Does an electrolytic capacitor work in vacuum?

[aramis]

Electrolytic capacitor in vacuum

I have to design a printed circuit board for a cubesat (small educational 1 kg satellite). The circuits on this PCB have a few capacitors from 10 uF to 2200 uF.

Question: Do electrolytic capacitors work in vacuum? If not what alternatives do I have?

[Time]

They might work but I doubt they put electrolytics in satellites in the first place.  Tantalum is an alternative that might work for you though they are slightly more expensive.

[Alex]

They will work if at the right temperature.

Check the standards in the aerospace industry (NASA does their own standards too), I am sure it is very tightly regulated and you are forced to use specific types.

[aramis]

The price is not a problem. They can be 3-4 times even ten times more expensive than ordinary (electrolytic) capacitors. I am just looking for some capacitors that can work in space and also have a volume as small as possible because those cubesats are so tiny (in comparison with ordinary satellites) that any cubic centimeter of their volume is valuable space.

[Time]

Look at tantalum capacitors than.  Liquid tantalums will have the best size/capacity ratio but they will probably be more temperature sensitive.  I think you are good on the pressure side for both tantalum and electrolytic.

http://www.radio-electronics.com/info/data/capacitor/tantalum-capacitor.php

[aramis]

So tantalum capacitors can withstand temperatures in the range [-60, +100 C] and also there is nothing inside which might evaporate in vacuum?

[Time]

So tantalum capacitors can withstand temperatures in the range [-60, +100 C] and also there is nothing inside which might evaporate in vacuum?

Yes, they have very wide temperature operating range.  You can read about them in that link I modified into my previous post.

[Simon]

For temperature you can check the datsheet, you might have large variances in capacitance with temperature change

[BrickBoiler]

I think if I was working on something like this I would start looking for a vacuum chamber to test things in. I've never done anything for space but I assume heat rejection is also a huge problem.

[DaveW]

I do aerospace stuff, not quite vacuum but rated to 70,000 feet, -55 to 85 deg C. We've tried electrolytics every now and then, including the higher rated aerospace ones and found them to be unreliable compared to solid tants or even liquid tants. How much capacitance and what voltage rating do you need? A tant is going to be pretty small, is pricey though,
http://uk.farnell.com/avx/tln6338m004r0055/capacitor-case-6-4v-3300uf/dp/1865010

[Zad]

I'm surprised there isn't a big spec document for this, but all components that "fly" will have to be explicitly space rated. Outgassing can be a big problem, not just for obvious things like electrolytic caps but from things like adhesives, paints etc. With conventional electrolytics, I imagine the component body size could be a problem in the shake tests, with high Gs and high jerk pulling the components off the PCB. 2200uF seems an awfully big capacitor for a tiny satellite.

[saturation]

Its more than just vacuums.  Parts have to space rated, or MILPSEC, at the least.  I don't design at this level, but I know a little about it because I was curious by the how high the cost of military parts were, and now and then some board would appear in a surplus parts bin in someplace like Apex Electronix, often its just aeronautic grade, you can get to play with them and see just how different they can be.

Cost tend to be 10-1000x the cost of a commercial part, i.e., 10c cap, becomes $10.

Electrolytics last I recall are general avoided for space and military applications.  Besides temperature, there are many other parameters, key of which is various types of radiation.  For mounting and assembly, shock and vibration ratings have to be highest, like warheads in missiles, given the nature of lift off to simply deliver the payload into space creates high Gs that are later dropped to zero by microgravity.  Finally operational reliability requires substantial derating and a lot of cross checking, kind of like underclocking a CPU rather than overclocking it, due to effects of radiation on switching circuits.

http://nepp.nasa.gov/files/13779/Liu-Aluminum%20Capacitors.ppt


    * MIL-STD-975, published by NASA, focuses on selection of parts used in the design and construction of space flight hardware as well as mission-essential ground support equipment.
    * MIL-STD-1547, published by the Department of Defense, is targeted to aid in the design, development and fabrication of electronic systems with long life and/or high reliability requirements while operating under the extreme conditions of space and launch vehicles.
    * AS4613, published by the U.S. Navy, sets forth derating requirements for the reliable application of electronic and electromechanical parts.
    * NAVSEA TE000-AB-GTP-010, published by the U.S. Navy, contains derating requirements and part selection and application information on the ten most commonly used electrical and electronic parts.
    * ECSS-Q-30-11A, prepared and maintained under the authority of the Space Components Steering Board in partnership with the European Space Agency, contains derating requirements applicable to electronic, electrical and electromechanical components.
    * MSFC-STD-3012, prepared by NASA's Marshall Space Flight Center, sets requirements for electrical, electronic and electromechanical parts selection, management and control for space flight and mission-essential ground support equipment for Marshall Space Flight Center programs.

[Lance]

What restrictions on size and weight do cubesats have?

[tecman]

There are hi-rel mil spec'ed electrolytics available.  The preferred are wet slug tants, but they are limited in value to generally under 1000 uf.  Vacuum will not be a problem, since most are sealed, but the temp range needs to be looked at carefully.  Additionally life will be a consideration, so how you are using them (ripple current, etc) will be a factor for life.  Some apps I have seen used NiCads as capacitors in certain applications.  Higher ESR than lytics, but more current reservoir.

paul

[Zero999]

Is it possible to put the PCB in a little enclosure and fill it with potting compound? Weight might be a problem.

Agreed, you should use tantalum capacitors in applications like this, they're more expensive but definitely worth it.

[DaveW]

With conventional electrolytics, I imagine the component body size could be a problem in the shake tests, with high Gs and high jerk pulling the components off the PCB.

If you're going to experience high g's you'll have to worry about this even with surface mount components. If you're using components with fewer pins (resistors/caps as opposed to soics, etc.) try and avoid mounting them near the edge of the board. For example, I've seen acceleration spikes of 400g approx for very short times (nanoseconds) ripping 0603 ceramic caps and pads off a printed circuit board. Potting the board will help to avoid the track breaking, but won't stop it pulled away from the board

[kaptain_zero]

I think Cubesat.org will be your primary information provider and or can direct you to the correct information you seek.  A quick quote from one of the compliance sheets goes like this:

Note: NASA certified materials should always be used in space flight hardware, especially epoxies and glue. If in doubt about the materials you are using, please contact the CubeSat Coordinator. Thermal-vacuum bakeout must be performed on fully integrated flight CubeSats before integration into the P-POD.

I used to run a SatGate in a previous life, but I never got involved with the construction of amateur satellites, so I am NOT a good source for info, but I am good at ferreting information when I need it. Amsat.org directed me to Cubesat.org and they in turn have links/docs for Developers. Make sure you get the right information as *anything* that does not meet the specs required will result in your project not getting a ride and there is NO LEEWAY in this area. Millions of $$$ ride on each launch... and nobody is going to take a chance on *anything* not certified for use.

Regards

Christian

[xygor]

I wouldn't say millions of dollars ride on the success of the project.  More like 10's or hundreds of thousands.  That's one major advantage of the cubesat.  Pumpkin seems to have flown a lot of hardware.  It looks like there are some tantalums on the obsolete board, but not on its replacement.  (Just looking at the pictures.)  I don't know if that is significant or not. http://www.cubesatkit.com/content/datasheet.html

[kaptain_zero]

Yes, it can look cheap, but one forgets (and I noticed that it's very hard to ferret out the actual total cost of a launch) that while cubesats themselves are cheap, collectively there's a bunch of them together in a launch scenario, and they are all sharing the cost of the launch vehicle/related costs. Often, they will be hitching a ride on a commercial launch by Nasa, ESA or some such, and the primary payload and launch vehicle will most certainly be in the many millions of dollars, and it's these primary players (and their insurance companies) who will insist that the cheapo cubesats, tagging along as ballast, are safe and conform to all the required standards.

Take for example ESA's Ariane 5 which has been used for Amateur Satellite deployment in the past (again, the Amateur sats were going as "ballast").  It's estimated that the launch costs of such a vehicle is around US$ 120M. Not exactly chump change and that does not include the costs of the primary and secondary payloads.

Regards

Christian

[aramis]

First of all, thank you for your suggestions.
Secondly, a cubesat (1 kg, 1 cubic decimeter satellite) has to be built using commercial off the shelf components. Space rated parts are not encouraged because they are too expensive and hard to find. Most cubesats successfully tested in orbit around the Earth do not have space rated or military certified components.

[saturation]

Wow, that's interesting:

http://en.wikipedia.org/wiki/CubeSat

Papers by NASA or IEEE say commercial available components are attractive because of lower costs and ease of access:

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5560227

and for over a decade, NASA has developed a testing program to evaluate COTS for space use, rather than going to 'military' grade listed parts.  Cubesats can also be used for post mortem of parts survivability.

But I guess it depends on the life expectancy of your device, shielding and design will reduce the effects of radiation but not
mitigate it fully, causing a slow decline in system life.  But getting to space within reach of individuals is peachy!  Now that's design fun.

http://www.spacemicro.com/news/pr/2010/2010_11_9_UHF_Cubesat_Antenna_Awarded_to_Space_Mirco_from_SPAWAR.html

http://education.ksc.nasa.gov/esmdspacegrant/LunarRegolithExcavatorCourse/Chapter3.htm

http://en.wikipedia.org/wiki/List_of_CubeSats

Success rates just at eyeball is 50-50 here.

Go for it.

First of all, thank you for your suggestions.
Secondly, a cubesat (1 kg, 1 cubic decimeter satellite) has to be built using commercial off the shelf components. Space rated parts are not encouraged because they are too expensive and hard to find. Most cubesats successfully tested in orbit around the Earth do not have space rated or military certified components.

[kaptain_zero]

For those interested in low cost launch details.... I found this 2004 article, which may be somewhat dated now, but still interesting to read: http://www.thespacereview.com/article/233/1

At the start of this thread I was briefly tempted to build another ground station for tracking these cubesats, but alas.... I just don't have enough sleep time to donate to such a project anymore....  Over a 10 year period in the 90s, I spent more nights into the 2AM to 3AM range, tweaking, troubleshooting and repairing my SatGate, only to grab 3 to 4 hrs of sleep and then head back to work for 7am than I care remember....  There still is a hollowed out section in the basement concrete wall that perfectly matches my forehead...... Beta testing software can be soooo much fun....... It's the mop and bucket time for cleaning up all the bits and bytes after a a major computer/beta software crash that sucked as my fully automated SatGate had to be operational 24/7.

And then there was the joy of trying to keep an AZ/EL antenna array working flawlessly when night time temperatures plummet into the -40C's range.... Most greases used in rotator gearboxes become solid at that point. I did find a grease that would work in the winter AND keep working when temps hit mid to high +40C's during the summer, but there would eventually be that day in the winter, when I had to crawl up the tower and wrap battery heater blankets around the rotators to warm them up so they could operate without burning out the motors. Ah such fun...... NOT.   


Regards

Christian

[Frangible]

Way back when I worked on a satellite that did indeed use electrolytics.  It worked for years before the nicad batteries finally gave way.  It was supposed to be a cheap satellite, using industrial grade components.  The original plan was to package everything in a pressurized structure, which would avoid outgassing plastics and bursting capacitors.  It turned out NASA was far more concerned with pressure vessels on board the Shuttle, so it was launched sans air.  Worked like a champ.

For cubesats, I would avoid them - they are kind of bulky for what you get, and they may be shaken to bits during ascent.  Rockets produce lots of strong vibrations, you'd be amazed at what breaks loose during shake tests.  Never mind potting - you do everything you can to make your systems as light as possible.  Weight == fuel == $.

[Trigger]

You have to get your CubeSat approved for launch which involves a lot of testing.
http://www.cubesat.org/images/developers/cds_rev12.pdf
The document lists all the requirements for materials and design.

When you're putting something in space always try to go solid state with everything as they have less of a chance for failure and design in resiliency and redundancy.  Though the best thing you can do is build a board, buy a vacuum chamber and stick it in there at full vacuum for a while.

I'm actually working on one myself though it'll be a high altitude balloon payload a few times to make sure communications and such are solid before it gets launch ready.  That and the cheaper testing gives something to show when you're trying to get funding to actually launch it.

[Lawsen]

Yes, electrolytic capacitors work in vacuum, because the dielectric material like aluminum foil, cardboard spacers, and electrolytes are not air dependent to be functional.  The absence of air might be better as no source of oxygen for the oxidation-reduction process to take place.  The forum members brought some important issues, temperature and mechanical vibration.  Space is extremely cold or hot, never an ideal place; never want to go there.  One concern is the mechanical shock and vibration from the G forces during the launch inside the rocket.  Dave Jones brought some good point about an inductor coil breaking off during his test of the Fluke 28-Series 2.  Your circuit needs to be well made to withstand the vibrations and gravity forces during launch.  The pdf for NASA TM X-64755, revision A, Title and subtitle is "Guidelines for the Selection and Application of Tantalum Electrolytic Capacitors in Highly Reliable Equipment" by Dr. A. M. Holladay in 1978 is available by download.  You could download it here:

http://www.everyspec.com/NASA/NASA+(General)/NASA_TM_X-64755_4768/

European Space Agency have their own web site for capacitor standards along with the Russian Space Agency Venus Venera programs, 1961 to 1984:

http://en.wikipedia.org/wiki/Venera

https://escies.org/ReadArticle?docId=627

You could down load the pdf and chose the capacitor information suitable for your product or project.  All are in English.  

You could contact the Russian Federal Space Agency for capacitor specifications by e mailing them:

http://www.federalspace.ru/main.php

Venera-D station:

http://www.federalspace.ru/main.php?id=2&nid=9948&hl=venera

Ejection seat designs with BF Goodrich in the U.S.:

http://www.zvezda-npp.ru/engl/k93.html

It is most likely in accordance to the European Space Agency's pdf specifications as the Europeans aerospace companies are often trying to standardized for commerce and business.  

French Space Agency ceramic chip, fixed type one, capacitor specifications:

http://www.avx.com/docs/certificates/ESCC%20St%20Apollinaire.pdf

Lawsen