Electrolytic capacitors at low temperature

[Simon]

I'm looking to design a circuit that needs to work down to -46C (oh how many times have I said that). So once I'm over hurdles of semiconductor devices that won't go under -40C my next concern is capacitors.

I think technology has changed much recently for capacitors so I'm not sure if what I find online is still relevant. So far from what I gather aluminium electrolytics are infact the best performing for low temperatures and even ceramics may not be a good idea.

Looking at parts on Farnell the datasheets seem to just say that they work from -55 to +105 with no information about characteristic changes over temperature.

[Psi]

Can you add a little heating element to keep the case internal temp above -20.
That would make it a lot easier to find components.

[wraper]

Most noticeable is significant ESR increase with lowering the temperature.

[Simon]

I'm going to try and do that, what I am trying to understand at the moment is what my options are and what happens at extremes.

[rob77]

i would definitely go for a hermetically sealed (because of moisture) case with thermal insulation and a heating element inside. and as i'm a bit paranoid, i would hermetically seal the case at 20celsius in a room with dry air. or even more paranoid and install 2 valves on the case and flush it with nitrogen

[Rerouter]

went off and had a read for you, Vishay state that most of there -40 rated capacitors can be used all the way down to -65 (i would assume the electrolyte solidifies there) but that the ESR dramatically increases towards the low end, but as an added bonus the leakage current drops dramatically, also found the below link with a crude calculation for the ESR vs Temp,

Furthur reading drags back up the old lifetime calcs, the lower temperature should mean the caps last multiple human lifetimes, but equally digs in that below some point between -20 and -150 the capacity of the cap will drop (to about 10% capacity at -150C for one example)

http://www.edn.com/design/components-and-packaging/4327019/ESR-calculations-for-electrolytic-capacitors-at-lower-temperatures

there is a perk with the ESR increase, that if the signal is AC, it will cause self heating which will decrease it a little (this actually implies capacitors have a crude form of load balancing :/)

[lewis]

I had a feeling you were going to ask about caps!

Use OS-CON solid electrolytic caps. Normal electrolytics increase in ESR significantly below -20 or so, but OS-CONs remain relatively flat. There isn't much info out there, but I found this: http://www.element14.com/community/groups/panasonicindustria/blog/2014/01/23/technical-advantages-of-panasonic-polymer-capacitors-1-stability-over-temperature-changes

Search Farnell for OS-CON.

Capacitance reduces too at low temperatures, by as much as 40% at -40C. If you do need to use a wet electrolytic (maybe you need a high capacitance or high voltage), parallel two of them to ensure you get the capacitance you expect.

One thing to watch out for with os-cons is the low ESR! It can be too low - they can cause instability in some circuits. Treat them like an MLCC, and test test test.

Or, use X7R ceramics.

[Psi]

+1 for OS-CON

They make some awesome caps.

[Simon]

I'll go for os-con caps, looks like digi-key is now my friend, farnell have 2 only and they are being "run out of stock", I have a similar problem for atmel chips, farnell don't even include temperature range in the parametric search which is pants and high temperature grade models of chips are also harder to come by.

[T3sl4co1l]

Polymer (not just os-cons, everyone makes them)
Tantalum (assuming current is limited so as to avoid China Syndrome)
There are also Ta-polys available, don't know much about them.  Presumably they don't burn up as dramatically as others.

Most ceramics, I think, freeze out at low temperatures?

Tim

[Simon]

ok, well farnell don't have all that I want anyway so I'll have to get some stuff from digikey or even mouser, both outstrip both RS and Farnell for range and stock availability and lead times.

[David Hess]

In the past when working at low temperatures my solution has been to use solid tantalum capacitors but now I would consider solid polymer electrolytics as well.

[SeanB]

Use wet slug tantalum in a glass seal package. They are rated for the full aerospace temp range, and will work well at -55C. Otherwise polyester caps for low values, and watch out for ceramics that are very bad at low temperature.