Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Jul Wed 30, 2014 6:11 pm


All times are UTC [ DST ]



Post New Topic Post Reply  [ 12 posts ] 
Author Message
 Post subject: C0G, NPO Ceramic caps vs. dipped mica
PostPosted: Jan Sun 02, 2011 6:49 pm 
Moderator
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 16357
Location: Somers, CT
For discussion/info, and hopefully no arguments, this comparision was found on the Clifton Labs website:

http://www.cliftonlaboratories.com/capa ... change.htm

Pete


Top
 Profile  
 
 Post subject:
PostPosted: Jan Sun 02, 2011 8:23 pm 
Member
User avatar

Joined: Sep Mon 27, 2010 3:30 am
Posts: 244
Location: Calhoun, TN
From my limited knowledge base, I see nothing to argue about concerning methodology, but DANG, who knew?! I now understand why Doug DeMaw was such a fan of NPO caps.

_________________
Smokestuffer


Top
 Profile  
 
 Post subject:
PostPosted: Jan Sun 02, 2011 10:16 pm 
Member

Joined: Jun Thu 15, 2006 1:21 am
Posts: 4021
Location: NE Ohio
Thanks Peter.

_________________
Bruce


Top
 Profile  
 
 Post subject:
PostPosted: Jan Sun 02, 2011 11:01 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 7578
Location: Minnesota
The shortcomings of the ceramics have been known for quite some time, the article fails to test for the piezoelectric effects that ceramics also have. They are fine for limited use in RF circuits,but I would not use ceramics unless there is no other choice.

Dave


Top
 Profile  
 
 Post subject:
PostPosted: Jan Sun 02, 2011 11:09 pm 
Moderator
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 16357
Location: Somers, CT
Piezo effect is caused by the use of Barium Titanate in the dielectric.

http://my.execpc.com/~endlr/ceramic.html

I've never seen anything to show that this a real concern in ceramic caps with low dielectric constants for RF use.

Personally, and I know others disagree, I avoid use of ceramics for coupling in high impedance audio circuits myself. Piezo effects aside, poor temperature stability and the fact they act like varicaps are other issues.

Peter


Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 12:11 am 
Moderator

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21016
Location: Detroit, MI USA
Interesting that some supposedly "high end" audio gear from the 60's like McIntosh preamps used disc ceramics in various functions. A friend has one, and some of the disc ceramics are seriously microphonic, if you tap on them they ring like a tube. I would have expected the engineers at McIntosh to know better. Maybe they figured no one would be tapping on the caps. :shock:

A well documented problem with piezo effect at non-RF frequencies happens when disc ceramics are used in the sweep circuits of 7" electrostatic TV's. In some sets they work OK, but in others they create a distortion, sort of a pie-crust or jagged edges on objects in the picture. In still other cases, non-linearity results from substituting disc ceramics. :shock:

One portion of the article shows how harmonics result from using disc ceramic caps, which could be related to the piezo effect. Overall, very interesting reading.

_________________
Dennis

Experience is what you gain when the results aren't what you were expecting.


Last edited by Mr. Detrola on Jan Mon 03, 2011 12:16 am, edited 1 time in total.

Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 12:15 am 
Moderator
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 16357
Location: Somers, CT
Hi Dennis

I know that ceramics can't be used in some sweep circuits because of the peak voltages...

But I wonder if some of those issues were related to bad design, or a failure mode in the part itself?

It would be a neat experiment to compare the difference in harmonic distortion on a jhigh performance hi-fi amp swapping between ceramic disks and mylars for the high impedance interstage coupling caps.

Pete


Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 12:17 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 75
Location: Colorado Springs
Unless I missed it, I didn't see any numbers for the *actual* capacitance at 0V for the parts - the graphs show a normalized change. But some of the cheap parts may have quite high rated tolerances, right? If so, the data becomes "not as bad" in some respects. Of course, it's still not good if the value changes, when you don't want it to change.

Some takeaways would seem to be:
1) If using cheap caps, use voltage ratings that are much higher than what you really need. I'm assuming that this minimizes the negative features of the cheapos.
2) Understand which types of circuits really need specific, stable cap values.
3) Understand how much margin you have, assuming the cap is going to change value.
4) Be careful in circuits where the applied voltage on the cap is going to swing widely in normal operation. Someone mentioned audio coupling caps - in a lot of circuits (not all), the cap has a continuous high voltage applied, and the signal is very small by comparison, so I would not be terribly worried about phase distortion and such. And it occurs to me that this could be a good design point also, for stuff like hi-fi audio: forcing the cap to always operate at a DC potential that is much higher than the signal may work in your favor.

Pete


Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 12:35 am 
Moderator

Joined: Jan Thu 01, 1970 1:00 am
Posts: 30984
Location: Livermore, CA
Ceramic caps hold up well in circuits designed for them.

I tested value vs temperature and was sure surprised at the change. At first heat caused capacity to increase. Further heating caused capacity to drop to 1/10 the value.

I used a modern ceramic cap, square shape rated .1 mf @ 100 volts. Value ended up around .01 mf with enough heat. Not a controlled experiment but interesting.

Even with this great change in value cap would still do it's job as RF bypass.

_________________
Norm


Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 2:27 am 
Moderator
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 16357
Location: Somers, CT
Curiousity got the best of me.

Supposedly the capacitance changes with the applied voltage, and can cause harmonic distortion when using caps with high K factors.

I did a quick bench test setup, using my HP 654A oscillator and HP 333A harmonic distortion analyzer.

The test jig included a high quality audio step up transformer with .33 uF mylar DC blocking caps used as needed. A 120 volt, well-regulated DC source was used for a bias voltage across the test cap, feed through a 6.8K resistor to approximate a low impedance source. A 2.2 meg resistor was used on the other side of the test cap for DC return and hi-Z termination. Audio level was 60 volts P-P riding on the 120 volt DC bias voltage.

The noise and/or harmonic baseline, with no cap (direct wire connection) or DC voltage was down 54dB. With a .01 mylar the noise level stayed at -54dBm. Subbing a new 0.01 uF cap purchased in the past six months showed no measurable change in harmonic distortion. All noise and harmonics still down -54dB relative to the 1kHz signal. It might be beneficial for someone else to verify these results, my test procedures may be flawed. I'm at the limits of my test equipment's abilities.

Sharply rapping the cap, using an insulated plastic screwdriver handle, produced noise spikes that were at best around 3mV PP max.

Pete


Last edited by Peter Bertini on Jan Mon 03, 2011 2:53 am, edited 2 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 2:51 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 75
Location: Colorado Springs
I don't know much about distortion measurement, but perhaps adding some inductance in the test jig is necessary, to maximize the effects of the changing capacitance?

Pete


Top
 Profile  
 
 Post subject:
PostPosted: Jan Mon 03, 2011 10:51 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1558
Location: Saskatoon
At 1kHz the 0.01uF capacitor will have a reactance of about 16k. Much less than the 2.2 Meg load that it's feeding into. So you won't see much distortion. It you reduce the load resistance to something less than 100k, then you may be able to measure something.


Top
 Profile  
 
Post New Topic Post Reply  [ 12 posts ]  Moderator: Peter Bertini

All times are UTC [ DST ]


Who is online

Users browsing this forum: Google [Bot] and 1 guest



Search for:
Jump to:  



















Privacy Policy :: Powered by phpBB