Understanding the temperature coefficient vs drift of ceramic capacitors

Hi,

a few weeks ago I have added a topic concerning the choice of capacitors with low TCC → Looking for a low TCC capacitance

From articles that have been posted in the topic above, that explain the code letters of ceramic capacitor classes, I have come to the point where I am not sure, if I understand the difference between TCC and drift.

TCC:
My understanding of the temperature coefficient until now was, that the TCC is a value in ppm/K (or ppm/°C) that stands for the change of the capacitance when changing the temperature. So for example a NP0 would have a TCC of 0±30 ppm/K.

Drift:
The drift of the capacitor is a value, that determines how much the capacitance changes over time.

According to the graphic above, the value in the second column (Temperature coefficient a 10-6/K) is chosen as drift while the third column seems to be the TCC.

Can someone clear things up?

Additional question: if I want to calculate uncertainty (worst case calculation), does this table in the graphic above contain all information that I need regarding the component?

Thanks in advance!

The temperature coefficient (2nd column) describes variation in observed capacitance with respect to temperature, the tolerance (3rd column) describes the margin of error that applies to the indicated temperature coefficient. For example, an NP0 capacitor can be expected to have a TCC of 0±30PPM/°C. “Drift” usually is the term used to refer to change in some component characteristic over time, but is not mentioned in the above table.

The table above describes industry standard classifications or groupings, rather than specific behaviors of any any particular component. Within the limits of the designation, there may be variation in the precise behavior of a part from supplier A vs supplier B.

If capacitance stability is of primary concern, the NPO/C0G variety are about the only ceramics worth considering along with their military equivalents. Skip the class II dielectrics (e.g. X7R) as they are subject to aging and voltage-induced changes in capacitance. Mica devices are also generally quite good. If one needs higher values, the polymer film caps will likely offer the best stability; looking at anything electrolytic would be a waste of time.

The easy route here would of course be to go and buy oneself a capacitance standard of the sort used for calibration purposes, which are specifically designed for and characterized in terms of stability.

2 Likes

Thanks! This helps a lot. I was considering heavily using class II ceramics for higher capacitances (1uF, 10uF and 100uF) but my main concern is stability. The final capacitance is not of great interest, since I can measure the capacitance with sufficient precision → the manufacturers tolerance does not have an impact on my module.
Could you suggest any polymer film capacitors in the scales of 1uF, 10uF and 100uF.
I was looking for such capacitors but havent been successful so far.
Any help greatly appreciated!

Here are some options for film capacitors in the 0.1uf, 1uf, 10uf and 100uf values.

Hi Steve, thanks for your post.
Do these caps have (extra) low TCCs? The stability (temperature/aging) is the my biggest concern regarding the new capacitors.

Polypropylene film caps are probably the best bet; where characterized, typical figures land around -200PPM/°C and 1% drift over a few years. If a given product is not characterized in terms of these behaviors, it’s not characterized in terms of these behaviors.

Capacitors exist on a quantity-quality spectrum; the more C one desires in a package, the poorer that C generally will be in terms of precision, stability, etc. It’s a curve defined by the state of technological development and market economics; if you want something that’s off the curve it’s likely available somewhere–for a price. Again, if you want the best possible stability, look at products specifically designed to offer the best possible stability, e.g a capacitance standard.

If cost is a concern, you’re not likely to do much better than the figures for polypropylene film dielectrics mentioned above in 1uF+ territory.

1 Like

Thanks, that’s something I can work with!