NP0 / C0G long term stability

"C0G (NP0) is the most popular formulation of the "temperature- compensating," EIA Class I ceramic materials. Modern C0G (NP0) formulations contain neodymium, samarium and other rare earth oxides. C0G (NP0) ceramics offer one of the most stable capacitor dielectrics available. Capacitance change with temperature is 0 =B130ppm/=B0C which is less than =B10.3% =C4 C from -55=B0C to +125=B0C. Capacitance drift or hysteresis for C0G (NP0) ceramics is negligible at less than =B10.05% versus up to =B12% for films. Typical capacitance change with life is less than =B10.1% for C0G (NP0), one-fifth that shown by most other dielectrics. C0G (NP0) formulations show no aging characteristics. The C0G (NP0) formulation usually has a "Q" in excess of 1000 and shows little capacitance or "Q" changes with frequency. Their dielectric absorption is typically less than

0.6% which is similar to mica and most films."

Thanks, that is more than I found. I am still not sure if the actual likely life stability is a lot less than that, or if that figure is just a limitation of the measuring method.

The highest lifetime-stability surface mount parts I know of are the Vishay metal foil resistors, at ~$10 each. These have a 100ppm lifetime stability. I was wondering how a C0G capacitor would compare, since my application could use either.

--

John Devereux

5=C2=B0C

Manufacurers like AVX and Johanson flat out state no aging whatsoever for this dielectric class. It's not the case for higher grade relative permittivity types but it's a completely different technology of dielectric so no extrapolation is possible.

Experimentally, the Susumu thinfilm resistors (more like 30 cents each) drift less than 10 PPM per year, maybe much less. We make a

16-channel DAC board, where each channel uses an ADR421 and a dac and 5 Susumu's to define the output, and all the channels hold 10-15 PPM per year.

I'd guess that C0Gs are in that range, too. I've done LC oscillators that drift PPMs per year, and I've had more troubles with inductors drifting than with caps.

John

But that extract you quoted was from an AVX datasheet wasn't it? That's what came up anyway. Or is there some distinction between "aging" and "capacitance change with life"?

Yes I am specifically talking about NP0/C0G rather than X7R etc.

--

John Devereux

Thanks, that is encouraging. The datasheets show up to 1000ppm for various type of drifts though.

--

John Devereux

You may be looking at the wrong data sheets or data sheets from an inferior manufacturer. Please provide a link to them.

[...]

I was looking at these:

--

John Devereux

I apologize. I thought the thread was still on capacitors. My mistake.

Well don't do that again!

Jamie

Hey, no problem.

Really, I should probably use the susumus, 1000ppm is just about acceptable and it seems that I likely will get much much better in practice.

I just found it an intriguing possibility, with references to "zero", "not measurable" drift of a 0.1 cent part, and having to spend $10 to get a 100ppm resistor. That's 10000 times more expensive!

--

John Devereux

Well, ceramic materials are man-made rocks. There are surface rocks that are as old as 4.5 billion years, so the 'aging' of a ceramic capacitor will depend, in the long term, mainly on the metal/wire/coating parts of the gizmo. Those are the same whether it's NP0 or any other formulation.

Quite logical, perhaps sensible, but completely wrong.

Interesting that reflow soldering will "de-age" the capacitor, thus maximizing the capacitance (and the drift) immediately after assembly!

So, if your design depends on, say, X7R capacitance, you can improve stability over a typical product life considerably just by waiting a few days after soldering before tuning it.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

That *is* interesting - unfortunately that reference does not help me since it was NP0 I was interested it. Which they say "are stable over time, temperature and voltage". Encouraging, but *how* stable?

--

John Devereux

Er... no, reread that article: it refers to spontaneous polarization of ferroelectric materials, and those aren't part of any NP0 capacitor. Ferroelectric materials are useful, but nonlinear, and they can polarize for lots of reasons (not just age, but time-under-bias, I suspect).

In other words, after the first million years, most of them are pretty stable. ;)

Hexagonal BaTiO3 is metastable, so in high-K ceramics you can get drift due to recrystallization. NPO/C0G ceramic doesn't have that problem.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

You said " Those are the same whether it's NP0 or any other formulation". X7R and Y5U are such.

That leaves open the possibility that the OP is worried about- that there may be small residual aging effects.

A few years ago (okay, twelve!) John Woodgate mentioned that such effects can be used to date when ancient pottery was fired.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com