How is Chipageddon affecting you?

[chickenHeadKnob]

--snip--
One guitar pedal is short a half dozen 2N5457 JFETs, which everyone is out of, showing obsolete in many places now.

JFET's  in general are something of a special case as classic part numbers have been obsoleted and hard to source for a long time now. Not related to the current "chipa-geddon"

[Red Squirrel]

Wonder if Remington Rand will make a come back with all this.  Pretty soon we'll have to go back to vacuum tube computers. 

Either that, or relay computers.  I could see a miniature relay tech, like DLP but with relays maybe become a thing.  Though there are still applications where you need to operate in the linear region of a transistor.  Guess that's where tubes would come in.  Wonder how small those can be made or if size matters a lot as to how they operate.  If they can get them down to the size of say, a flashlight bulb, I can totally see companies using them in their designs if it really comes to a point where even discrete fets or transistors can't be bought.

[bdunham7]

Wonder if Remington Rand will make a come back with all this.  Pretty soon we'll have to go back to vacuum tube computers. 

No, that's not how post-apocalyptic dystopian societies work.  There won't be any new computers, just what you can harvest from dumpsters or steal from your fellow survivors.  The phrase "learn to appreciate what you have" will take on a whole new meaning.

[perdrix]

Avago HEDS-9100-F00 encoders not available anywhere - 26 weeks lead time 

David

[T3sl4co1l]

MEMS switches are very sticky (van der Waals forces), I understand they have problems with consistent operation.  But yes, it has been tried!  Excellent on-off ratio (including at AC, i.e. low capacitance), as you would expect from a mechanical switch.

I wonder how fast they can operate.  kHz certainly, but maybe not MHz.  Much larger scale size as well, um per transistor rather than 10s nm.

Or speaking of which, can they do nano-MEMS these days?  I haven't checked.  With everything just completely bathed in proximity forces (and with such small levers compared to the contact areas, even direct chemical bonding will be considerable) I would guess it's not likely any better behaved.

Switches can possibly be controlled in an analog fashion, van der Waals forces and such notwithstanding, as tunneling current appears before direct conduction does.

Tim

[VK3DRB]

Many MEMs chips are not available. And all IMU sensors. Robert Botch has a 38 week lead time on the BNO055.

[Siwastaja]

Robert Botch has a 38 week lead time on the BNO055.

Not really, it's the "normal" lead time still quoted, but they can't deliver. If you order now, you won't likely get the parts in 38 weeks. Bosch Sensortec was in the situation of completely ignoring distributors and not answering questions for some 4 months, which resulted in Mouser to put a notification ALL IN CAPS that the factory is unresponsive, warning everybody about this dodgy company which shuts down communication completely in a crisis.

But finally a few weeks ago, BOSCH finally opened their mouth and came up with an estimate when they can satisfy orders in queue since January 2021 - and that will be in June 2022. So while possible, I seriously doubt their ability to provide 38 week lead time for new orders, because their current estimate of the actual lead time is 1.5 years.

[jmelson]

Guess that's where tubes would come in.  Wonder how small those can be made or if size matters a lot as to how they operate.

Roughly during the Vietnam war era, there were CK57xx tubes used in some gear.  They were roughly the diameter of a pencil and about an inch long.  Search for "CK57 vacuum tube" and you will find some pics online.
The US NSA made some crypto gear that used many dozens of them.  Check out the KWR-37 at :
http://www.jproc.ca/crypto/kwr37.html

Jon

[Gribo]

Thanks for the heads up regarding Bosch. I have a new design with one of their IMUs. Lets hope TDK does not behave the same way.

[VK3DRB]

Robert Botch has a 38 week lead time on the BNO055.

Not really, it's the "normal" lead time still quoted, but they can't deliver. If you order now, you won't likely get the parts in 38 weeks. Bosch Sensortec was in the situation of completely ignoring distributors and not answering questions for some 4 months, which resulted in Mouser to put a notification ALL IN CAPS that the factory is unresponsive, warning everybody about this dodgy company which shuts down communication completely in a crisis.

But finally a few weeks ago, BOSCH finally opened their mouth and came up with an estimate when they can satisfy orders in queue since January 2021 - and that will be in June 2022. So while possible, I seriously doubt their ability to provide 38 week lead time for new orders, because their current estimate of the actual lead time is 1.5 years.

June 2022... what gets me about these dummköpfes is they still advertise their Sensortec brand, fully knowing they cannot supply parts. They should have had reserve stock for the engineering community to continue the R & D design cycle, else when these chips finally become available, engineers will have abandoned them for alternatives. My guess is few companies will want to use Sensortec devices in mid 2022 and Bosch will have a difficult time trying to regain any trust and confidence in them.

I will investigate alternatives of the BNO055 from another vendor, even if it takes more than one chip. Unfortunately it means a heap of research and coding to do; as well as a PCB respin.

Incidentally, in 2019, Bosch was fined 90 million euros for their part in the VW diesel emissions fraud. Dodgy company.

[Siwastaja]

Yes, all vendors are having troubles delivering now but Bosch Sensortec seems very arrogant and ignore normal communication practices. They for example don't relay the part status information, as usual, to the suppliers so parts are listed "active" on all suppliers while the Bosch Sensortec website (as the only place in the world) says parts are NRND.

But in the end, if parts are good, it's what matters the most. They are not the only ones who have 1-1.5 year actual lead times now.

[Ranayna]

As someone looking at it from the outside, the issue at Bosch surprises me.
Just two weeks ago there was big news that they opened a new Fab in Dresden. Specifically producing chips for automotive applications. 

So apparently, contrary to the news reports, the Fab is not running yet, or can they really have that big of a backlog?

Other than that: We notice severe delivery time increases for virtually everything: PCs, Laptops, screens, network infrastructure, printers, you name it. Along with sometimes massive price updates.

[VK3DRB]

Good customer-focused companies will be honest and keep stakeholders informed. Bad companies like Bosch leave customers not knowing what to do. Silence can be a form of dishonesty. I feel for great companies like Digikey and Mouser who are caught up with the likes of Bosch.

There should be an online rating system specifically for electronic components vendors, so that engineers can make an informed decision about whether to use a company based on vendor reputation. There are review sites around, but it is all fragmented and sparse. Such a website would sort out the wheat from the chaff.

[jrs45]

FWIW Mouser seems to be shipping same-day again, or next day, which is nice.

[cortex_m0]

As someone looking at it from the outside, the issue at Bosch surprises me.
Just two weeks ago there was big news that they opened a new Fab in Dresden. Specifically producing chips for automotive applications. 

So apparently, contrary to the news reports, the Fab is not running yet, or can they really have that big of a backlog?

The fab is not running yet. https://www.design-reuse.com/news/50133/bosch-wafer-fab-dresden.html

Production in Dresden will start as early as July – six months earlier than planned. From that time on, semiconductors made in the new plant will be installed in Bosch power tools. For automotive customers, chip production will start in September, and thus three months earlier than planned.

[tszaboo]

As someone looking at it from the outside, the issue at Bosch surprises me.
Just two weeks ago there was big news that they opened a new Fab in Dresden. Specifically producing chips for automotive applications. 

So apparently, contrary to the news reports, the Fab is not running yet, or can they really have that big of a backlog?

Other than that: We notice severe delivery time increases for virtually everything: PCs, Laptops, screens, network infrastructure, printers, you name it. Along with sometimes massive price updates.

It takes like 10-14 weeks to send a wafer though all the production steps on a 14nm production line. IDK how that compares to MEMS because that might be more complicated in some ways, easier in others.

Also, if you set up a new production line, it could be a long time before your yield is up to standards. Or if you received all the qualifications that your product requires. If all the testing is done on your newly produced ICs.

Which is no excuse for bad communication.

June 2022... what gets me about these dummköpfes is they still advertise their Sensortec brand, fully knowing they cannot supply parts. They should have had reserve stock for the engineering community to continue the R & D design cycle, else when these chips finally become available, engineers will have abandoned them for alternatives. My guess is few companies will want to use Sensortec devices in mid 2022 and Bosch will have a difficult time trying to regain any trust and confidence in them.

I will investigate alternatives of the BNO055 from another vendor, even if it takes more than one chip. Unfortunately it means a heap of research and coding to do; as well as a PCB respin.

Incidentally, in 2019, Bosch was fined 90 million euros for their part in the VW diesel emissions fraud. Dodgy company.

The Bosch Sensortech, the power tool company, the automotive company, the appliance company shares very little. They have different management, locations, goals. The only common denominator is the Bosch conglomerate.

[Gribo]

TI CC1101, CC1190, CC112x and CC1200 are all out of stock.

[TimFox]

Good customer-focused companies will be honest and keep stakeholders informed. Bad companies like Bosch leave customers not knowing what to do. Silence can be a form of dishonesty. I feel for great companies like Digikey and Mouser who are caught up with the likes of Bosch.

There should be an online rating system specifically for electronic components vendors, so that engineers can make an informed decision about whether to use a company based on vendor reputation. There are review sites around, but it is all fragmented and sparse. Such a website would sort out the wheat from the chaff.

The company from which I retired had a monthly meeting to review the performance of all vendors in its supply chain, and what actions were required for the bad performers.  (This was long before the plague hit.)

[harerod]

In several of my designs every single silicon based component has become unavailable. The other day a manufacturer confessed a "stock discrepancy", which translates into "several different components allocated for scheduled manufacturing have disappeared in the last couple of weeks". This is the second time this happened, albeit to a different client. I have been recommending to withdraw material from manufacturers for a while, which of course leads to new problems.
So that client actually planned ahead and now lost a year's worth of components, including regulators, transistors, microcontrollers. We are talking about products where every single component change requires a recertification. Brokers are not the kind of source we want to use here.
While the ensuing mud wrestling is kind of entertaining to watch, we have to find alternatives, otherwise people will lose their jobs.

Ah, incidentally - I opened this topic in the microcontrollers section.
https://www.eevblog.com/forum/microcontrollers/replacement-for-stm32f407/msg3597922/#msg3597922

[tom66]

Many MEMs chips are not available. And all IMU sensors. Robert Botch has a 38 week lead time on the BNO055.

We have a similar problem.  One distributor quoted 99 weeks lead time.  And the company, like others have said, doesn't give a shit.  Unless you order 100k/year or more, they won't even talk to you about technical support, and neither will their technical support agents, who nominally get a kickback from any sales.

Part got designed out;  not dealing with Bosch again.  Shame, it's a neat part (though full of software bugs!)

[TimFox]

Guess that's where tubes would come in.  Wonder how small those can be made or if size matters a lot as to how they operate.

Roughly during the Vietnam war era, there were CK57xx tubes used in some gear.  They were roughly the diameter of a pencil and about an inch long.  Search for "CK57 vacuum tube" and you will find some pics online.
The US NSA made some crypto gear that used many dozens of them.  Check out the KWR-37 at :
http://www.jproc.ca/crypto/kwr37.html
M
Jon

Perhaps the final US production of small vacuum tubes were the “Nuvistors”, originated by RCA.  I believe the final runs were done at Phillips ECG (originally Sylvania) since they were common in military gear.  A 7586 triode, rated at 1 W plate dissipation and  110 V plate voltage, has an overall length of 0.80 in and a maximum diameter of 0.44 in.  It requires roughly 900 mW to heat the cathode.

[IDEngineer]

Many MEMs chips are not available. And all IMU sensors. Robert Botch has a 38 week lead time on the BNO055.

Now there's an ironic typo....

[T3sl4co1l]

Guess that's where tubes would come in.  Wonder how small those can be made or if size matters a lot as to how they operate.

Roughly during the Vietnam war era, there were CK57xx tubes used in some gear.  They were roughly the diameter of a pencil and about an inch long.  Search for "CK57 vacuum tube" and you will find some pics online.
The US NSA made some crypto gear that used many dozens of them.  Check out the KWR-37 at :
http://www.jproc.ca/crypto/kwr37.html
M
Jon

Perhaps the final US production of small vacuum tubes were the “Nuvistors”, originated by RCA.  I believe the final runs were done at Phillips ECG (originally Sylvania) since they were common in military gear.  A 7586 triode, rated at 1 W plate dissipation and  110 V plate voltage, has an overall length of 0.80 in and a maximum diameter of 0.44 in.  It requires roughly 900 mW to heat the cathode.

I have some Raytheon subminis dated into the mid 80s, were those not US?

Tim

[TimFox]

I have some Nuvistors dated in the mid-80s.  Sometime in that decade, the major tube manufacturers gave final notice to the DoD, who responded with final orders to maintain their stock of spares for replacement.  Early in the next century, the DoD sold off much of that stock and crazy people like me stocked up.
I assume your Raytheon sub-minis are US made.

[T3sl4co1l]

Ah yes, that makes sense.

I recall there were a handful of [production] TV sets still using tubes [other than the CRT] til the... late 70s at least, not sure if early 80s.  That would coincide with an LTB around there.

Tim