14

This question is summarized from a discussion on the Electrical Engineering StackExchange chat.
It is reproduced here in hopes to generate a discussion with long-term value

Question for anyone with experience with long (> 30 years) life cycle product design:
Do you generally design keeping in mind possible obsolescence / discontinuation of key parts within the life-cycle of the product?


Specifically, if a pin-compatible equivalent for a specific IC is no longer made, how is that planned for? Not every semiconductor manufacturer does lifetime buy notifications by package, and some manufacturers simply fold up and disappear.

The context here is of a product for which my client has over 100k PCBs in inventory, and more than 2 million devices deployed over 30 years. A couple of the key parts used on the board no longer exist, and near equivalents are all SMT. All ICs on the original boards are DIP and socketed. Some of the ICs in question are obsolete analog continuous-time signal processing parts, the rest are digital logic and thus easily substituted by equivalents, ASICs or MCUs depending on complexity.

There's a repair workstream (industrial product, 20 to 30 year serviceability warranties), and there is a production workstream (repeat orders, thousands of boards per year).

Respinning the board, while an ideal suggestion, is not an option in this case, as the Purchases department of the end-customer would consider a base PCB change as a "new product", hence requiring evaluation of competing vendors, and renegotiation of the contract - This will trigger a fresh tendering process and potentially loss of 10x millions worth of annual business for my client, to a competitor.

Current repairs of customer-site devices are all done by manual reworking in the field, the device is not allowed to be brought back to a workshop. Replacement of boards does happen, but the "new" replacement board absolutely must be identical in PCB layout to the one being replaced, as the end customer does not take kindly to any changes.

A proposal being considered, though yet to be validated with the end customer's purchases team, is replacement of DIP ICs with identically sized little PCBs with pins, plugged into the DIP sockets on the main board. This is intended to reduce field-work risks and time.


So, back to the question: What are practical EE experiences in planning for such product lifecycles and associated challenges? Great ideas for "next time" are also welcome.

starblue
  • 6,462
  • 2
  • 21
  • 35
Anindo Ghosh
  • 50,188
  • 8
  • 103
  • 200
  • Why is it so important for this customer that the PCB does not change? – jippie Dec 09 '12 at 08:34
  • 4
    I know what 30 year old electronics look like, I know what electronics look like today. I have no clue what electronics will look like in 30 years. I would replace the guy who agreed with the 30 year support contract that does not allow for hardware updates. – jippie Dec 09 '12 at 08:37
  • 5
    I always liked this comment on [sparkfun](http://www.sparkfun.com/news/350): "Microchip PIC on all ranges is amazing at supply. I heard a story of a guy who made a product on a very old PIC16C for controlling the airline food carts. The part was very old and out of all the disties. To avoid him re-doing expensive approvals Microchip pulled the old recipe and the masks and ran him a lifetime supply. Now argue all you want about anything technical – but that’s as real as it gets. We use PICs a lot now." My plan would be find companies with good track records and hope they don't disappear. – Kit Scuzz Dec 09 '12 at 09:09
  • @KitScuzz so basically what you are saying is to buy a lifetime supply once you are about to get product approval. – jippie Dec 09 '12 at 09:14
  • 1
    @jippie : Huh? The quote was about how Microchip made a new run of the chip way after the part was obsolete so that the maker could continue to support an old product. I would say that if you make it to the point that Anindo Ghosh mentioned (you already have millions of parts that require support in customers' hands), and you have a manufacturer willing to make a run of an old part for you, buying a lifetime support supply makes sense. The point is: pick a manufacturer who with support targets as long as your own, and hope they still exist when you need them. – Kit Scuzz Dec 09 '12 at 10:04
  • 1
    @KitScuzz My thought would be to prevent the situation where you have to rely upon a supplier to be willing to do a custom run of an obsolete product. I think it is cheaper to buy in advance than to beg for an extra run. – jippie Dec 09 '12 at 10:26
  • 1
    Can you tell us what the product is or give some clue if direct description is not possible? The "may not be brought back to base" requirement makes it sound like a military or Casino app or a high security one. – Russell McMahon Dec 09 '12 at 12:26
  • 3
    It sounds like the client has a bigger problem: They did not put the effort into keeping the product line up to date as technology improved. If they had, losing business to competitors wouldn't be an issue; they would win on backward-compatibility alone. – Dave Tweed Dec 09 '12 at 13:15
  • 1
    The client sounds unrealistic to me - they really expect to have the design last until the end of time? Either they pony up for a redesign or deal with the reality that parts are simply unavailable... – dext0rb Dec 09 '12 at 15:58
  • 2
    @jippie Regarding firing the person who agreed to a 30 year support contract - He's a hero, he turned a small company he started when I was 10 years old, into a multimillion-dollar business on ***just one contract***. He's also my potential client, so no, I'm not firing him :-) – Anindo Ghosh Dec 09 '12 at 17:27
  • @KitScuzz A couple of the ICs have actually been "refabbed", by replacing them with custom designed, DIP packaged, "bespoke system on a chip" equivalents; Others have been replaced with newer versions of the silicon specially packaged by manufacturers into DIPs. Some manufacturers are real nice that way. – Anindo Ghosh Dec 09 '12 at 17:31
  • @Jippie Regarding lifetime buys - the expected roll-out had been 1 million units in 30 years. That number was crossed due to a huge, impossible-to-predict, project scale-up by the end customer, 10 years ago. End of stockpile! – Anindo Ghosh Dec 09 '12 at 17:33
  • 1
    @DaveTweed My potential client actually does have several way way better alternative products, proven and already deployed at other customers... However, the kind of tender in question is rarely won on merit alone, there are unmentionable other considerations - **unless** the vendor is the only one that can meet a given requirement. This was true 25 years ago, hence the contract. This is no longer true today, competitors can at least match the end-customer's specification, and said competitors have a track record of *interesting* business dealings. – Anindo Ghosh Dec 09 '12 at 17:36
  • @RussellMcMahon Excellent insight! Broadly, the application involves audio and RF spectrum analysis and reactive actuation on pattern matches. High security deep-deployment device, non-military. The functionality is trivial to achieve through modern technology even at a DIY level, incidentally - and the end-customer is not unaware of this, but they typically never rock the boat if things are working as per signed terms of business. – Anindo Ghosh Dec 09 '12 at 17:42
  • 1
    @dext0rb Actually, the end customer is happy to continue buying for another 7 years, and will not go through a tendering and re-evaluation process unless something breaks in the process as originally penned in the contract. So, yes, the vendor definitely is happy to be unrealistic about letting the ancient design live for this one customer. I would be, too, if I could have a mere 1% of the *profits* alone, year on year. Let me see, would I like my second Lamborghini in red or black? – Anindo Ghosh Dec 09 '12 at 17:46

3 Answers3

7

Handling component obsolescence for 30 years is a difficult task. Even 10 years is amazing for anything but the most simple of products.

Simply put, suppliers and manufacturers are unreliable over a 30 year span. As mentioned before, they will stop making chips without notice or simply cease to exist. There was one manufacturer about 6 years ago that had a fire at one of their older factories. Rather than rebuild and start making those chips again, they simply obsoleted all the chips that were made there.

Despite everyone's best intentions, parts become unavailable.

If a PCB respin is not allowed, then the only reliable solution is to stock a 30 year supply of all parts. This is more complex than it sounds because some parts have a limited shelf life. Parts would need to be stored long-term in temperature and humidity controlled storage, and even then some parts will still not last 30 years. Storing components for 30+ years will require the consultation of the manufacturers to determine the ideal storage conditions. Documentation on how to properly store and solder the parts would need to be archived. For example, if a bake is required before assembly then those instructions would need to be saved.

If a respin is allowed, then the ideal solution is to archive all design data and write additional documentation to facilitate redesigning every 4 to 10 years. In my opinion this is the best solution (assuming a respin is allowed). It allows the product to evolve over time to accommodate new technology and new manufacturing processes. This would either lower the cost of the product or give the new versions new features and capabilities.

Chris Laplante
  • 2,151
  • 1
  • 24
  • 32
  • +1 This is very insightful: You've perfectly summarized my own discussion with the client a few days ago. Unfortunately, some of it could not be predicted, so it did not happen. At this stage, I am at a loss on whether I can help make things better at all. – Anindo Ghosh Dec 09 '12 at 17:49
6

I haven't worked on anything that needed a 30-year product life, but I have worked on products that went into semiconductor manufacturing equipment, and needed 10+ years of support.

At least in those cases our customer was sane enough to realize that sometimes a slight redesign is the most cost effective and reliable solution.

Not every semiconductor manufacturer does lifetime buy notifications by package

Every reputable semiconductor manufacturer will give notifications on a per-part-number basis. If there is a manufacturer out there who doesn't do this, I wouldn't purchase from them for any project, let alone one with your stringent support requirements.

Reputable manufacturers don't just notify if a part/package is going obsolete, they notify for any change in the product that might impact quality (in the ISO 9000 sense). For example, they notify if the lead finish is changing from tin-lead to matte tin. They notify if the device packaging process is being moved from one factory to another.

some manufacturers simply fold up and disappear

This, or course can't be helped. And there are other similar risks.

David mentioned a case where a factory fire resulted in product obsolesences without notice. In another recent case I know, a (large and extremely well-known) fabless semiconductor company was forced to obsolete several lines of products because their foundry vendor decided to close down an older (and no longer sufficiently profitable) manufacturing line. At least in this case they were able to give advanced notice and allow for lifetime buys...but their products were not $0.05 transistors, they were complex $20 - $200 ICs, so you can imagine the investment required for some lifetime buys.

Other risks -- regulation

One more thing to watch out for is the march of environmental regulation. The European RoHS directive and similar laws may not apply to your products, but they affect your supply chain. RoHS, for example, forced a lot of mature products to be redesigned, reducing sales volume for a lot of mature components, which probably resulted in some obsolesences as specific parts became unprofitable.

New versions of RoHS are expected within a few years. So you can expect a new wave of obsolesences as the market settles out in relation to the new rules.

Other risks -- changed market landscape

If you designed a product 30 years ago and you designed it entirely with multi-sourced components like 2N2904's and 74LS04's, you would very likely still be able to get all those parts today.

But the trend in the last many years is away from multi-sourced components. Very few new products are being replicated across manufacturers. And the parts that are are at similar complexity levels to what was available 30 years ago --- hex logic gates, individual transistors, etc. Even "simple" devices like linear power regulators are now complex enough that nobody tries to duplicate another company's design exactly.

If you want to design in a microprocessor or a programmable logic device, you are simply stuck with a single-source component, and all the risk for long term support that entails.

Also, as my anecdote above pointed out, the rise of the fabless semiconductor company also adds risk because it means your chip vendors may not actually control their own manufacturing resources.

Strategy

To me, it sounds as if your sales team needs to get tougher with this customer (of course, engineers always say this).

One option is tell them you can fix the problem their way, but to maintain this level of support, the product price will be increasing XX% each year for the remainder of the product lifetime. You can make a lifetime buy for these obsolete products, but you'll need to pay for engineering evaluation of the requirements, technicians to manage the inventoried parts, etc., and you need revenue from these products to support that.

Another is to call their bluff on the re-tender process --- there can't be that many competitors out there who could realistically support these lifetime requirements, and those that can will price accordingly. On top of that, your team has experience with this particular application and the customer's detailed requirements, giving you an advantage when it comes to bidding the project. Who knows, you could come out of a re-tender with a contract at a substantially higher sales price.

The Photon
  • 126,425
  • 3
  • 159
  • 304
  • +1 for some excellent structured thought, which I will shamelessly reuse if I choose to proceed as consultant on the case! (A) The "Sales team" at the time of the contract was one fresh-out-of-college engineer who could demonstrate a product nobody else offered back then. He owns the company he started, BTW. (B) Given the profit margins on the product - of the order of 800%, I've guessed from some remarks - I don't think anyone is fussed about pricing, (C) No, if it gets re-tendered, international vendors will "gift" their way into the contract for sure. My client has compunctions. – Anindo Ghosh Dec 09 '12 at 17:58
  • Also, the actual device is so simple by today's technology standards, that it can be built at home using merely toner transfer, single sided boards and medium pitch SMD components. So yeah, the "competition" is somewhat ubiquitous. – Anindo Ghosh Dec 09 '12 at 18:37
  • @AnindoGhosh, building one with toner transfer is one thing. But nobody's going to build and ship 70k units per year that way. In this case you're not selling unique technology, you're selling your track record of support, reliability, understanding of the application, etc. (But yes I read your other comments and I see how you're in a bind here). – The Photon Dec 10 '12 at 05:28
  • The toner transfer example was just to indicate the simplicity of making a competing device to the same *functional* specifications. :-) The problem lies elsewhere, as you have noted from my comments. – Anindo Ghosh Dec 10 '12 at 08:55
2

In automotive industry obsolessence is a curse. Most vendors do not support beyond 10 years (might sound crazy to commercial/industrial world). We have several designs that are more than 10 years in service usage in small volumes. Even a small component could trigger a redesign that would cost millions of dollars to design and validate. When the technology is obsolete it is even harder. e.g.: Ceramic Substrate with an obsolete micro controller.

There are companies like Rochester Electronics who handle old parts and even have capability to re-manufacture parts that are donkey years old. Tekmos is one of them too (they handle Frescale more).