Why not seal super capacitors?

[adjustttt]

This is probably an extraordinarily stupid question but if the longevity of super capactiors is limited by the electrolyte evaporating why can't they just be sealed off completely without evaporation they would last piratically forever right?
I am probably missing something very big here 

[glarsson]

It is difficult to make a perfect seal and keep the cost low.

[adjustttt]

So that means I could seal them completely myself?

[wraper]

So that means I could seal them completely myself?

Why do you think that with your added seal it will perform significantly better? If you think that covering it with silicone will completely prevent very slow electrolyte escape, you are wrong.

[adjustttt]

Why is that wrong?

[wraper]

Why is that wrong?

Why do you think that silicone cannot pass it through? If you use hard compound, it could even make things worse by adding mechanical stress to original seal.

[Brumby]

Why is that wrong?

What would you use as a sealant?  Silicone?  Glazed ceramic?

Silicone is good at keeping water out - but what about other substances?  It needs to cure to set - which happens when exposed to air.  If it was a perfect sealant, then only the surface would cure and the internal mass would remain uncured.

A solid glazed ceramic would have a better chance - but the capacitor isn't going to survive the process too well.  Even if it did, were there any expansion issues with the capacitor, you would get anything from a hairline crack (say goodbye to being sealed) to an explosion (say goodbye to the cap).

Maybe someone can tell us what caps NASA used on Voyager I and II.  They seem to be doing ok.

[wraper]

Silicone is good at keeping water out - but what about other substances?  It needs to cure to set - which happens when exposed to air.

Actually it isn't good at keeping water out. Curing process is based on water contained in the air, not oxygen or nitrogen. Of course there also are two part silicones but I don't think they are that good in this regard.

[JoeN]

Cerakote or Duracoat?  Is there some spray ceramic that might make a good seal that only needs to be lightly baked or not baked at all?  And you can choose really cool colors...

[David Hess]

If they were hermetically sealed, then they would be called bombs instead of capacitors.

[Paul Moir]

People wrongly think water doesn't migrate through plastics or rubber.  The only real options are metal, which doesn't work around the leads, and glass, which is tricky.
Hermetically sealed electrolytic caps do exist though.  Check this out:  https://www.digikey.com/product-detail/en/cornell-dubilier-electronics-cde/MLSH322M030JK0C/338-4248-ND/6096195

[Gyro]

If they were hermetically sealed, then they would be called bombs instead of capacitors.

No, you could still have the engineered 'weak spot' in the top, just like normal electrolytics, for safety venting. Still hermetic.

[bd139]

The vents barely work on half of the electrolytics made, even from named brands like Rubycon. I've blown up hundreds of the things over the years. No thanks:

[T3sl4co1l]

The vents barely work on half of the electrolytics made, even from named brands like Rubycon. I've blown up hundreds of the things over the years. No thanks:

On a related note, I have some film-in-oil can capacitors (not actually motor run caps, but that's the basic idea) which have a patented safety mechanism: the top is double-crimped with an extra bead, so that it's kind of scrunched down in length.  If gas generation occurs, the top poofs out, disconnecting the internal leads (which are only just long enough to connect).

Hopefully the top doesn't poof out too far, too rapidly, and explode and catch fire anyway.

Tim

[wraper]

The vents barely work on half of the electrolytics made, even from named brands like Rubycon. I've blown up hundreds of the things over the years. No thanks:

In the video only K50-18 cap has safety vent and it did not explode. Capacitor types which exploded violently, don't have any safety vent and you show this video as example of safety vent not working  .

[T3sl4co1l]

This is probably an extraordinarily stupid question but if the longevity of super capactiors is limited by the electrolyte evaporating why can't they just be sealed off completely without evaporation they would last piratically forever right?
I am probably missing something very big here

But they are!

Some, anyway.  I find a few results for military supercapacitors, although I can't seem to find a MIL STD for them (if it exists).  One picture shows a glossy backfill on the component, presumably an epoxy encapsulant.

Mil grade electrolytics do the same thing: either an aluminum can backfilled with epoxy (that always seems to be pink or red, on the ones I've seen?), or a metal can with a glass frit seal on one end and a soldered or swaged seal on the other (hermetically sealed, also dry and wet tantalums made the same way).

No need for high temperatures, solder does a fine job as hermetic seal.

Tim

[Gyro]

The vents barely work on half of the electrolytics made, even from named brands like Rubycon. I've blown up hundreds of the things over the years. No thanks:

They probably barely work half the time because of poor, non hermetic rubber seals at the bottom, allowing them to slowly spill their guts all over the PCB as the pressure builds instead. It ought to be possible to volume manufacture them to blow at a very consistent pressure.

[wraper]

But they are!

Some, anyway.  I find a few results for military supercapacitors, although I can't seem to find a MIL STD for them (if it exists).  One picture shows a glossy backfill on the component, presumably an epoxy encapsulant.

There were Soviet made K50-6 capacitors filled with epoxy. They had way lower reliability that Soviet capacitors which had rubber seal.

[David Hess]

Another reason they may not bother with better seals for supercapacitors is that they degrade at a rate dependent on the voltage.  It works out to about 1/10th the operating life for every 0.2 volt increase.  Most applications operate them at about the 10 year point where loss of electrolyte through the non-hermetic seal will not be the cause of failure.

[stj]

there are, or atleast was sealed ones from NEC,
i see them in stuff i repair backing up sram.

STEEL cylinders with a crimped base over a plastic bung with i think more behind it.

[Assafl]

Are the red Roederstein and Siemens "Bakelite" capacitors considered sealed?

Audiophiles seem to like them. They claim the sound is as if a veil is lifted and the transistor sound undergoes vastation.

[voltsandjolts]

But they are!

Some, anyway.  I find a few results for military supercapacitors, although I can't seem to find a MIL STD for them (if it exists).  One picture shows a glossy backfill on the component, presumably an epoxy encapsulant.

https://www.tecategroup.com/products/data_sheet.php?i=PC&t=SERIES

[Cliff Matthews]

I have some Litton 1980's el-caps sealed with glass-like resin below the rubber, then crimped, and then poly-wrapped. Marked 180uF/25v 105C but now they measure 30% high with 0.14 ESL. As can be seen, they're still full of electrolyte. Made in USA - can they still be trusted?

[T3sl4co1l]

Sure -- they may need reforming though.

I wonder if reforming is ill-advised for that type...

Tim

[wraper]

I have some Litton 1980's el-caps sealed with glass-like resin below the rubber, then crimped, and then poly-wrapped. Marked 180uF/25v 105C but now they measure 30% high with 0.14 ESL. As can be seen, they're still full of electrolyte. Made in USA - can they still be trusted?

No. Well, you can reform them but why to use such old stuff which already degraded. Aluminium oxide layer which is insulator partially dissolved over time thus becoming thinner. The result of this is increased capacitance and lower withstanding voltage.

[Cliff Matthews]

Meh, no real loss, just one drawer full of old but unique history. Hope there was no PCB's inside, since I didn't use gloves while sawing it open..

[wraper]

Meh, no real loss, just one drawer full of old but unique history. Hope there was no PCB's inside, since I didn't use gloves while sawing it open..

You could check a datasheet if you can find one. Sometimes electrolytic capacitors may have something like +80/-20% tolerance instead of usual +/-20%. I personally would not use 30+ years old capacitors regardless of how good they are, though.

[perieanuo]

Best you can do is put resin, is what we did in manufacturing motor control boards, laser diodes and medical equipment. Which resin, that's other discussion... I think here people referred as glass resin compound (I used a compound with 1:6 ratio, black resin with transparent hardener). 24 hours for drying. Good solution but not cheap or easy to do with mixed components and some connectors...

Envoyé de mon HUAWEI NXT-L29 en utilisant Tapatalk

[SeanB]

Resin sealed caps do degrade with time and temperature cycling, the resin is not totally vapour proof, and the liquid ( and more importantly the hydrogen gas generated internally as the capacitor reforms) does diffuse through it and it slowly degrades. The only capacitors with a indefinite life are the wet solder sealed glass frit insulated tantalum types, as used in the Voyager spacecraft. Those are truly hermetic, as there is literally zero outgassing from them in a bare non sleeved case, and they will last decades in operation at 125C. Electrolyte in them is sulphuric acid, and the inside is silver plated steel or tantalum only, with a PTFE inner support structure.

The rectangular sealed capacitor referred to from Digikey I will tell you a nasty secret about some of them, they are internally just regular axial electrolytics with rubber seals, simply with a PCB each end that commons a few capacitors to make the final value, and then they are potted with a flexible encapsulant and an outer insulation sleeve to keep the terminals and board insulated from the case, and the leadouts are 2 short PTFE coated wires that are soldered to the inside of the terminals, and then the outer seal is low melting point solder for final seal after crimping the top on. They do not leak in use externally, but do lose vapour from the internal capacitor units into the housing, so they do fail by going high ESR and eventually nearly open circuit, but do not outgas at all from the case. Bloody expensive as well, and they were a pain, as the lead time was so long to get them. however those MSLH ones are slightly different in that they just take the rolled foil pack and flatten it ( so no lost mandrel construction like a regular electrolytic, these need a removeable mander in manufacture, a hand process to remove) then stuff it in a can, and laser or TIG weld it in a CA enclosure, and then use a vacuum fill to get the electrolyte in there, with a welded on fill port that is also a pressure vent ( weak point in the case for emergency vent). Will still have an operating life limit from degradation of the electrolyte with time and reforming though.

[Paul Moir]

Epoxy normally has a fairly lousy MVTR.  Not surprising a rubber, which likely also has a barrier layer added somewhere, can beat it.

[Raj]

better seal,would cause heating issues too.Why not create a circuit design,making them easy to replace?

[pieman103021]

better seal,would cause heating issues too.Why not create a circuit design,making them easy to replace?

That is not always an option. Depending on the application the circuit might not be able to be repaired or even be accessed.

Sent from my BLN-L24 using Tapatalk

[james_s]

Are the red Roederstein and Siemens "Bakelite" capacitors considered sealed?

Audiophiles seem to like them. They claim the sound is as if a veil is lifted and the transistor sound undergoes vastation.

I've had to replace a bunch of those in Bosch ECUs and VDO gauges, they sure don't seem any more reliable than the normal rubber sealed type. I don't think loss of electrolyte is the only failure mode, I suspect it also chemically changes, perhaps it breaks down via electrolysis.

[Electro Detective]

Manufacturers won't cook their golden goose just yet   

You can count on leaking capacitors and spewing batteries to keep bunnies consumers spending solid for a few years more, and keep the landfills topped up.   

That said, I have seen better quality capacitors on some PC motherboards and graphics cards, which 'look' less likely to leak 

I only buy stuff for myself to keep long term ONLY if I know I can get to the caps and replace them EASILY, and keep a log book/reminder notes on batteries in multimeters   

[BravoV]

Read this -> http://www.chemi-con.co.jp/e/catalog/pdf/dl-e/dl-sepa-e/dl-precaution-e-170401.pdf

Especially the whole 5.1

[james_s]

Manufacturers won't cook their golden goose just yet   

You can count on leaking capacitors and spewing batteries to keep bunnies consumers spending solid for a few years more, and keep the landfills topped up.   

That said, I have seen better quality capacitors on some PC motherboards and graphics cards, which 'look' less likely to leak 

I only buy stuff for myself to keep long term ONLY if I know I can get to the caps and replace them EASILY, and keep a log book/reminder notes on batteries in multimeters   

There's no conspiracy to make capacitors fail early and keep people buying new equipment, that's just a side effect. It's all about cost, most companies if given a choice between a component that costs 9 cents each and lasts at least as long as the warranty and a component that costs 10 cents each and lasts 3 times as long they'll go with the cheaper part. Most consumers don't know any better, they shop entirely by price and will buy the cheaper item rather than a more expensive one that looks similar. The ultimate result is that virtually everything is engineered down to the lowest possible cost while being just good enough that the masses will still buy it.

[Electro Detective]

and that is why capacitors and batteries will keep leaking into the next century and beyond... don't fix it if it ain't broke = $$$$$$ 

[stj]

they make them too small to handle the current IMO.
look at the smd electrolytics - nothing but trouble.
they leak a lot and you cant even see it for ages.

what we need is shit polymers.
ones with higher esr so they can replace general to mid-tier electrolytics.

[wraper]

most companies if given a choice between a component that costs 9 cents each and lasts at least as long as the warranty and a component that costs 10 cents each and lasts 3 times as long they'll go with the cheaper part.

Price difference is much higher, many times, not 10%. And there is no explicit border between bad and good. Often cheaper parts last longer than many much more expensive counterparts and there is no guarantee that expensive parts won't fail prematurely. Like plagued Nichicon HM,HN and Chemi-con KZG, KZJ.

[T3sl4co1l]

The paradox of quality is this: a low quantity item is either going to be very unreliable (or its actual reliability very poorly established) -- because a large sample size does not exist -- or very expensive and carefully analyzed (like aerospace components are).

In massive commercial production, you get both high quantity and high quality.  It is generally in the manufacturer's best interest to pursue high quality, because it's cheaper to fix it on the front end -- refine the industrial process -- than to test quality into the part.

"Testing quality into a product" is something that can be done for products that are quick to test, and which fail due to random fabrication defects rather than subsequent stresses and aging.  Semiconductors are a great example.  Capacitors are a terrible example -- spending a few extra seconds on the tester tells you nothing about the 3 month+ survival rate of the component in actual use.  (Instead, a sampling approach with accelerated aging is needed, assuming the offending mechanisms can be accelerated proportionally.  Results still take a long time to obtain, at which point the product has long since been shipped and used.)

Tim

[David Hess]

I suspect the changes leading to lower design lifetimes were:

1. A greater use of switching regulators leading to capacitors selected for ESR rather than the capacitance needed for holdup time.  Notice that switching regulator input capacitors rarely fail.
2. A better understanding of the relationships between ESR, ripple current rating, and operating life.
3. Better specified capacitors.

The above allows the operating life of the device to be closely determined and limited.