Capacitor Quality

[mikeselectricstuff]

MOD:
This thread was separated from the Siglent SDS3055 teardown:
https://www.eevblog.com/forum/blog/eevblog-829-siglent-sdm3055-bench-multimeter-teardown/msg820233/#msg820233

I wonder how much difference there really is between the lower-cost parts from "established brands" and lesser known (Outside China) brands like Lelon, Capxon etc.

Designing something in just because they're an established name isn't necessarily good engineering when cost is a major factor.

Given a fixed budget, what might they have skimped on elsewhere because they decided to use a more expensive part because of the brand name without really knowing if it's actually any better?

IME of 30-odd years of repairing things, the biggest differentiator of capacitor reliability is  heat exposure rather than capacitor brand.
For caps that are not thermally or electrically stressed, I'd be surprised if there was much difference between brands.

A "torture" test comparing performance of various cap brands might make an interesting episode...

[c4757p]

He won't... too afraid a non-Japanese capacitor might do well....

[EEVblog]

I wonder how much difference there really is between the lower-cost parts from "established brands" and lesser known (Outside China) brands like Lelon, Capxon etc.
Designing something in just because they're an established name isn't necessarily good engineering when cost is a major factor.

No, but in this case there was evidence that they haven't skimped on every last cent for everything.
And we have seen Nippon-Chemicon caps in a really horribly cheap PSU teardown before.

Given a fixed budget, what might they have skimped on elsewhere because they decided to use a more expensive part because of the brand name without really knowing if it's actually any better?

About the "knowing" aspect. What engineer is going to do the hundreds or thousands of hours thermal testing required to test if some cheapie brand is up to scratch?
You don't of course. If you deem capacitor life to be a potentially important factor in your product, then as a design engineer you do not take chances (budget allowing), you design in the top brand parts that have many decades of quality brand reputation in the industry. Do you think the tops brand have not earned their industry reputation?
This is standard design practice that extends to many different parts and their specs on their datasheets, not just caps.
For instance, you need a reliable reed relay. Do you trust a one-hung-low brand who's datasheet says 1,000,000 operations, or do you trust a Pickering with the same 1,000,000 rating? The answer is obvious. Therefore in reviewing or tearing down gear, it's important to look at this stuff.
Same goes for using a genuine Analog Devices True RMS converter chip, or some asian knockoff as another example.

A company can of course do extensive testing to qualify a lesser brand part, no problem with that, but that's an issue up to them. And unless they publish their results, the industry does not have anything to go on except the slow build up of experience over decades.

IME of 30-odd years of repairing things, the biggest differentiator of capacitor reliability is  heat exposure rather than capacitor brand.
For caps that are not thermally or electrically stressed, I'd be surprised if there was much difference between brands.

I'm almost certain that you are wrong.
The "bad capacitor" problem was a real issue in the industry:
https://en.wikipedia.org/wiki/Capacitor_plague
The major brands were not affected.
And who discovered the problem? And ex-scientists form Rubycon.
http://www.molalla.net/members/leeper/alumin~1.pdf

But sure, temperature and electrical stressing is always the major factor in capacitor failure.

A "torture" test comparing performance of various cap brands might make an interesting episode...

The number of hours involved in doing that would be huge. And the likelihood of getting a null result would likely be high. As this issue is more to do with industry experience on a mass scale than one that can be easily reproduced.
Image if I tested them and there was no difference? Would that prove that the crap brands are just as good? No it wouldn't, because collective industry experience proves otherwise.
In this case it's much easier to prove a positive result (failure) than a negative one (no real failure/difference)

But I agree that it would be interesting.

[madires]

IME of 30-odd years of repairing things, the biggest differentiator of capacitor reliability is  heat exposure rather than capacitor brand.
For caps that are not thermally or electrically stressed, I'd be surprised if there was much difference between brands.

Of course, heat exposure is the most stressing point in a electrolytic cap's life. But there are differences between brands. In some 20-25 years old audio stuff I've found a few bad low stressed Philips caps while beefy power supply filter caps from ROE (Rödermark, now part of Vishay) are still in specs.

[EEVblog]

He won't... too afraid a non-Japanese capacitor might do well....

But that's actually a real possibility.
Then how do you explain a result that shows no major difference?
Was the testing statistically significant at all? (likely not)
Does it say anything about long term quality control of that manufacturer? (the answer is obviously no)
I would hate if that happened and then everyone falsely used my video as some justification that the entire industry is wrong, and there really is no difference in production quality control of capacitors.

[c4757p]

 

[tautech]

He won't... too afraid a non-Japanese capacitor might do well....

But that's actually a real possibility.
Then how do you explain a result that shows no major difference?
Was the testing statistically significant at all? (likely not)
Does it say anything about long term quality control of that manufacturer? (the answer is obviously no)
I would hate if that happened and then everyone falsely used my video as some justification that the entire industry is wrong, and there really is no difference in production quality control of capacitors.

So you rate every leading manufacturer to stay at the top of their game. 
And none of the others to evolve and challenge them.

History has proven every dog has their day.

[mikeselectricstuff]

And we have seen Nippon-Chemicon caps in a really horribly cheap PSU teardown before.

Well at last that's what the label says - these days it's hard to know if parts used in really cheap gear are genuine or not. A label on a plastic sleeve is pretty low down the scale of difficulty for faking,

About the "knowing" aspect. What engineer is going to do the hundreds or thousands of hours thermal testing required to test if some cheapie brand is up to scratch?
You don't of course. If you deem capacitor life to be a potentially important factor in your product, then as a design engineer you do not take chances, you design in the top brand parts that have many decades of quality brand reputation in the industry. Do you think the tops brand have not earned their industry reputation?

I'm just questioning the assumption that all lesser known brands are necessarily inferior, or even have that production in recent years has a better track record. As consumers/reviewers we only see an infinitely small sample and neither of us have worked in the real industry recently enough to have experience of what is coming out of any capacitor factory nowadays.
The established names are having to compete at the lower end, and there's only so much that their customers will pay for the name, so their production methods for their cheaper parts aren't going to be too different from anyone else, though you might hope (with no way of actually knowing) that QC may be better.
A quick search on Digikey for 100uf 16V 105 deg radial caps shows parts from all the major established brands below GBP0.04 at 10K. I don't think it's reasonable to expect these to perform as well as the ones you find still going string in a 30 year old piece of testgear.

A company like Siglent may well have been using a particular Obscure (to us) brand by the hundred thousand ( and possibly their subcontractors, millions) for long enough to have at least as much confidence in it than any of the ones we know from Digikey, maybe more.
 
All I'm saying is that maybe there's a little too much emphasis on choice of cap brand as an indicator of quality, based on minimal actual evidence. NCC/ELNA/Panasonic etc. = good ( at least in intention, probably). Unknowns = maybe no better or worse, we don't know.

[mikeselectricstuff]

The number of hours involved in doing that would be huge.

Yes but it's something that could be just set up and  left to get on with itself over a few months with minimal actual attention. Though possibly not in an office environment - you wouldn't one of those expensive fire-service callouts because a neighbour smelled a smouldering capacitor...

[EEVblog]

So you rate every leading manufacturer to stay at the top of their game. 

Of course not, but it's not a bad bet.

And none of the others to evolve and challenge them.

Of course they can, and I encourage that.
How about they actually do the testing, or pay someone like UL to do independent testing to prove how good their capacitors are?

But of course this will end up as "Dave doesn't want to do the testing because he's afraid blah blah".
Why does it have to be about me?
How about you go do the testing and publish some results? How about Mike?

Go to the Lelon website for example. Where are the white paper on quality control etc?
Just some crap on their ISO quality control:
http://www.lelon.com.tw/en/about2.php
What are they doing to prove to the industry they are any good and can be trusted for long term quality control? Which they have even more incentive to do being one of the major player in the "bad cap" problem over a decade ago?

[mikeselectricstuff]

What are they doing to prove to the industry they are any good and can be trusted for long term quality control? Which they have even more incentive to do being one of the major player in the "bad cap" problem over a decade ago?

It could be argued that someone who has had a problem and fixed it is a better bet than someone who has never had a problem, as they've learned from their mistakes, while the others may have just been lucky.
"That employee's mistake just cost  $50,000, aren't you going to fire him?"
Boss : "Why would I do that? I just spent $50,000 training him..."

[EEVblog]

I'm just questioning the assumption that all lesser known brands are necessarily inferior

Sure, nothing wrong with, and I support that.
Doesn't change the fact that the big brands have the rep and the small ones don't.
It's not just me on some capacitor fanboy crusade, people keep asking for and talking about what brand caps are used in products, so I always point it out as a matter of course.

, or even have that production in recent years has a better track record. As consumers/reviewers we only see an infinitely small sample and neither of us have worked in the real industry recently enough to have experience of what is coming out of any capacitor factory nowadays.

Yep, and you can work in the industry for 40 years and still have very little real quantifiable experience in capacitor quality and reliability, apart from the anecdotal .

The established names are having to compete at the lower end, and there's only so much that their customers will pay for the name, so their production methods for their cheaper parts aren't going to be too different from anyone else, though you might hope (with no way of actually knowing) that QC may be better.

Of course, it's a brand trust thing.

A company like Siglent may well have been using a particular Obscure (to us) brand by the hundred thousand ( and possibly their subcontractors, millions) for long enough to have at least as much confidence in it than any of the ones we know from Digikey, maybe more.

They may well have. 

All I'm saying is that maybe there's a little too much emphasis on choice of cap brand as an indicator of quality, based on minimal actual evidence. NCC/ELNA/Panasonic etc. = good ( at least in intention, probably). Unknowns = maybe no better or worse, we don't know.

Sure. But once again, there are top names in the business, and they have built that rep over decades. For me as an experienced design engineer not to share that info with viewers would be bad form.
If someone asked me (and they do, a lot) what brand capacitors are deemed to be the best quality, then there is a very real list for that, whether anyone likes it or not. or whether it's fair or not.

[mikeselectricstuff]

If someone asked me (and they do, a lot) what brand capacitors are deemed to be the best quality, then there is a very real list for that, whether anyone likes it or not. or whether it's fair or not.

Not that they do very often, but if someone asked me I'd suggest that they are asking the wrong question. Like when someone asks what's the best microcontroller
You basically get what you pay for, and  that the way that caps are used, in particular heat exposure and ripple current, is probably a more significant  factor than differences between brands at a particular price point.
A badly laid out PSU full of NCCs running very hot is probably going to die before a well-cooled one full of CapXons.

[mikeselectricstuff]

Go to the Lelon website for example. Where are the white paper on quality control etc?
Just some crap on their ISO quality control:

Lots of even fairly high-end Chinese companies have crappy websites. I don't think you can read anything useful into that. they clearly don't need marketing wank to sell their product - they're too busy making stuff to mess about with corporate fluff like White papers etc. It's a totally different culture.
The big brands do all that stuff as marketing to maintain their image to justify their price premium, and to satisfy their higher end customers, who the Chinese don't care about selling to as the volumes are too low.

[T3sl4co1l]

About the "knowing" aspect. What engineer is going to do the hundreds or thousands of hours thermal testing required to test if some cheapie brand is up to scratch?

Well... a good engineer, of course.

Typical example: product that's been in production for a year.  Time for a round of cost-reduction.

That's one round of parts out in the field.  Accelerated life tests can be done to qualify new (potentially cheaper) substitutes.  Once a choice is made, another round of production can be sold, and evaluated by the marketplace as a whole (hopefully, without requiring recalls).

This is not something that will normally apply to < 10k/yr products, or to new design (unless you've personally done all that qualification yourself, or someone has that data available), which is likely the range of experience of most designers here.

But to categorically disregard such a step out of hand, is bad engineering!

Tim

[rsjsouza]

I'm almost certain that you are wrong.
The "bad capacitor" problem was a real issue in the industry:
https://en.wikipedia.org/wiki/Capacitor_plague
The major brands were not affected.
And who discovered the problem? And ex-scientists form Rubycon.
http://www.molalla.net/members/leeper/alumin~1.pdf

It is somewhat ironic when you link this thread with the last sentence of the PDF article above:
Teapo suspects it is an attempt by Japanese aluminum capacitor suppliers to gain market share, and that the unfortunate incident involving a few companies will unfairly categorize the entire aluminum capacitor industry in Taiwan.

Coincidentally, your immediate suspicion of Jamicon branded caps in episode #824 (14:30) reflects that same fear. Still from the same article:
Further intelligence revealed that Kamei/Jamicon sources its electrolytes from Sanyo in Japan, and likely was not affected by the faulty electrolyte supply.

All that said, in my opinion it is caveat emptor. Some top brands have their track record (despite some goof ups happened in the past with a few of them) while others are the "Rigols/Siglents" of the cap world with their hits and misses. In both cases you always have to look at Mike's post: unless you are a very high volume customer that could potentially request additional test information, you get what you pay for.

[tautech]

It should be made mention that this thread (probably wisely) was ripped by Dave from the Siglent SDS3055 teardown:
https://www.eevblog.com/forum/blog/eevblog-829-siglent-sdm3055-bench-multimeter-teardown/msg820233/#msg820233

Just to put the opening discussion in context.....

BTW
Mike, you have another thread to your name now. 

[EEVblog]

A badly laid out PSU full of NCCs running very hot is probably going to die before a well-cooled one full of CapXons.

No doubt.
But that's actually a different question to which have the best reputation for quality.

[calexanian]

What really is more important in cost saving in components, and particularly in caps is knowing the supply chain. If even a lesser brand can be just fine so long as they meet their own specs, but a high quality brand can suffer if an unscrupulous vendor or broker sells you old, damaged, or fake parts. Knowing the supply chain can really make or break your product.

[EEVblog]

But to categorically disregard such a step out of hand, is bad engineering!

I of course didn't mean to imply that.
But obviously no practising engineer has the time and resources to do that sort of testing for regular product design.

[EEVblog]

Lots of even fairly high-end Chinese companies have crappy websites. I don't think you can read anything useful into that. they clearly don't need marketing wank to sell their product - they're too busy making stuff to mess about with corporate fluff like White papers etc. It's a totally different culture.
The big brands do all that stuff as marketing to maintain their image to justify their price premium, and to satisfy their higher end customers, who the Chinese don't care about selling to as the volumes are too low.

Yep, that's the real world whether you like it or not.
No one can complain that a component maker "unfairly" doesn't get a good rep for quality if they don't actively show it and explain it and provide evidence (worthwhile of not). The only other way to build up a quality rep is over time by having your product used for decades without any major issues.

[EEVblog]

Typical example: product that's been in production for a year.  Time for a round of cost-reduction.
That's one round of parts out in the field.  Accelerated life tests can be done to qualify new (potentially cheaper) substitutes.  Once a choice is made, another round of production can be sold, and evaluated by the marketplace as a whole (hopefully, without requiring recalls).

And that's the trick, it's more of risk going with a lesser known brand.
I've done stuff like this. But in the case of capacitors that are fairly complex chemically and rely on good and consistent manufacturing quality, my report to Management after said testing is going to say something like this:
"Accelerated life testing was performed on N products of X brand, under Y conditions, and indications are that they may be as reliable as brand Z. But this testing is only representative of small sample over a small batch time period, and may not reflect the long term quality reliability of company X. There will still be some risk with going with company X long term until long term field testing data is in."
Or something like that.
Basically, any professional engineer knows there is some risk with going with that new brand based on some accelerated life cycle testing on a small sample.

A lot of engineering is about lowering the risk and using "best practice". It's like the old "Nobody ever got fired for buying IBM/Fluke/Keysight etc".
In the case of caps, no engineer ever got fired for designing in Panasonic caps if it's within the budget. It's smart design.
If for some reason those Panasonic caps went tits-up because they had a production problem, then no one is going to hound you for the choice of using those Panasonic caps. If you used CapXon, well, good luck defending that choice.

[EEVblog]

High quality parts matter, but not nearly as important as careful engineering. For most of the time, the problem is between pebkac.

No one will argue that.

[BravoV]

Even using a branded 1st tier cap, apart from external heat exposure, still curious if a crappy design job can be a factor say like the circuit works constantly exceeding the cap's rated max ripple current ?

Wondering if there is such dedicated circuit just to torture a cap on it's ripple current ? Like an AC/pulsing adjustable current source to heat up the cap internally.

[SeanB]

Simple way to stress test a capacitor is to put it in a small plastic box, chuck in the nastiest SMPS and power it 24/7. Also known as a knockoff phone charger, most have a low brand or non name brand pair of capacitors inside, which will die in a year.

I just took a Multichoice PVR apart here, mostly because I do not watch TV, and because i wanted the 500G WD drive inside. PSU is all Rubycon KY series and the main boards are a mix of Rubycon and Nichicon, though there are a fair number of Teapo as well, and 2 small 10mm diameter 10mm high 100uF 16V pale blue no name 85C capacitors there as well. 2 gold Lelon caps there as well on the main ASIC board by the power input. The main reason these fail is because they are placed nicely in a warm TV cabinet, with no airflow, and, even though they do have a little Sunon fan to move the air inside over the hard drive, they cook. Other failure is the tuner modules cook themselves, or they get killed by lightning.