General > General Technical Chat

Bad caps documentary - very good

(1/10) > >>

flipper:
Nothing new to us, and a good watch:

tigrr:
Samsung and LG have the highest percentage of failed capacitors. It's a bit disappointing...

richnormand:
Interesting. Nothing really new but more people should be aware.

An interview with Rossmann would have been appropriate in addition to Ifixit.

Over 20 years ago when I was travelling a lot I noticed many airports using the Viera Panasonic plasma for 24/7 flight display.
I got one for home and it still works. It was not the cheapest but you can make a product last if you want to.
Also, in the documentary, you can see a list of defective TVs and the capacitors but no failure in the Panasonic/Sony/.. column.

A month ago I replaced the battery in my son's Surface. I now understand the (very) low repairability score from Ifixit now. |O

Planned obsolescence indeed.

EPAIII:
There is nothing new about electrolytic capacitors failing. I spent over 45 years at TV stations and related places like production companies repairing not just TVs, but all types of equipment, even digital before the public had even heard that word. And of all the components that I replaced electrolytic capacitors were near the top of the list if not at the top.

Please notice that I inserted the word "electrolytic" into that statement. Other types of capacitors have a much, much lower failure rate. In that 45+ years of replacing defective components I don't think I came across even a single film capacitor that had failed. I would say the same about ceramic disc capacitors but there I did have one that failed. But just ONE in 45 years. And it had an intermittent short. Since it was in a low Voltage circuit and over-rated by at least a factor of 5X, I suspect it was a rare, factory defect and not some kind of slow, constant deterioration. No, all the other types of capacitors combined don't even come to 1% of the number of electrolytics that I have replaced. So you can't say that capacitors in general fail a lot. That honor goes only to electrolytics.

Individual transistors were used a lot in that equipment. I have seen individual devices that had at least 500 2N3904s in them. Some may have had 1000 or more. Transistors are probably second on my personal list of replacements. More than electrolytic capacitors? Probably not but I did replace a lot. Transistors can and do deteriorate from the same things that can attack electrolytic capacitors. And heat is at the top of that list.

Inductors (coils) also have a very low failure rate. They don't slowly deteriorate: they usually fail from a current spike that vaporizes the weakest point in the wire they are wound with. Or, less often, they can have an internal short. But both of these are very rare events.

So why do electrolytic capacitors fail so often? Why do they deteriorate? And why on earth do designers still use them when there are other types of capacitor that will last almost to forever? I will tackle the last question first. The answer is simple; it is the capacitance and Voltage ratings that are available in each type. In order to have a film capacitor as large as an aluminum electrolytic the size of my little finger, that film capacitor would need to be as big as a house and would probably cost as much as ten luxury homes. The capacitance of a capacitor is directly proportional to the area of the plates and inversely proportional to their separation. Yea, I know everyone here already knows that. So when larger values of capacitance and higher Voltage ratings are needed, the electrolytic type capacitor really shines. I mean it is like a super-nova compared to a dying star at the end of it's life. And the other side of the engineering here is the cost vs performance trade-off. Micro Farad for micro Farad, ceramic or film capacitors are easily 10,000 times more expensive than electrolytics. At that rate for parts that are present in ten or more in them, things like the cell phone which we all carry around every day would only be affordable by the richest of the rich. You and I would simply be out in the cold.

But in engineering circles it is well known that every thing is a compromise. The famous saying is,
"I can make it fast.
I can make it cheap.
I can make it well.

PICK ANY TWO!"
And that is the truth, you only get to pick TWO, not all three.

Electric capacitors are made fast and cheap. They must be cheap for otherwise there would be no need to even make them because no one could afford them.

Planned obsolescence? NO!

Planned affordability! YES!

And it is such a fundamental and obvious engineering choice that no one who designs electronic devices would spend even a single micro-second thinking it over. So, like it or lump it, that's it.

Now, could electrolytic capacitors be made better? To last longer? Of course they could. But you are back to that fast-cheap-well thing again. And yes, they probably do last for the average lifetime of the devices they are used in. And if you do want your devices to last twice as long, find someone who knows how to replace them when they fail. I have personally restored things like TVs to life by just ordering and replacing ALL the electrolytic capacitors in them. And, for what it is worth, I replaced those TVs before they failed for a second time.

tom66:
I still repair electronics from time to time and I think in the last decade I've seen one or two failed electrolytics.  Whilst I'm sure they do contribute to product failure the more common faults now for TVs are issues with the main flash memory.  Bad capacitors are essentially an issue of the past.  Even Chinese-made electrolytic capacitors can still last a long time because they have spent effort on the research and development to improve quality.  I would still not specify them myself because their performance usually isn't as good as the Japanese makes, but I don't design pricing sensitive equipment.

Some engineers do not appreciate that writing to flash memory wears it out because each write requires a page to be erased and data saved to it, this process can only be done around 1,000 - 3,000 times per page.  Worse, the hotter the device, the quicker the failure; and, in particular, the type of failure is for flash cells to become progressively leakier, which is itself temperature dependent.  The issue can then occur that as the chip warms up - or, often, the chip is placed near to a hot system-on-chip IC - the pages begin to read out corrupt or fail to read out due to ECC fault.   This will cause the OS to crash or bug out in various ways.   Unfortunately, this is rarely recoverable because manufacturers often do not supply replacement firmware even as a binary image, nor do they provide a way to load this in case of NAND failure (the issue almost never occurs on NOR flash).

Navigation

[0] Message Index

[#] Next page