What is the lifetime of common parts/circuits these days ?

[coppice]

I think cars are a pretty good example. The infotainment in modern cars is so central to key operations, like heating and cooling, that you can't really use the car without it. .

I believe you have misunderstood my point. Of course the electronics in cars behave like electronics everywhere. That is to say, the ageing mechanisms are few and far between. Very few. Electronics can still fail, despite having hardly any ageing mechanisms. Perhaps you are confusing ageing with intrinsic failure rates. 

But a car is much more than electronics. It is a complex machine with a lot of mechanical parts. And the reason a car is a bad analogy is because all these mechanical parts do have numerous ageing mechanisms, all of which are well understood. It is easy to design in a specific service life to these mechanical components, and it is easy to predict when they will wear out.  Neither of these apply to electronic components.

So yes, you can compare car electronics to any other electronics and say they are analogous. But you cannot usefully compare the mechanical elements of a car with electronics in general, because they have almost nothing analogous.

I believe you have misunderstood my point. The car industry was always going to embrace a lot of electronics in cars. However, they have chosen to package it in ways where its particular reliability characteristics suit their desire to make cars reliable enough to not be too much of a PITA, but create a "not worth fixing it" moment that stops them surviving too long. The electronics in a car in very modular, with little MCU circuits distributed through practically every area. A nice car may have 100 of them these days. The central display and UI part could have been a modestly priced thing, that you could change like replacing a tablet. In fact, in many cars it looks like a tablet. However, it has been constructed as a module that costs thousands of dollars to replace. I think that is neither an inevitable consequence of a car's design, or an accident.

[SteveThackery]

I believe you have misunderstood my point. The car industry was always going to embrace a lot of electronics in cars. However, they have chosen to package it in ways where its particular reliability characteristics suit their desire to make cars reliable enough to not be too much of a PITA, but create a "not worth fixing it" moment that stops them surviving too long. The electronics in a car in very modular, with little MCU circuits distributed through practically every area. A nice car may have 100 of them these days. The central display and UI part could have been a modestly priced thing, that you could change like replacing a tablet. In fact, in many cars it looks like a tablet. However, it has been constructed as a module that costs thousands of dollars to replace. I think that is neither an inevitable consequence of a car's design, or an accident.

I'm probably not quite as cynical as you about the motives underlying their design decisions. But even if you are correct, we have wandered well away from the ageing mechanisms pertaining to electronic components, which is what the OP was asking about.

In terms of replacement costs for the electronic modules in a car, I agree with you: they are outrageous. You sometimes are faced with a "not worth fixing" moment, but that is usually because of the ridiculous cost of the parts rather than any crafty "designed to fail" strategy. Furthermore, you are quite often forced to use a main dealer because the car won't just accept a replacement module - it has to be configured before the rest of the car systems will talk to it. A bit like "signing" a file. And I'm pretty sure they would refuse to "sign" a second hand module from a scrapyard. I've no idea why they do this, but you are right: a perfectly good car might be scrapped simply for want of a small electronic module for which you must find a thousand pounds, maybe two. It is unacceptable, in my opinion, but we have no real way of avoiding it, except by buying a vintage car with no electronics in it (OK, maybe electronic ignition).

[SteveThackery]

Trust me, it isn't naive. People like me spend a whole career studying reliability and failure modes...

I'm sorry - it occurs to me that I've made a misleading statement. When I said "people like me" I meant Reliability Engineers. I want to be clear that I did not spend my whole career as a Reliability Engineer.

[SteveThackery]

Suppose a device has a 10 year MTBF. What do you think is the probability of it lasting 10 years before it fails? If you don’t know the answer, you will be surprised!

MTBF is the inverse of the failure rate, and the failure rate is numerically equal to the probability that an individual item will fail within a given time interval (typically a year). So, an MTBF of 10 years means that 1/10 (ie 10%) of the items will fail within any given year, and there is a 10% chance that any particular item will fail in any given year.

Start off with 100 items. 10% will fail in year 1, leaving 90 left. The next year another 10% will fail, leaving 81 left. Like this:

Year 1: 90 left
Year 2: 81 left
Year 3: 73 left
Year 4: 66 left
Year 5: 59 left
Year 6: 53 left
Year 7: 48 left
Year 8: 43 left
Year 9: 39 left
Year 10: 35 left

So you end up with only a 35% probability that a component will last as long as its stated MTBF, which is counter-intuitive to most people.

This demonstration assumes that the items are not repaired, so it is more proper to refer to the MTTF (mean time to failure) rather than MTBF (mean time between failures). They are mathematically equivalent. The former is used for individual electronic components, because they are not repaired - they are discarded and replaced with a new one; the latter refers to electronic systems, because they are (potentially, at least) repaired and brought back into service.

[Circlotron]

I could imagine an epoxy coated silvered mica capacitor lasting a thousand years provided there were no impurities trapped inside.

https://en.wikipedia.org/wiki/Silver_mica_capacitor#/media/File:Silver_mica_capacitors.jpg

[CatalinaWOW]

Trust me, it isn't naive. People like me spend a whole career studying reliability and failure modes...

I'm sorry - it occurs to me that I've made a misleading statement. When I said "people like me" I meant Reliability Engineers. I want to be clear that I did not spend my whole career as a Reliability Engineer.

I actually do respect your knowledge and experience.  But will point out that people with similar experience predicted the space shuttle had less than one chance in ten thousand of a loss of life accident.  When you go into unexplored territory no one has experience and there are holes in knowledge.  1000 plus years is unexplored territory (and jets fighters are not just silicon chips).

[radiolistener]

Obviously you can design something to fail - that bit is easy, just under-specify capacitor voltage ratings, resistor power ratings, or over-volt your chips. But to design something that works for around a year and then fails?

Well, how would you do it? Electronic components aren't aware of the passage of time. You can't tell a resistor "work for 12 months, then burn out". Obviously you could install a software bomb if you are using a microcontroller, but that would be better described as sabotage rather than failure.

That's pretty easy and no conspiracy here. Just a business...
You can just skip some mandatory process steps during manufacturing electronic components or use more cheap materials. Modern technology can predict how it will affect lifetime pretty precise, with precision of weeks or even days.

So, you can optimize manufacturing process in that way to get planned lifetime  at minimal cost. And it appears that de-facto standard for electronics lifetime planning is about 1 year or even less.

The goal is to manufacture product at min price and sell it at max price and and avoid expenses for RMA during guarantee period.

Just a business... 

Regarding on how to make electronic components so short live, its not hard at all. For example skipping some steps during silicon manufacture can give you about a month of lifetime then it just fails and don't start anymore. Some technology like flash memory has very short lifetime by design.

Some modern sdcards can retain the data with no power just for one year and has so small endurance that you can get total failure within one week or even less with just aggressive using it for writing the data.

The issue with flash memory involve almost any microcontroller-powered device, almost all modern electronics. If you don't power it for years, it's just fails with zero time usage due to data loss on the flash memory.

Cheap electrolytic capacitors lifetime is pretty short and dramatically reduced if you use it at high temperature. Almost all modern electronics is compact, it means that there is bad ventilation and manufacturer may "optimize" the cost by removing some mandatory thermal design measures to use less and more cheap materials and components. As result, capacitors lifetime is dramatically reduced.

Not because conspiracy, just to reduce manufacturers expenses. Just a business...

As result the more modern technology give us a total garbage junk products which looks very good at a glance...

And unfortunately this issue affects not only consumer electronics, but also aircraft, space and military technology, just because almost all electronics become more and more cheap and turns into junk and there is no sense to manufacture good quality and long lifetime products, because it leads to bankruptcy. Officials helps to make things even much worse with regulations due to mafia corruption which they calling "green technology".  So even if you want to find good quality and long lifetime products, you can't. Due to official regulations and simply because it is not profitable from a business point of view.

That's why we have more and more aircraft failures and man-made disasters these days, even despite the very strict quality control management and the most stringent control processes in such areas like medical, aircraft, space, nuclear and military.

[Phil1977]

There are just much too many unknown variables to "target" the lifetime precisely to a RMA deadline or anything like this.

Just as examples, but not exhaustive:
- How much is the device in use? One person uses the USB-C charger only during a few short trips for the smartphone, the other one charges his laptop and e-bike battery on a daily basis with 100W.
- What´s the temperature? Is it used in Scandinavia or in the sun of the Mexican dessert?
- How much EMI backfires at the device?
- Is a PC used for hardcore gaming or for web browsing?
- Is a car raced or stored?
- Is a power tool used on construction sites or at a hobbyist?

If you would know all these issues and more, you could predict the lifetime in days. But you don't know.

Moreover: It´s easy to neglect all these points and build an unreliable product - but that will hurt your RMA rates. And last but not least, it´s not even true that the absolute worst case combination of the points above is the one to address. Sometimes also non-usage or too mild usage causes problems.

[tszaboo]

I'm probably not quite as cynical as you about the motives underlying their design decisions. But even if you are correct, we have wandered well away from the ageing mechanisms pertaining to electronic components, which is what the OP was asking about.

Here is a story for you, that will convince you that it's deliberate:
Mercedes solders light bulbs into light assemblies. They don't place in them in sockets, it's soldered into a PCB. Your light goes out, you are not allowed to drive, police pulls you over, you fail inspection. You have to replace the entire assembly. And it's a smart assembly with CAN serial number hard coded into it, so you cannot buy a second hand one, you have to drive to a an authorized service to have it reprogrammed.
A light bulb went out in my car, Toyota charged me 1.5 EUR to have it replaced, including work.
At Mercedes? Up to 2000 EUR.

You think electric cars will solve the issue because the engine is more reliable? Have I got news for you!

[Phil1977]

This damn stupid headlamp example...

A LED in a modern car headlamp is just an electronic component like all others in the car. Do you also want to put all IC´s in the car in expensive and unreliable test sockets so that you can easily swap them, just in case?

It´s the task of reliability engineers to make the LEDs last as long as everything around it. And yes, at some point a limit is reached and then so many things fail at once that it´s really not economic to repair it.

That´s the real target of lifetime optimization: To *equalize* the endurance of all components. If -some day- all components are equally worn out, then it´s time for recycling.

But coming back to the annoying headlamp example: The difference is: A halogen bulb is designed to fail early, so it must be exchangeable. Failing of a LED is practically the same as failing of a power transistor, a capacitor or anything else electronic. Usually it´s fault of the operating conditions that a LED runs hot and fails, but if the manufacturer ensures stable operating conditions then 10 000h to 100 000h of operation are feasible with light output of more than 80%.

[SteveThackery]

You can just skip some mandatory process steps during manufacturing electronic components or use more cheap materials. Modern technology can predict how it will affect lifetime pretty precise, with precision of weeks or even days.

So, you can optimize manufacturing process in that way to get planned lifetime  at minimal cost. And it appears that de-facto standard for electronics lifetime planning is about 1 year or even less.

Look, I'm sorry, but I just think that entire text is completely wrong and nonsense.  It seems like you can't or won't provide any authoritative references to support your claims or you'd have done that by now, so until you can do so it seems pointless to engage with you any further on these particular claims. I just want future readers of this thread to know that your assertions are in dispute by me. If you do provide the appropriate references, and I'm wrong, I will gladly say so.

[tszaboo]

This damn stupid headlamp example...

It's not an LED. I was very specific when I said lightbulb. As in 12 V, wolfram and glass.

[SteveThackery]

Here is a story for you, that will convince you that it's deliberate:
Mercedes solders light bulbs into light assemblies. They don't place in them in sockets, it's soldered into a PCB.

That's an excellent story and I agree that it does smell of planned failure. Not least because these are presumably filament bulbs? In that case their ageing mechanisms have been understood for 130 years or so. I believe normal filament bulbs have a designed service life of 1000 hours, although we don't really know if that applies to the Mercedes bulbs. I would think it does.

[Phil1977]

This damn stupid headlamp example...

It's not an LED. I was very specific when I said lightbulb. As in 12 V, wolfram and glass.

Sorry, my mistake. I´m just allergic to the story regarding LEDs because I hear it so frequently.

Soldered in incandescents really have a bad taste.

[tszaboo]

Here is a story for you, that will convince you that it's deliberate:
Mercedes solders light bulbs into light assemblies. They don't place in them in sockets, it's soldered into a PCB.

That's an excellent story and I agree that it does smell of planned failure. Not least because these are presumably filament bulbs? In that case their ageing mechanisms have been understood for 130 years or so. I believe normal filament bulbs have a designed service life of 1000 hours, although we don't really know if that applies to the Mercedes bulbs. I would think it does.

https://forums.mercedesclub.org.uk/index.php?threads/w211-rear-tail-light-bulb-replacement.142802/
Here are some people discussing it for their older cars. So they've been doing this for about 20 years now.
Remember the W124 from Mercedes, that lasts forever? Now we have this.

[SteveThackery]

1000 plus years is unexplored territory (and jets fighters are not just silicon chips).

On that count we totally agree. I would make no claims at all regarding 1000 year timescales - they are totally outside my, or anyone else's, experience.

Having said that, there are plenty of human artefacts older than 1000 years. Of course they don't have any electronics in them, but I mention it because it shows some materials can remain stable for at least that long, and very much longer. So although 1000 years sounds absurd for a jet fighter, maybe for simpler things it isn't inconceivable. For example, I definitely think it would be possible to make a WW2 aircraft last that long.  In effect they are infinitely repairable.

Which brings us to an important part of the 1000 years question: is maintenance and repair allowed during that time?

I think it's the advent of electronics that really started the throw-away world we now live in. Unlike everything that went before, electronic components are basically unrepairable. Yes, in theory it might be possible to open up a failed semiconductor and repair it, but it would be so difficult we can probably say it is effectively impossible.

[SteveThackery]

https://forums.mercedesclub.org.uk/index.php?threads/w211-rear-tail-light-bulb-replacement.142802/
Here are some people discussing it for their older cars. So they've been doing this for about 20 years now.
Remember the W124 from Mercedes, that lasts forever? Now we have this.

I do find this really interesting. Do you think that Mercedes deliberately build failures into their cars in order to generate a revenue stream from the cars throughout their life? It does seem an extraordinary thing to do, but it's obviously possible.

I just wonder why they aren't worried about damaging their reputation.

The move to non-repairable modules began decades ago, but I assume it to be a cost reduction strategy rather than a revenue generator. What do you think?

[Phil1977]

Maybe it´s just a side effect of the standardized qualification procedures in car industry.

Just assume one supplier got all the papers together to proof that its slightly undervolted and robustly made incandescents match the required failure rates for Mercedes. Then - in principle - it´s the right way for Mercedes to check these claims and make use of the cost savings that are possible without sockets and user serviceable compartments.

If it´s always good to blindly trust these qualification procedures is another question. Especially in the field of accelerated ageing tests there are a lot of uncertainties.

[SteveThackery]

I remember back in the late '70s or early '80s there was a stink that got mentioned in the media at that time. The new Audi had rear lights where the red plastic lens was not removable, so if you crunched it in a parking mishap you had to replace the entire unit - reflector, lens, bulb holder - and it was expensive.

I don't recall Audi offering any response, but it was a long time ago. I guess, but I don't know, that it was cheaper to dispense with the rubber seal and the two screws. Also, perhaps VW judged the higher repair costs could be borne more easily by the customers of their prestige brand, rather buyers of cheap VWs? I don't know - just speculating.

[coppice]

I remember back in the late '70s or early '80s there was a stink that got mentioned in the media at that time. The new Audi had rear lights where the red plastic lens was not removable, so if you crunched it in a parking mishap you had to replace the entire unit - reflector, lens, bulb holder - and it was expensive.

I don't recall Audi offering any response, but it was a long time ago. I guess, but I don't know, that it was cheaper to dispense with the rubber seal and the two screws. Also, perhaps VW judged the higher repair costs could be borne more easily by the customers of their prestige brand, rather buyers of cheap VWs? I don't know - just speculating.

And now in the time of super expensive beam forming LED and laser headlights they ultrasonically weld all the light modules shut, and charge $1000 or more for a new one. Well, at least the ultrasonic welding reduces the chance of rain leaking into the light, but they sure aren't reliable. The number of Mercedes and Audis with faulty LED lights is appalling. I guess a lot of them are being replaced under warranty, but beyond that?

[coppice]

I do find this really interesting. Do you think that Mercedes deliberately build failures into their cars in order to generate a revenue stream from the cars throughout their life? It does seem an extraordinary thing to do, but it's obviously possible.

I think they just focus on a market where most people lease rather than buy. The initial buyers don't expect to keep the car beyond its warranty period, so they don't worry about the cost of fixing it, only the frequency of failures. This seems to be reaching a limit right now, though. Poor resale values for the leasing companies disposing of cars seems to be breaking this business model. Maybe they will have to swing back to things being replaceable in small, more reasonably priced pieces. There has always been a dynamic like this in the car market. If you sell huge numbers of cars in a market, and don't make small replacement parts available, third parties will enter the market to fix that. If you only sell a small number of cars in a market you keep your inventory of spare parts small, by stocking a few large chunks. Before BMWs were popular in the UK, dealers would never sell you things like a bearing, but a big seller like Ford would. You had to buy a big chunk of a BMW that included the bearing. We used to joke about their 2 spare part model - the front half of the car and the back half. Now BMW are a volume seller in the UK people can get things replaced in smaller increments.

[SteveThackery]

I think they just focus on a market where most people lease rather than buy. The initial buyers don't expect to keep the car beyond its warranty period, so they don't worry about the cost of fixing it, only the frequency of failures.

Very good point about leasing - it does change the dynamic.

[Phil1977]

E.g. BMW has returned to swappable headlamp glasses BECAUSE of the leasing contracts.

At least in Europe we have a lots of B2B-leasing contracts that include all insurance costs. Especially for these contracts BMW has recognized that repair costs must not be neglected. If repair costs are high then the B2B-customer must cut down on vehicle costs or vice-versa. So many parts that easily brake in small but frequent accidents get also designed for cheap replacement - at least for the high-volume models like 3 and 5 series.

I know this story first hand from BMW but I´m quite sure it´s (mostly) the same for other brands.

[Howardlong]

Coincidentally, something new on my radar is that class II MLCC ceramic caps such as X7R, Y5V & Z5U degrade over time.

We all know about the DC bias derating of MLCCs, but I wasn't aware of the aging issue.

This aging varies significantly not only by dielectric, but also by manufacturer.

Thankfully most of these caps are used in non-stringent applications such as AC coupling and DC decoupling, but there may be latent failures in waiting in marginal designs.

https://youtu.be/yLuWcTo2cb4?si=A9nzO6_W9xLe6vnY

[coppice]

E.g. BMW has returned to swappable headlamp glasses BECAUSE of the leasing contracts.

At least in Europe we have a lots of B2B-leasing contracts that include all insurance costs. Especially for these contracts BMW has recognized that repair costs must not be neglected. If repair costs are high then the B2B-customer must cut down on vehicle costs or vice-versa. So many parts that easily brake in small but frequent accidents get also designed for cheap replacement - at least for the high-volume models like 3 and 5 series.

I know this story first hand from BMW but I´m quite sure it´s (mostly) the same for other brands.

Are they doing anything about the ultrasonic sensors, radars, cameras, and so on, which have been driving up insurance repair costs?