The New Chip Shortage is Passive Components

[TheLostSwede]

If you thought that the chip shortage was bad, then the building shortage of passive components, such as capacitors, resistors, inductors and so on, is going to have you in tears. Due to lockdowns in Malaysia and Indonesia, where most of the well known Japanese aluminium capacitors are made, the factories of Chemi-Con, Nichicon and Rubycon have been shut down for most of July and August. The three companies together control some 50 percent of the capacitor market and it's expected that the current situation in Malaysia will lead to a reduction in capacitor shipments by 30 to 60 percent.

At the same time, the increased demand for everything from computer parts to renewable energy technology has ramped up demand for these components. Some of that business has been picked up by Taiwanese and Chinese manufacturers, but whereas in the past you could get your order in some four to six weeks, the lead times are now three to six months and that's if you're lucky. DigiTimes is reporting that several Taiwanese component makers have seen a YoY revenue growth of 20 percent or more for the first half of this year.



This has in turn led to some of the Taiwanese capacitor manufacturers eyeing buyouts of their suppliers to be able to provide a more streamlined and secured manufacturing pipeline, where they don't have to worry about their competitors buying up stock from their suppliers. The shortage isn't expected to ease in 2021, although if things improve in Malaysia and Indonesia, 2022 might see better supply of these vital components, but right now it's anyone's guess what will happen in the longer term.

View at TechPowerUp Main Site

 

[Bomby569]

It shows the same problem as always, the world can't be dependent on 1 or 2 countries to manufacture vital components but i doubt any lessons will be learn from this. Race to the bottom, make shit as cheap as possible.

 

[Vayra86]

Lovely.

 

[Ferrum Master]

Yes...

My all personal projects are on HOLD.... Premium part orders are back in supply only next year...

Well... I guess drinking beer is only that's left.

Chatting with my Mikrotik friends, they suffer from it even longer, like a year and cannot even push out premium product lines... Private sector was the last to take a hit.

 

[persondb]

I wonder how hard it is to increase capacity for those components. Those have really existed for a long time though that doesn't mean they are easy to manufacture.

Well, the demand for electronics will keep increasing so it would make sense to have more capacity in manufacture of those anyway in my view.

 

[TheDeeGee]

And when is the next flooding because a someone stumbled over a bucket of water?

 

[Chomiq]

Don't miss:

Global chip shortage hits all-time high with a wait time of 21 weeks

The automotive industry is impacted the most. It's probably common knowledge by now that there is a severe chip shortage worldwide, which affects not only...

www.gsmarena.com

 

[TheLostSwede]

I wonder how hard it is to increase capacity for those components. Those have really existed for a long time though that doesn't mean they are easy to manufacture.

Well, the demand for electronics will keep increasing so it would make sense to have more capacity in manufacture of those anyway in my view.

It's not just the manufacturing capacity that's the issue, but also the testing, as that needs to have some "operator" input and can't all be automated.

Don't miss:

Global chip shortage hits all-time high with a wait time of 21 weeks

The automotive industry is impacted the most. It's probably common knowledge by now that there is a severe chip shortage worldwide, which affects not only...

www.gsmarena.com

Order lead times on finished chips is more like 52 weeks plus. Even simple microcontrollers have at least that long lead times, if your buying from well known companies like ST, NXP, Microchip etc.

 

[Bomby569]

it's also a supply chain problem, if you don't have chips you decrease production of everything else, then you have chips but other things aren't there, it's hard to manage production in this climate, a mess

 

[Lionheart]

Tell them to put better capacitors in the PS5's internal PSU, that thing has more coil whine/buzzing than my HD 5970 at launch.

 

[Ferrum Master]

Tell them to put better capacitors in the PS5's internal PSU, that thing has more coil whine/buzzing than my HD 5970 at launch.

Caps are not responsible for that in this particular case dude...

Mine doesn't, RMA it if you dare. QC issues.

I have to admit, I had a case with my secondary acer monitor, that whined... but at least they all do. I fixed it myself tho.

 

[zlobby]

I guess these companies simply... didn't have the capacity. This is FUD-inducing! Sorry, I couldn't resist! I should be grounded until I conduct properly!

 

[TheLostSwede]

Tell them to put better capacitors in the PS5's internal PSU, that thing has more coil whine/buzzing than my HD 5970 at launch.

Coil whine often comes from poor pairing of components in the power regulation and not from having better/worse capacitors.
Had this issue at a company I worked for and apparently the engineers were deaf, as none of them heard it. Luckily it was on a mezzanine board, so it was easy for them to change the design.

 

[Ferrum Master]

the engineers were deaf, as none of them heard it.

Lough or not, maybe they were... if they were older dudes it would be only natural that hearing decays at high spectrum, especially where the piezoelectric pitch is.

My Acer panel also had a dumb PCB design, it may look pretty, like streets and houses assembled in close groups... some stupid PCB designer with artist traits and no field knowledge Blame them. My panel driver board supply whined, of course it was the MLCC section besides vreg... it didn't whine on lower refresh rate 60Hz, because less current is consumed, setting 165Hz a concert starts. Caps are aligned too close with each, they resonate, the PCB is too thin, the PCB was glued with double sided adhesive to the IPS panel metal shroud it all made it act like a speaker. I made a effect from defect, I removed the adhesive and put Thermal pads beneath and above the PCB to dampen the vibrations, with doubled benefit of cooling. It mitigated the noise immediately and for my surprise it is a good monitor now, I use it in landscape mode always besides my main monitor, for web and datasheets.

But the moral, that on day one after buying a monitor, I had to fix it... isn't very pleasing, maybe my karma played here. But after reading some Reddit complaints also, I decided that RMA won't help and they all are like that - for deaf people lol. There are many people where they live it is all noisy, silence isn't found even at nights, let it be AC, work environment, like chat, PC Fans, kids, traffic, haunted house, thus they even do not notice. Maybe that was the real excuse for engineers in your case.

 

[TheLostSwede]

Lough or not, maybe they were... if they were older dudes it would be only natural that hearing decays at high spectrum, especially where the piezoelectric pitch is.

Nah, they were both younger than me...
Coil whine drives me nuts, but in all fairness, it only happened when you transmitted data over the 5GHz radio, which they apparently hadn't tested...
This is why it's important you test all parts of a product before you ok an a design.

 

[Ferrum Master]

Nah, they were both younger than me...
Coil whine drives me nuts, but in all fairness, it only happened when you transmitted data over the 5GHz radio, which they apparently hadn't tested...
This is why it's important you test all parts of a product before you ok an a design.

They tested on a retail device? Well prototypes have often better parts, housing, PCB(like especially copper layer thickness)etc... as I said also often one additional mylar or thermal pad can mitigate the issue. Or it was a total stupidity that it started to act like that on higher current loads and always as 5GHz leaks more and eats more current... if so... then yeah... monkeys...

I forgot, there is one more possibility for... tinnitus, they hear the noise all the time, so yolo it ain't different as usual

 

[Chomiq]

They tested on a retail device? Well prototypes have often better parts, housing, PCB(like especially copper layer thickness)etc... as I said also often one additional mylar or thermal pad can mitigate the issue. Or it was a total stupidity that it started to act like that on higher current loads and always as 5GHz leaks more and eats more current... if so... then yeah... monkeys...

I forgot, there is one more possibility for... tinnitus, they hear the noise all the time, so yolo it ain't different as usual

I guess it depends on a person but for me noise from tinnitus is on a different frequency, also it only comes up at night when it's quiet in the room. With fans at low speed I can sometimes hear a bit of coil whine from the GPU but only when it's really quiet. I can hear my eneloops charging from across the room, even something as simple as an USB charger can generate some sort of high frequency noise.

 

[Ferrum Master]

I guess it depends on a person but for me noise from tinnitus is on a different frequency, also it only comes up at night when it's quiet in the room. With fans at low speed I can sometimes hear a bit of coil whine from the GPU but only when it's really quiet. I can hear my eneloops charging from across the room, even something as simple as an USB charger can generate some sort of high frequency noise.

Yes, I have thrown away a loads of chargers, mostly Huawei were prone to that, I took one home and could hear it from my bedroom, driving me nuts. China devices, yea... First thing I do when ordering something from AliExpress is open it up and fix it too. GPU's these days all do whine at higher loads, and that's a poor design, we pay such money yet we are fooled and receive inferior solutions, because the costs and the problem hasn't gained much needed traction. As long one of the makers will get a PR about some sort of SILENT VRM design, others then would follow, but naahh... let them feed the same, they game on headphones anyways.

Actually while on topic, the higher PD revision(like the other news topic) could imply use of better controllers and parts to ensure efficiency... but still it will not save from bad design or some batch of poor quality parts. And in these times, when there is part unobtainium... people make things from anything. So brace ourselves.

 

[TheLostSwede]

They tested on a retail device? Well prototypes have often better parts, housing, PCB(like especially copper layer thickness)etc... as I said also often one additional mylar or thermal pad can mitigate the issue. Or it was a total stupidity that it started to act like that on higher current loads and always as 5GHz leaks more and eats more current... if so... then yeah... monkeys...

I forgot, there is one more possibility for... tinnitus, they hear the noise all the time, so yolo it ain't different as usual

This was during product development, so obviously we had to test all aspects of the product before it went into production. Luckily I caught it, without even looking for it, as I was doing some performance testing and the router started to squeal like a piglet... At least no customer got a product with the issue, which is what mattered.

Actually while on topic, the higher PD revision(like the other news topic) could imply use of better controllers and parts to ensure efficiency... but still it will not save from bad design or some batch of poor quality parts. And in these times, when there is part unobtainium... people make things from anything. So brace ourselves.

Hopefully this is true.

This is also why I never sourced the cheapest possible power adapter for my clients, as they simply are shite and for 50 cents or so, it's not worth the complaints.

 

[Ferrum Master]

This was during product development, so obviously we had to test all aspects of the product before it went into production. Luckily I caught it, without even looking for it, as I was doing some performance testing and the router started to squeal like a piglet... At least no customer got a product with the issue, which is what mattered.

Then a total dud. No excuses. The prototype testing guys should have noticed it... but as long it booted, and put radio tests, they even didn't bother to test out the most basic things as they are usually copy pasted over and over from older designs.

Yeah, last time I got that when I was designing an unorthodox high voltage step up supply for guitar effect pedal, years ago, when overdriving it with such unusual load, it went nuts, now there are some nice specialized IC's for... guess what... nixie clocks, and they work perfect and are noiseless regarding voltage noise also, I don't even understand why they are made. There are still some innovations in high voltage department even in these days, that's actually unusual, the best thing being the LR8N high voltage linear vreg allowing to get rid of 500g of 10H choke for each preamp tube section and making it totally silent and few other know how things regarding tube degradation and biasing. Oh well my, my free time side job high voltage ventures...

 

[TheLostSwede]

Then a total dud. No excuses. The prototype testing guys should have noticed it... but as long it booted, and put radio tests, they even didn't bother to test out the most basic things as they are usually copy pasted over and over from older designs.

Well, this was a very "cutting edge" design at the time, as it was the first router (to my knowledge) that used a two PCB design and it was also the first to use a heatpipe for cooling.
Even so, this was something like the third or fourth revision of the mezzanine board, as we ended up swapping several things around due to EMI issues.

This was a much earlier board revision. and pretty much everything changed from this revision until we shipped nearly a year later...

And now we're way off topic...

 

[chrcoluk]

Cardboard shortage next to be announced?

 

[Ferrum Master]

Cardboard shortage next to be announced?

Prices on wood materials have nearly doubled here... soo... obviously.

 

[dirtyferret]

Cardboard shortage next to be announced?

I know you are saying this in jest but it actually started a few months back (we buy pallets of cardboard boxes of various sizes)

Cardboard prices reach record high amid e-commerce demand

Box producers are running on lean inventories, as online buying booms and creates greater demand for packaging materials.

www.supplychaindive.com

 

[TheLostSwede]

Cardboard shortage next to be announced?

Nah, second round of toilet paper shortage...