all, especially in any specific applications NORMALLY encountered.
This is like arguing in a vacuum that polystrenes should NEVER be used in audio because they can't stand high temperatures.
PUT IT IN PERSPECTIVE!
pooge said:
The man was clear, get this through your thick skull: ceramic caps are not good for high end audio, they are piezolelectric, period. They may work for space electronics though. The Proton rocket launcher vibrations in Baikonur are negligible.
http://www.electronicstalk.com/news/kme/kme148.html
john curl said:
Sorry, not enough linear for high end audio.
To add insult to injury, they are multilayer, something that is well known as affecting the sound, in particular when used in power supply decoupling, due to the poor piezoelectric properties or ceramic.
Here
Do yourself a favour and get some Russian NOS teflon caps; they cost only $50/pc. (a real bargain) and look wonderful on the 2 sq. inch PCB space they each take.
Please don't confuse these people any more.
I'd hate to come back from these *******' tax forms only to have to herd ADHD stray cats again to get them on point again.
I'd hate to come back from these *******' tax forms only to have to herd ADHD stray cats again to get them on point again.
It's cool that KEMET supplies spice models.
Are there any oppionons on multilayer vs single layer ceramics?
Are there any oppionons on multilayer vs single layer ceramics?
syn08 said:
In fact, I'll take this seriously and not use ceramics ... and yes, I prefer teflon caps.
However, you can start use of magic 1uF NPO, which I bet will be at least as large Russian NOS teflon cap. 😉
But, you are free to do what you like, or consider is sounding good.
Again kemet and again military:
Updated: 15.04.2009 😉
http://www.kemet.com/kemet/web/homepage/kfbk3.nsf/vaFeedbackFAQ/242F5F2E69DCEC7485256EDF004CA495
My dissertation is over !
Regards,
Tibi
Teflon is a little over the top for power supplies, but I always use polystyrene or polypropylene, when I can.
john curl said:
For digital power supplies decoupling, I prefer silver-mica up to 10nF.
For analog power supplies decoupling, polypropylene all over the range.
Arcotronics are cheap and good performers.
Regards,
Tibi
They used KM caps in Soviet rockets launched on Baykonur where size and every gram of weight matters. However, their properties were well known, so why Kemet caps can't be used today for similar purposes?
I used them in audio then; however very carefully after found some of them suddenly with zero resistance. I could not understand why: I did not exceed specified voltage, current, temperature values, and mechanically they were not tortured. Probably, they were from rejected by military QC parties?
Speaking of distortions, everything distorts. Otherwise design of new equipment would be meaningless. And everything, including thyristors and relays, may be used in audio, when used properly.
I used them in audio then; however very carefully after found some of them suddenly with zero resistance. I could not understand why: I did not exceed specified voltage, current, temperature values, and mechanically they were not tortured. Probably, they were from rejected by military QC parties?
Speaking of distortions, everything distorts. Otherwise design of new equipment would be meaningless. And everything, including thyristors and relays, may be used in audio, when used properly.
I would like to point something out about caps. Often we need to use caps for power supply decoupling applications. Many here talk about ceramic caps being extra good with high frequency performance, but this is only partially true.
Virtually all plastic, ceramic and mica caps have a self-resonance that is tied to its value.
In other words, virtually all .1uF caps have a self-resonance at around 4 mega-Hz, more or less depending SLIGHTLY on construction and HIGHLY on lead length.
Since some .1 caps are tiny, then effective lead length can also be POTENTIALLY short, but IF you have to add lead length to get from point A to B, then the resonance will be the same.
The only thing that really counts is DA, which adds dissipation with increased frequency, and series R, that is usually contact and lead diameter sensitive. For example, some Rel styrene caps have lower series R than other Rel styrene caps. I usually use the high series R caps, because I prefer their natural damping characteristics. However, if I were decoupling a switching power supply, I would use the lower series R cap. All caps look inductive above their resonant frequency.
Virtually all plastic, ceramic and mica caps have a self-resonance that is tied to its value.
In other words, virtually all .1uF caps have a self-resonance at around 4 mega-Hz, more or less depending SLIGHTLY on construction and HIGHLY on lead length.
Since some .1 caps are tiny, then effective lead length can also be POTENTIALLY short, but IF you have to add lead length to get from point A to B, then the resonance will be the same.
The only thing that really counts is DA, which adds dissipation with increased frequency, and series R, that is usually contact and lead diameter sensitive. For example, some Rel styrene caps have lower series R than other Rel styrene caps. I usually use the high series R caps, because I prefer their natural damping characteristics. However, if I were decoupling a switching power supply, I would use the lower series R cap. All caps look inductive above their resonant frequency.
john curl said:
On top of my head, 4MHz resonance with 0.1uF maps to an ESL of 16nH. Considering the rule of thumb of 25nH/in of conductor this makes for .64" terminal and PCB trace length. It would take a really poor PCB design to reach this value for ceramics, in particular multilayers.
OTOH, film caps usually have a significant internal ESL, due to the rolled foil construction. This makes reaching those 16nH (internal inductance plus terminals plus PCB traces) much easier.
Bottom line, film caps are inferior when it comes to self resonance. This applies to SMD parts as well, although the resonance frequencies are much higher.
I still fail to see the importance of DA in power supply filtering caps.
Please look at the Tecate Group Surface Mount Multilayer Ceramic data sheet, type CMC, for .1uf cap. .0603 case. Please let me rub your nose it it!😀
john curl said:
I'm looking.
http://www.tecategroup.com/capacitors/datasheets/smd/ceramic/CMCSTD.pdf
X7R multilayer 100,000pF self resonates at about 20MHz. I hope you don't use Y5V ceramics, those are much worse than aluminum electrolytics, they have -90% capacity dependence on bias.
You point is?
- Status
- Not open for further replies.