What's the most temperature stable capacitor?

[Nemo]

I've been asked to design a low pass filter, but, it's to work over the
range -40C to +65C (maybe +75C). The frequencies of interest are around
2Hz so it looks like I'll need capacitors in the range 1uF to 4u7. Oh,
and it's meant to be small so surface mount components would be best
(but I'd be happy to sacrifice that requirement if necessary).

What's the best kind of capacitor to use here? I don't think the usual
dielectrics - X7R, etc - are characterised to -40C and I'm sure they
will be 20% or more out by that temperature. Are there more exotic
capacitors available, perhaps from non mainstream suppliers, for this
kind of application, that might reach 1uF?

Thank you

Nemo

 

[Stig Carlsson]

"Nemo" <Paul@nospam.nospam.nospam.nospam.co.uk> skrev i meddelandet
news:hZyWmkBmn64GFwm+@furfur.demon.co.uk...

I've been asked to design a low pass filter, but, it's to work over the
range -40C to +65C (maybe +75C). The frequencies of interest are around
2Hz so it looks like I'll need capacitors in the range 1uF to 4u7. Oh, and
it's meant to be small so surface mount components would be best (but I'd
be happy to sacrifice that requirement if necessary).

What's the best kind of capacitor to use here? I don't think the usual
dielectrics - X7R, etc - are characterised to -40C and I'm sure they will
be 20% or more out by that temperature. Are there more exotic capacitors
available, perhaps from non mainstream suppliers, for this kind of
application, that might reach 1uF?

Thank you

Nemo

One of the stable ceramic dielectrics is COG which has a temp. coefficient
<30 ppm/C but you will not get near 1uF.
Maybe you can mix some some plastic dielectrics like polyester and
polypropylene. This is not SMD!
Polyester is +400(+-200) ppm/C at 1 kHz, polypropylene -200(+50,-100) ppm at
1 kHz according to the Evox-Rifa site.

Try 1/3 polyester and 2/3 polypropylene and you might get something you can
use.
The capacitance will change over time but if you temperature cycle the
capacitors from lowest to highest a few times before you use them they will
stay more stable.

Best regards
Stig Carlsson

 

[Roger Hamlett]

"Nemo" <Paul@nospam.nospam.nospam.nospam.co.uk> wrote in message
news:hZyWmkBmn64GFwm+@furfur.demon.co.uk...

I've been asked to design a low pass filter, but, it's to work over the
range -40C to +65C (maybe +75C). The frequencies of interest are around
2Hz so it looks like I'll need capacitors in the range 1uF to 4u7. Oh,
and it's meant to be small so surface mount components would be best
(but I'd be happy to sacrifice that requirement if necessary).

What's the best kind of capacitor to use here? I don't think the usual
dielectrics - X7R, etc - are characterised to -40C and I'm sure they
will be 20% or more out by that temperature. Are there more exotic
capacitors available, perhaps from non mainstream suppliers, for this
kind of application, that might reach 1uF?
Use smaller components, and synthesise the required value. Possibly use a

gyrator, and an inductor.
Or go for a complete active filter design. The most _accurate_ solution,
will be to do this using a DSP (at this frequency, you could even
potentially use quite normal processors, rather than DSP designs). With
crystal timing, it'll beat anything possible with pssive components.
You are not going to get small accurate bipolar capacitors. Even if you
sacrifice the 'size' requirement, the cost will be silly....

Best Wishes

 

[Hal Murray]

In article <hZyWmkBmn64GFwm+@furfur.demon.co.uk>,
Nemo <Paul@nospam.nospam.nospam.nospam.co.uk> writes:

I've been asked to design a low pass filter, but, it's to work over the
range -40C to +65C (maybe +75C). The frequencies of interest are around
2Hz so it looks like I'll need capacitors in the range 1uF to 4u7. Oh,
and it's meant to be small so surface mount components would be best
(but I'd be happy to sacrifice that requirement if necessary).

How sharp a cutoff do you want?

Can you do it in the digital world? (Quartz crystals are pretty
stable over temperature.)

--
These are my opinions, not necessarily my employer's. I hate spam.

 

[Nemo]

Stig Carlsson suggests
....

Try 1/3 polyester and 2/3 polypropylene and you might get something you can
use.

Comprehensive answer with nice plot twist at the end - thank you -
sounds extremely useful!
--
Nemo

 

[Tilmann Reh]

Nemo schrieb:

Stig Carlsson suggests
...
Try 1/3 polyester and 2/3 polypropylene and you might get something you can
use.

Comprehensive answer with nice plot twist at the end - thank you -
sounds extremely useful!

Also have a look at the MKM capacitors by WIMA - they are made of such a
mixed dielectric, especially for (relative) temperature independent
capacity.

Tilmann

--
http://www.autometer.de - Elektronik nach Maß.