Why are Western companies getting offline power supplies designed in China?

[free_electron]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

[tooki]

treez, you have a knack for asking questions without ever considering the answers yourself — or pondering the baseline plausibility.

With regards to power supplies, look at Apple, the world’s most valuable company. Apple has the talent in-house to design its own damned CPUs, whose year-over-year performance improvements are beating the crap out of the competition. Yet if you buy an Apple product, its power supply will be made by any number of power supply specialists, including Delta, TDK-Lambda, and Flextronics. Apple could easily afford to hire power supply experts to design their own PSUs, and then just have a contract manufacturer build them. But they’re not, they’re even designed by the PSU companies, as evidenced by having multiple suppliers for most of their power supplies, whose actual designs vary by supplier. (Meaning that Apple just gave them mechanical, electrical, interface, and performance requirements and let them figure out implementation. I just replaced the internal power supply in my sister’s Time Capsule after the PSU died from water ingress. The replacement PSU was made by a different company, with a completely different circuit and totally different layout. But it’s still a custom part that is functionally interchangeable.)

If not even Apple, with its massive engineering resources and endlessly deep pockets, bothers doing power supplies in house, do you really think there is any  chance it makes sense for you to?!?

[rstofer]

In the govt agencies where I worked, promotion was obtained by competitive examinations, & if you then applied for a higher position, after tentatively  being selected, were still subject to challenge by others with equal or better qualifications and/or seniority.

For the US...

I was thinking more of promotion at the managerial level.  Promotions within the working group are whatever they are, usually  by changing employers.  In the US, public service jobs are all unionized, promotions are more often based on 'time in grade' and not merit.  But the employees have a defined benefit retirement plan, something private sector employers are abandoning.

Our public utility just found a way to cut their labor costs in half.  They invented a shovel that will stand up by itself!

One could argue that our education system (in the US) is a government run operation.  A local school board and a state regulating agency overwatched by a federal agency control everything that happens.  It is an abysmal failure!

[rstofer]

If not even Apple, with its massive engineering resources and endlessly deep pockets, bothers doing power supplies in house, do you really think there is any  chance it makes sense for you to?!?

Not building PSUs is one of the reasons their pockets are so deep!

The concept of 'core competency' runs rampant through businesses.  It is usually used to jettison overhead departments (facilities, payroll, janitorial, etc).

Designing a PSU could be a 'core competency' but it doesn't have to be because there are multiple vendors who do have that ability.  Just tell the vendors what you want and select the low bid (or not, there are reasons to avoid the low bidder).

[coppercone2]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

This is how its done but you can't seriously tell me as a designer you like it this way. I want to scrutinize every aspect of my product to make it solve the intended problem as good as possible. I even feel this about IC's, custom made asics to your specification will always be a better solution.

[free_electron]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

This is how its done but you can't seriously tell me as a designer you like it this way. I want to scrutinize every aspect of my product to make it solve the intended problem as good as possible. I even feel this about IC's, custom made asics to your specification will always be a better solution.

yup. i like it that way. i can't get myself to figure out the turns and wire gauge to make an smps transformer ... and then building it, testing it. getting the empty bobbin alone is a drag ...
go look at the prices at distributers for those things. i can get the whole supply made in china for that price. From reputable manufacturers like antec and delta and meanwell and others
like it or not but making a good SMPS is NOT easy ! especially the transformer part is a specialty.

it can be done in the west. XFMRS inc has good transformer building service. and you can get competitive on price for certain specialty apps. but for your run of the mill power brick ? buy it ready made.

[TimNJ]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

This is how its done but you can't seriously tell me as a designer you like it this way. I want to scrutinize every aspect of my product to make it solve the intended problem as good as possible. I even feel this about IC's, custom made asics to your specification will always be a better solution.

I'm still not sold on the idea that Widget Company XYZ can build a power supply better than Power Supply Company ABC. And, even if they can, I'm still not sold on the idea that they should. Yes, some companies may have the talent to build custom power supplies for their widgets, but again, why do it? Like many things, power supply design is kind of an art, and can be challenging to do right.

Yes, power supplies aren't perfect black-box devices. They have quirks and peculiarities that might cause you to scratch your head occasionally. But generally, they do behave pretty well.

[tooki]

If not even Apple, with its massive engineering resources and endlessly deep pockets, bothers doing power supplies in house, do you really think there is any  chance it makes sense for you to?!?

Not building PSUs is one of the reasons their pockets are so deep!

The concept of 'core competency' runs rampant through businesses.  It is usually used to jettison overhead departments (facilities, payroll, janitorial, etc).

Designing a PSU could be a 'core competency' but it doesn't have to be because there are multiple vendors who do have that ability.  Just tell the vendors what you want and select the low bid (or not, there are reasons to avoid the low bidder).

Yep, absolutely!!!!

There’s also the related concept of comparative advantage, of doing only the things where you are significantly better than others. I’ll never forget the analogy my economics professor used: Suppose you’re on the starship Enterprise. The decks need scrubbing. An ensign can scrub the deck in 4 hours, while Spock can scrub it in just 1. But Spock has science and telepathy talents the ensign lacks, so Spock’s comparative advantage is in sciences and mind melds, which the ensign can’t do. Even though Spock could scrub the decks, the opportunity cost of doing so is enormous, so it makes sense to have the ensign scrub the decks. (Or four ensigns, if you need it done quick.)

[sokoloff]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

This is how its done but you can't seriously tell me as a designer you like it this way. I want to scrutinize every aspect of my product to make it solve the intended problem as good as possible. I even feel this about IC's, custom made asics to your specification will always be a better solution.

Honestly, I do like it this way. I'd rather ship than obsess over minute details of a commodity element of my design... If you need an aluminum enclosure, do you start by mining bauxite as well?

[ocset]

Thanks, yes , I understand what you all mean….

I'm still not sold on the idea that Widget Company XYZ can build a power supply better than Power Supply Company ABC. And, even if they can, I'm still not sold on the idea that they should. Yes, some companies may have the talent to build custom power supplies for their widgets, but again, why do it?

…sure, if you live in a country  full of companies that make great widgets, then you don’t want to bother with specialising in general power supplies for those widgets.
What I am saying is, that for countries like  UK that don’t  have widget making companies in any  significant quantity, then UK (or others) should  start out by starting a power supply company, -one that can supply  “widget making companies”. The power supply company  should be government  run, because in countrys like UK,  the private sector has totally dropped the idea of a  “power supply company”  and wont do it. In fact, in UK, the private sector has gone on an orgy of selling off  UK owned industry to overseas buyers.
The reason the government run company should be a power supply one, is because its relatively straightforward (compared to eg making radar equipment), and  , I believe, can be done as cheaply as the Far East…or at least, nearly so. Also, engineers who work in it could eventually leave and go and be power supply engineers in widget companies. Widget companies, I find, very often need their first prototype widget to have a custom made power supply for it….so the ex-gov’t_run_PSU company engineers could go and do that kind of stuff. Who knows, maybe  some engineers could start  up a private power supply company and put the government  run one out of business…..thats an excellent scenario because it means the electronics sector has been stimulated.
But in the first place, there is a need for a government  run company to get the ball rolling. Uk (and other countries) cant attract  youngsters to come into electronics because there are too few  electronics companies around…..so government run ones need to be started. The government wont be running it as such……it will be run by engineers employed by the government.
Also, the UK should declare that if anyone in UK starts up a power supply company, then it will be illegal to sell it off overseas. 66% of uk manufacturing industry  (>500  staff) is now foreign owned, and the UK now has not enough companies to  pay off its enormous national debt. The uk will be a third world country within a decade, unless radical measures are taken....
https://massey276.wixsite.com/ukdecline

..other countries too, can look out for this, because politicians have a habit of  nuzzling up to huge companies and making decisions as a result that jeopardise their own country....eg the "enterprise act 2002" in UK  that made sure that the UK govenrment now has no duty of intervention when uk companies are  about to be sold overseas......The sale of UK companies  to overseas  buyers ramped up dramatically after this legislation was passed.

[mikeselectricstuff]

The West has a shortage of engineers.
It would make no sense for them to sped their time reinventing the wheel, rather than working on more innovative and profitable products. A PSU is something that you just buy from people who have experience and infrastructure to do it.

[ebastler]

@treez -- Fundamentally, your problem seems to be that you (personally, and/or the company you work for) are active in a field which the Chinese are perfectly capable of addressing, and which is a large enough market to be attractive for them to address. LED drivers and power supplies seem to be that general field, and your personal area of expertise.

Your response to this, and the topic of many similar-flavoured threads you have opened here, is that government regulation should somehow force the public and private sector to bring this business back to the UK. Honestly, I don't think this is ever going to happen.

Instead, you should think about ways to differentiate the products you develop: Add some unique technical twists, or tailor them to niche markets which are too small for Chinese competitors to bother with. (Or which the Chinese just are unaware of, since they are not close enough to these customers.)

This may require skills which you personally don't have -- digital design maybe, e.g. to add remote monitoring/remote control capability to your LED lighting. Or the skill to listen to potential customers with an open mind, to figure out their unmet needs. Maybe you can develop some of these new skills yourself; maybe you are better off working for a company which has (or can bring in) these skills, and which then needs your "core" analog expertise to design the complete product.

Either way, establishing yourself in product segments which evade the Chinese competition is the way to go, in my opinion. Asking for government action to force customers or companies to "buy British" or "buy European" just won't work.

[BravoV]

Ask an average person, does an ambulance is important/crucial to a hospital ? Doubt anyone will deny it.

But, ask again, does "every" hospital need a full blown ambulance car service/repair station equipped with highly trained/experienced car service personnels that can fix almost everything at the ambulance car ?

Your problem is, you've been whining for maybe years here ? That the hospital you're working at, and the country you're living at are not encouraging and supporting inhouse ambulance car repair, and in your case by an enthusiast grade but passionate to fix the ambulance at the hospital it self.

Pretty sad actually, imo you should move to China, or at least change your career, that hopefully will made enough money to support your hobby & passion in power electronics, so you can live happily.

[doobedoobedo]

Well some of us don't think Treez has ever set foot out of whichever Eastern European country he's from.

Also who in their right mind would advocate for an industry to be run by government? Unless of course the motivation was to piss away tax income.

[ebastler]

Well some of us don't think Treez has ever set foot out of whichever Eastern European country he's from.

Why would you think that? No, treez is for real and is working for a lighting company in Southern England. I am not going to dox him here, but he has certainly posted enough detail to be Google-able.

[rstofer]

There’s also the related concept of relative advantage, of doing only the things where you are significantly better than others.

When Jack Welch ran GE (back when it was running a lot better than it has lately), he had a plan.  If a business segment wasn't #1 in its market or #2 with a chance of becoming #1, he sold it.  If you aren't #1, you can't set the price!

They have dumped ballast manufacturing, are selling off the lighting division for overseas manufacturing, trying to find a way out of railroads and so on.

GE even sold the dog!  The famous "His Master's Voice" dog was an image that started in the UK (Grammophone), was bought by RCA and sold later to GE.  Then GE sold the dog to Thomson SA, a French multinational corporation.  There are a number of licensees...

https://en.wikipedia.org/wiki/His_Master%27s_Voice

How could anybody do that?

I guess if you want to make the argument that companies should build their own PSUs, you could argue that they should make the other stuff as well.  Monitors, disk drives, DVD drives, etc.  All of this is possible, there are people who know how to do it.  But these accessories, like PSUs, are just commodities.  Not unlike a dozen eggs.  And no, I'm not going to raise my own chickens!  You ever smelled a chicken farm?  Even cattle don't smell that bad!

[coppercone2]

Speccing a power brick is : here are the dimensions and we need these voltages at these currents.
done.
i dont give a rats ass about the guts of the module , what switching controller they use or what topology. i give an electrical and mechanical spec and that 's it. how it works inside the box i don't care.
As a matter of fact : as technology improves and newer ic's with higher switching frequencies and better topologies come available it pays to have re-shop the brick ever couple of years. you may get cheaper and smaller for the same specs.

This is how its done but you can't seriously tell me as a designer you like it this way. I want to scrutinize every aspect of my product to make it solve the intended problem as good as possible. I even feel this about IC's, custom made asics to your specification will always be a better solution.

Honestly, I do like it this way. I'd rather ship than obsess over minute details of a commodity element of my design... If you need an aluminum enclosure, do you start by mining bauxite as well?

If I could justify it some how, then yes. I think in general though the repeatability and reliability and quality of aluminum is such that the benefits would be slim (but I don't know much about aluminum). Very interesting however.

I think your analogy is too steep, I don't think its right to compare a very pure chemical element and the chemical/mining industry to a electrical PCB.

[ocset]

The point is, that both in USA and Europe, there are huge Western owned Power supply designer/manufacture companys…so this shows that it can be profitable…unless of course they are maintained by protectionist grants(?).
There are even hugh lighting designer/manufacturers…eg Tridonic, Philips.
I think the end-game of the West not doing any Power supply design/manufacture is bad news for the West……..and if you want to have a sucesful power supply design/manufacture industry, then you need to have available jobs for  Western people in general power supply design/manufacture….that is, you need, if necessary, to stimulate that by having a government funded power supply design/manufacture industry….if it eventually gets succesful enough to go private, then so be it.
Power supplys are in virtually all electronics products, the West cannot risk loosing this skill by over-outsourcing power supply design/manufacture to the Far East……..i mean, for one thing, it would just be helping to fund the building of islands off the coast of the Bahama’s and Philipines etc

[CJay]

Our labor has priced itself out of the worldwide market.  I'm not suggesting we degrade to 3rd world living standards but we better focus on where we excel:  inventing magic!

The problem is, not everyone is capable of inventing magic.

This, always this.

The 'bubble gum' rock bottom price mass market stuff can be done anywhere, what's important is to remain on the leading edge producing stuff that can't be done elsewhere, we are now an IP economy, the Bunnie Huang video explaining cellphone manufacture makes it abundantly clear that the west can never compete with developing economies and China etc. on a level playing field in manufacturing, where we can compete is in the creation of new IP and then take the first year or so of profits from outsourcing it to manufacture in those develoing markets, that way we all win.

To suggest bringing back manufacturing of parts like the every day consumer PSU is madness and utterly counterproductive, it will hold us back rather than propel us forward, we should be concentrating on new technolgies, for the same reason bringing back jobs in coal, steel etc. is moronic when the only way to be cometitive is to slant the playing field against your competitiors with artificial barriers, that might work short term if you can put them out of business but mid to long term its idiotic as it creates inflation in your own market..

[sokoloff]

The point is, that both in USA and Europe, there are huge Western owned Power supply designer/manufacture companys…so this shows that it can be profitable…unless of course they are maintained by protectionist grants(?).
There are even hugh lighting designer/manufacturers…eg Tridonic, Philips.
I think the end-game of the West not doing any Power supply design/manufacture is bad news for the West……..and if you want to have a sucesful power supply design/manufacture industry, then you need to have available jobs for  Western people in general power supply design/manufacture….that is, you need, if necessary, to stimulate that by having a government funded power supply design/manufacture industry….if it eventually gets succesful enough to go private, then so be it.
Power supplys are in virtually all electronics products, the West cannot risk loosing this skill by over-outsourcing power supply design/manufacture to the Far East……..i mean, for one thing, it would just be helping to fund the building of islands off the coast of the Bahama’s and Philipines etc

Does the west have a successful power supply industry? (Is your green text true and blue text not? Or green text is true, but you think the owners are going to walk away from a successful and profitable business for some reason?)

If your green text is true, why would we need the government intervention called for in the red text?

I'm honestly trying to follow your arguments in this thread (and the others), but they're coming across as "I have a strong opinion and I'm going to shout about it for a while" rather than "Here's the fact-based evidence that I see and that evidence leads me to the following conclusion that I'm concerned about." IMO, the latter is a much more effective way to engage with and convince others.

[rstofer]

I'm not in favor of some form of protectionist racket to support overpriced domestic manufacturing.  In terms of power supplies, who cares?

There are two classes of power supplies in my view:  Those required for national defense and the consumer stuff.  I don't care where the consumer stuff comes from and we do build overpriced national defense supplies.  But for the defense application, design requirements will obviously drive costs upward.

And I don't believe for one minute that we can't manufacture a consumer power supply in the US.  Who invented all those SMPS chips?  We have the capability, we just can't reach a price point.  Government subsidies only work in the EU; we aren't doing that kind of thing in the US very much.  Yes, there are examples of subsidized products, like tax breaks for battery cars, but those exist to create a market that will ultimately create a product which will ultimately improve the environment.

It's a tough world out there!  Sink or swim...

[rstofer]

I think your analogy is too steep, I don't think its right to compare a very pure chemical element and the chemical/mining industry to a electrical PCB.

But we obsess over sand every day.  Somebody has to go to the beach, dig up sand, melt it into slugs, slice it into wafers, print an image (yes, I know the process is more complex than that) and wind up with a PC.

[rstofer]

I'm not in favor of some form of protectionist racket to support overpriced domestic manufacturing.  In terms of power supplies, who cares?

There are two classes of power supplies in my view:  Those required for national defense and the consumer stuff.  I don't care where the consumer stuff comes from and we do build overpriced national defense supplies.  But for the defense application, design requirements will obviously drive costs upward.

And I don't believe for one minute that we can't manufacture a consumer power supply in the US.  Who invented all those SMPS chips?  We have the capability, we just can't reach a price point.  Government subsidies only work in the EU; we aren't doing that kind of thing in the US very much.  Yes, there are examples of subsidized products, like tax breaks for battery cars, but those exist (temporarily) to create a market that will ultimately create a product which will ultimately improve the environment.

It's a tough world out there!  Sink or swim...

[ocset]

There are huge companies in the west (eg USA and Europe), who design/manufacture lots of cheap consumer lighting and power supply stuff...eg Tridonic, Philips, Osram, parts of GE, Ecco, Traco, RECOM, etc etc...so there must be a case for doing this in the west.

Also, think what would happen if the West imposed a total ban on Chinese imports tomorrow, due to the "China Island building" issue.
The Chinese would stop selling Cheap domestic power supplies to the West, and the West wouldnt be able to cope, because we wouldnt have enough
trained engineers to  make all the power supplies and lighting products etc , that are needed.

Also, if you dont have a big , general, cheap domestic power supply design/manufacture operation going on in the west,
then you will not be able to get enough engineers for the military stuff.
Also, if you only have the "magic electronics" stuff going on...(hi grade electronics jobs)...then you won't get enough youngsters coming into
study electronics, because they will worry that they wont "cut it" to be a "magic" electronics engineer......so they wont choose electronics.
Thats why you need a big , domestic electronics industry, because then the young will see more simple electronics  jobs that they can go for...and they
will choose  to study electronics.
You need a big cohort of youngsters to come into electronics...then from that cohort you will get your 5% of "magic" electronics engineers....
..but if your cohort is too small in the first place, then your 5% is going to be too small (ie you wont get enough "magic" level electronics engineers.

By the way, when i worked at Alstom, on the electric warships, we were told that we may struggle to get
high power thyristors , because in USA, they were struggling to train up enough semiconductor specialists.
My take is because they arent getting enough people into electronics in the first place....because we in the west have sacked our
large basic domestic electronics industry.

It wouldnt be pure protectionism.....if the west started this kind of industry, it wouldnt be that much lower than breaking even....its not like it wouldnt be able to sell the stuff.
-And the spinoffs, as ive said, would be great.

Of course, in the ideal situation, the west would only do the high end electronics stuff...but you have got to have loads of low end electronics
jobs in your economy if you want to produce  a sufficient  quantity of  those "magic" level electronics engineers.

[rstofer]

You're claiming that we need more engineers for an industry that you claim doesn't exist.  If we had an excess of engineers, salaries would go down and I'm not paying a huge amount of money for my grandson to get a first rate engineering education to have him work for burger-flipping wages.  I expect him to start at over $100k and work up quickly!

We have plenty of engineers - what we might not have is someone trained in a specialty.  So, they have to learn on the job.  That's true of all engineering.  What we turn out of colleges is a bunch of young people who are pretty good in math, know a little about a lot of things but know a lot about nothing.  And that's the way it should be.

In those engineering fields where registration is a requirement, there will also be a requirement for a 4 year apprenticeship under a registered engineer.  And this only starts AFTER the student passes the Engineer In Training exam and that isn't particularly easy since it covers ALL fields of engineering.  Electronic Engineers are at a serious disadvantage because a) electronics isn't tested and b) they didn't take enough Civil and Mechanical Engineering courses.  There's a reason I know this!  The test was REALLY difficult and I had to study a lot of subjects that never came up in an electronics program.

Then there are programs at universities for post grad work sponsored, or at least highly encouraged, by industry.  Hughes Aircraft wanted a bunch of MSEEs.  Turns out the local university wanted a bunch of grad students.  The industry, as a whole, wanted advanced education so my employer (not Hughes) paid ALL of my costs for grad school - no conditions imposed.

Industry must plan to provide advanced education.  They may not want to do that but, overall, when they increase funding for training/education, they raise the level for the industry as a whole.  Silicon Valley works a lot like that.  People change jobs, people learn new stuff and change jobs again.

At one point, National Semiconductor teamed up with Stanford University to help create a "non-registered option" program where working engineers could take the same televised engineering classes at the plant that students were taking on campus.  Same homework, same tests, the only thing missing was the microphones for interrupting the instructor and those weren't used on campus either.  This was way back in the mid '70s so it's nothing new!  Oh, it was in real time so the classes were during working hours!  And you got paid for attending!

We have plenty of talent.  If you drill down on this page to "Electronic Engineers - except computer" and then look up California, you will see a mean income of $122K and if you scroll down further for Sunnyvale-San Jose-Santa Clara (Silicon Valley) you will see $130K mean income.  Now we're talking!  Alas, that won't buy a house in the local area...  But that's a separate issue.

We have plenty of engineers, we have plenty of programmers, AFAIK, we have plenty of everything.  But you don't use $130K engineers to develop consumer grade power supplies.  They have better things to do.

I haven't heard the military complaining about not being able to find engineers.  They don't hire them directly (for the most part) but rather expect them to work for Raytheon, Grumman, General Dynamics and other contractors.  AFAIK, these companies can find plenty of engineers.  They just aren't cheap!  And, usually, they need to be US citizens (to get the security clearances) so we don't have to worry about H1B imports taking the jobs.