23
\$\begingroup\$

I am part of my company's work to become more environmentally friendly and sustainable. On a high level for the company that means pursuing ISO certifications (such as ISO 14001) and working according to other standards. But I'm trying to figure out what this should mean for me, as a electronics developer?

So far I'm looking into switching PCB manufacturer to a company that regularly produces sustainability reports that I think are transparent and good, adapting PCB designs to minimize waste materials, improving the recycling/waste sorting in our lab, and perhaps using OSP instead of ENGI as surface finish. But what else is there that I can do as a developer? Are there components I should stop using in my designs, standards/certifications that are suitable to follow as a developer, or anything else that you have done yourself or could recommend me to look into?

\$\endgroup\$

4 Answers 4

24
\$\begingroup\$

You may want to look at recommendations from the circular economy ... perhaps starting at the Ellen Macarthur Foundation

The underlying idea is design with a view to repair, recycle and re-use. Arcatus' answer hits one key point ... reliability reduces the volume of junk to be recycled.

But there are two aspects to reliability ... avoiding failures is only one, but the easiest for the designer alone.

Ease and speed of repair is another, which requires a systemic approach different from some modern companies. There's no point making a subassembly replaceable if it's inside a case that's glued or welded shut. And there's no point if the company has no spares supply chain or no commitment to keep spares available beyond the guarantee lifetime. I touched on this in another Q&A.

To illustrate the systemic approach : I replaced my car a couple of years ago; the roof bars are adapted to each car model with a relatively cheap and simple fitting kit. Great design! However when the manufacturer introduced a new system, they immediately obsoleted the fitting kits for the older system. So, not like that...

In the kitchen, I found I can cheaply and easily replace the controls and heating elements on a 30 year old stove. A $10-$20 part, a few screws, spade clip connectors, job done. Microwaves, not so much.

Going back to Arcturus' answer; he deprecates electrolytic capacitors, with good reason. But if you can't eliminate them, consider moving them along with all the hot power supply components to a second cheap PCB, replacable separately from the main unit. (You have to factor in the reliability of connectors though; it's not an open/shut decision).

Or farm the lot out to an external PSU, like a laptop brick or USB phone charger. It's ideal since it's easily accessible, replacable, and pretty standard.

Design either for zero failure rate, or easy access, easy repair, and good spares availability (where possible, using standard or very common parts).

Another example : this laptop (after seven years) has a new power supply, a new (and greatly improved) disk, and new and bigger memory. Maybe I'll get another seven years out of it... The case was designed with 3 screws and subtle disassembly hints moulded into the underside, making it a dream to work on compared with some.

There is room to be imaginative : for example, while you can't realistically 3D print components for a production run, you might consider open sourcing a design at the end of its maintenance lifetime so that 3rd party spares can be produced. (This is being done for at least one model of pre-WW2 lathe!) A systemic approach might put together a suitable archive at design time, for public release after production ends, to allow lifetime maintenance. (Of course it may not cover everything, for IP protection purposes)

I don't know of ISO standards for these ... yet ... they may emerge from the aforementioned circular economy project.

\$\endgroup\$
7
  • 1
    \$\begingroup\$ Great points. I actually meant to include reparability as well, but my post was already quite ranty, and boy, does that topic make me itch, so I took it out so I could stay on topic. Thanks for adding it, and phrased much more calmly that I could have done. I am watching the EU right to repair initiative with interest, but it is shocking how corporations are lobbying hard against the legislations but at the same time branding themselves as "green" and ecofriendly. It is business, off-course, and it illustrates how bad it is for business to actually be ecofriendly. \$\endgroup\$
    – Arcatus
    May 18, 2021 at 12:18
  • 4
    \$\begingroup\$ Wouldn't that be nice? Full blueprints in an accessible form placed into a form of escrow to be released should the manufacturer go under or fail to support the product for a (long) adequate period. But more realistically, we can all design for repairability \$\endgroup\$
    – Chris H
    May 18, 2021 at 15:09
  • 1
    \$\begingroup\$ @ChrisH Escrow might be pushing it a little, at this stage. But a corporate "honour system" where a company declares : we have no more interest in producing and selling this design. We have spares for X years; after that, if you love it enough to want to keep it going, here are the drawings, good luck". See Jaguar for how NOT to do it, right now. thetimes.co.uk/article/… \$\endgroup\$ May 18, 2021 at 17:44
  • 3
    \$\begingroup\$ Since we are producing fairly expensive equipment towards life science research, and work according to "customer is king", I actually wouldn't entirely rule out that we could have some form of policy that all design documentation for legacy products not longer being sold or actively maintained could be released. Perhaps. I'll work towards it. \$\endgroup\$ May 18, 2021 at 18:03
  • 1
    \$\begingroup\$ @user_1818839 Do you perhaps own a Thinkpad? \$\endgroup\$
    – Tejas Kale
    May 19, 2021 at 20:10
16
\$\begingroup\$

Reliability?

There is an entire field of expertise on building robust electronics. This requires vigorous testing, and thorough design. This will be more expensive and increase time-to-market. But avoiding defective products should be on the agenda. (Period. Not just "avoiding defects until the warranty is expired")

But: this is exceedingly rare. While it is possible for a commercial product (where lifetime is estimated to a few years) to be built using industrial techniques (where a few decades of lifetime is estimated) I have never, ever, seen it. For obvious reasons: Cost.

My TV could have had a conformally coated PCB. Or ceramic capacitors instead of electrolytics. But where is the business in that?

But I'm going off on a rant here. To answer your question: as you seem genuinely interesting in making environmentally friendly electronics, then:

  • De-rate components, and read up on such practices.
  • Avoid or at least cherry-pick electrolytics and other components prone to failure.
  • Use IPC class 3 where it makes sense. (Some of those requirements can skyrocket the cost, so I am not recommending you implement practices where you shoot your foot off. Consider Class 3 a guide more than a requirement)
  • Conformal coat your boards.

I would gladly purchase a dishwasher advertised as as "Developed using HALT techniques" even if it was more expensive. But obviously I am weird, and unfortunately the company selling those would be bankrupt within the year.

This is probably not the answer you were looking for: if you want to be environmentally green you are going to be more expensive than your competition, and you are (per design) going to sell less. It's a sad truth.

\$\endgroup\$
4
  • 2
    \$\begingroup\$ Thank you a really good answer! Increased quality and reliability is of course something we are pushing for, even outside the scope of being more sustainable. But you include some very good tips for achieving reliability beyond "test & verify everything" which is the usual advice when pushing reliability. I do however work within the life science field, so these actions you list are most likely not at all an issue to get through, cost is rarely an issue for me when it comes to PCB design at my company. Which is an insane luxury. \$\endgroup\$ May 18, 2021 at 9:38
  • 1
    \$\begingroup\$ Absolutely. Instead (or let's say in complement) of issuing stricter and stricter power consumption rules, the regulatory bodies should mandate longer warranty times, at least for products that don't auto-obsolete themselves quickly. This will help environmental compliance in all sectors but reduce GDP. Capitalism is not amused. \$\endgroup\$
    – tobalt
    May 18, 2021 at 11:46
  • 2
    \$\begingroup\$ I'm not sure about conformal coating. For some products maybe, but the board is then not repairable (at least to a good approximation). The coating itself isn't recyclable (at least last I looked) and will reduce the admittedly limited recyclability of the board. The OP works in a very different scientific field to mine, but here we certainly keep things going for longer than they were ever designed to by modification as well as repair. \$\endgroup\$
    – Chris H
    May 18, 2021 at 15:04
  • \$\begingroup\$ @ChrisH True, Conformal coating might come with it's own environmental hazards, including the chemicals used. But I think most, if not all, I've come across is solderable; you simply burn trough them. It may make repairs more difficult, but it prevents a broad spectre of undesired effects - reducing the chance that a repair is needed in the first place. \$\endgroup\$
    – Arcatus
    May 18, 2021 at 19:53
8
\$\begingroup\$

Here are a few more points. I approach this as a former systems engineer, so these deliberately look at the big picture rather than details of electronic component selection etc., but this needs a joined-up approach. One of your comments hints that you're thinking about rather specialist design/production, as I used to be involved with.

  • Anything cut to size in manufacture should be sized/shaped to maximise yield form your manufacturing process and minimise offcuts. This could mean PCBs, but it could also mean cables. This is a job for production engineering and not just design, but avoid things like (in a former job of mine) having 2m of cable on the BOM, only to discard 0.7m in assembly by the time it was terminated and cut to final length.

  • Consumables in production should be looked at too - hand built kit often ends up using rather a lot, whether it's disposable gloves or cleaning supplies.

  • Housings often account for a significant proportion of the material in a product. These should be easily recyclable. The finishing processes should be considered too - solvent use/evaporation for painted finishes can be significant for example.

  • Your broader shipping/packing processes can be as important as your product design - design the product considering the possibility of using all recycled/recyclable packaging (cardboard box inserts rather than moulded foam, but this can require a simple product shape.

  • Similarly sending out external power supplies and mains cables with every order may or may not be appropriate, but sending 3 different mains cables certainly isn't. It's not as common as it was, but I've bought kit where the majority of the bulk/weight in the box for a small product was destined for the bin because of the 2 useless additional mains cables (US, EU version when I'm in the UK) and associated packaging.

  • On a related note, using standard interfaces can help cut cable waste. USB has helped with this but a few manufacturers still use unusual connectors with no good reason. Interconnects between parts of your kit are another matter, but shouldn't be too rare in case repair is needed in the future.

  • For big runs, custom tooling can save waste, but for short runs less so; there's a judgement to be made about whether to use mouldings or not, but also for test: In an old job we had some lovely hand-made wooden jigs/test stands for products built in quantities of a few tens per year, that saved time, effort, and metal/plastic.

\$\endgroup\$
7
\$\begingroup\$

I think you should make it win-win, both for your company and the customer. Some ideas:

  • Lower power use

This is marketable, and translates directly into cost savings for the customer. Especially if your product uses batteries, but also important for mains. This is both a design aspect: can you shut down some hungry bits of circuit when not in use? ...and a usability aspect: it should come out of power-save mode transparently and quickly, like my laser printer. It heats in three seconds and then prints. Unlike the old ones where the heater was running all the time to stay ready, and you had to leave it on, because otherwise it took forever to get ready.

A display that says "power save" is a plus, makes the user feel good.

  • Lower embedded energy

This one is much more difficult, because who knows what the embedded energy of a chip is?

  • Reduce use of materials of dubious morality

Tantalum comes to mind, since it is mined by slaves.

  • Increase reliability

This is a no-brainer, especially since you say you make life sciences equipment. Thermal design is important, and it should not be ignored just because marketing says "make it thin and slick".

  • Make it easier to repair

I'm thinking about Apple, who specifically takes every step to make repair impossible: unobtainium custom chips, no spare parts, everything glued together, top secret schematics, authentication between parts so you can't swap them between two devices, fragile flex PCBs that break, etc. I recommend youtube channel "Louis Rossman".

I'm not saying you should make everything available, but, say, a product that uses an off the shelf switching power supply is easily repairable when the power supply fails, versus a custom SMPS design on the mainboard, where you can't swap it out. Or helpful silkscreen on the board, labeled test points, stuff like that. Even better, don't hide the self-diagnostic menu behind a password.

Basically, if the customer has two non-working devices, they should be able to strip them and make one working device. That's, like, insurance for the future, in case something happens, like a pandemic.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.