Vishay Mosfet group going under?
in [Design]
|
From: Joerg on 6 Jun 2010 19:18 krw(a)att.bizzzzzzzzzzzz wrote: > On Sun, 06 Jun 2010 14:24:59 -0700, Joerg <invalid(a)invalid.invalid> wrote: > >> krw(a)att.bizzzzzzzzzzzz wrote: >> >>> On Sun, 06 Jun 2010 13:58:39 -0700, Joerg <invalid(a)invalid.invalid> wrote: >> [...] >> >>>> I mean, if one of them would go overboard too many times the big boss >>>> would see that on the regular reports and could have a chat. In my case >>>> that was never needed, in years. But I do remember some consternation >>>> when the accounting guys of the company that bought us heard about this. >>> Overboard, yes, but this stuff is *CHEAP*. What's a few hundred bucks? Ok, >>> when I was using $3800 FPGAs, sure, be careful with them. $.002 resistors? >>> and $.03 gates? >> >> I did give them some pointers about my expectations, such as "Now don't >> go out there and order five new MS-Office licenses just because >> Microsoft came out with a new version" :-) >> >> Seriously, this can happen. I had my surprise recently when arriving at >> a client to check out my prototype. They wanted to time a mechanical >> reaction initiated by my circuit and I had told them that we could do >> stuff like that on the cheap. A $1 piezo would have sufficed. Long story >> short when I arrived there was a professional sensor in hi-sheen >> stainless steel, cradled in a velvet-clad and immaculately polished >> wooden box ... > > Perhaps they didn't understand the requirements and thought they needed > something traceable to NIST. > Well, the timing would be kind of hard to trace to NIST unless the connected electronics are traceable. Which I guess was technically the case because it was still within the calibration period. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Joerg on 6 Jun 2010 19:29 krw(a)att.bizzzzzzzzzzzz wrote: > On Sun, 06 Jun 2010 14:29:08 -0700, Joerg <invalid(a)invalid.invalid> wrote: > >> krw(a)att.bizzzzzzzzzzzz wrote: >>> On Sun, 06 Jun 2010 14:02:47 -0700, Joerg <invalid(a)invalid.invalid> wrote: >>> >>>> krw(a)att.bizzzzzzzzzzzz wrote: >> [...] >> >>>>> Several manufacturers are apparently dumping these lines (CS10?). All but a >>>>> few crystals (14.7456MHz, and such) are custom ordered. This one happens to >>>>> be 12.0MHz. In reality it can be anything from 4-40MHz. We also use a >>>>> 12.5MHz crystal in the product. I could easily use it, but with 12.7K in >>>>> inventory there isn't any point. >>>>> >>>>>> When I did that with an inductor for a client a long time ago we looked >>>>>> at it and it was "Wow, custom from Taiwan costs less than catalog from >>>>>> the US". So they are still using that Taiwanese inductor. >>>>> Yes, these crystals are from China, too. >>>> If a certain type seems to get dropped you might want to ask around. In >>>> China there always seems to be someone who keeps making older stuff. For >>>> example, through-hole parts and single-sided phenolic are kicking and >>>> alive over there. When you ask a board house about phenolic in the US >>>> that can result in a blank stare ;-) >>> Yeah, we have no trouble finding someone else. It's just some work showing >>> that the new parts are really equivalent. It also takes 12 weeks (or more) to >>> get the evaluation parts before useful quantities can be ordered (then another >>> 12 weeks). >> >> For a crystal? It's been a while but I could always get them in two >> weeks if needed. In a real crunch within the same week. > > At one time, sure. No longer. > >> But yeah, China >> tends to take longer and often the willingness to expedite drops down >> when the supplier grows and takes on more important (a.k.a. bigger) >> customers. Been there :-( > > Apparently these parts are only sourced in China, now. Even crystal > oscillator lead times have slipped out that far and the disty can, under > normal circumstances, program them. Hmm, putting my capitalist hat on (which I usually only take off at night ...) all this opens numerous opportunities: a. A "good-old-days" type of service for crystals. b. A futures trading system for components. With hedge contracts and the whole nine yards. c. Parts obsolescence insurance plans with a leadtime rider, with an (expensive) option to also insure against excessive leadtimes. d. Hunter businesses, where there are numerous very smart engineers who no longer participate in the productive life of a circuit designer but canvas the market all day long. In search of parts that shoot up in leadtime but where they find out that these are crucial to some products. Then buy the whole stock in one fell swoop, and re-sell it to companies that failed to move away from JIT. Ka-ching! -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Robert Baer on 6 Jun 2010 19:40 Joerg wrote: > Robert Baer wrote: >> Joerg wrote: >>> Yzordderrex wrote: >>>> On Jun 5, 10:06 am, Winfield Hill <Winfield_mem...(a)newsguy.com> >>>> wrote: >>>>> Yzordderrex wrote... >>>>> >>>>>> I heard that Vishay is cancelling open orders for fets and shutting >>>>>> lines down. I suspect my intelligence sources though. Has anyone >>>>>> heard anything that might confirm this? >>>>> Can you get more specific information. With part numbers? >>>>> For example, Vishay bought a raft of old MOSFET products >>>>> from IRF a few years ago, parts that IRF was likely ready >>>>> to retire then, and Vishay may finally be retiring now. >>>>> >>>>> -- >>>>> Thanks, >>>>> - Win >>>> >>>> Sorry Win, I don't have much to go on. I was told that they are >>>> canceling open orders. I think this is probably an embellishment of >>>> the truth. I do notice quite a few fets are out of stock at digikey, >>>> so I expect there is a lot of panic buying going on. Not just >>>> Vishay, but other sources as well. A different intel source told me >>>> that they are trying to slow down this environment where purchasing >>>> groups are helter skelter ordering 2x what they need in order to get >>>> what they need. ... >>> >>> >>> That actually _is_ what I recommend to client for some parts. Better >>> to have $3k of extra inventory sitting on the shelf than being hit >>> with a hard 20wk leadtime and next month you have to furlough half >>> the production staff because of a hardcore line stop. I know that >>> accountants and CFOs really don't like this but man's gotta do what >>> man's gotta do. >>> >>> The only explanation I could imagine is that some mfgs must have >>> slammed the brakes end of 2008. >>> >>> [...] >>> >> 20wk lead time sounds like they have to open new mines for raw >> materials, then build plants, then... > > > Well, I see that sort of time frame a lot lately. Essentially it means > "No, we don't have anything we could test, package and bond. Nada, zip, > zilch". So one has to react by hoarding stuff. > Is it not a bit difficult to get what does not exist due to the long timeframe?
From: krw on 6 Jun 2010 19:49 On Sun, 06 Jun 2010 16:29:49 -0700, Joerg <invalid(a)invalid.invalid> wrote: >krw(a)att.bizzzzzzzzzzzz wrote: >> On Sun, 06 Jun 2010 14:29:08 -0700, Joerg <invalid(a)invalid.invalid> wrote: >> >>> krw(a)att.bizzzzzzzzzzzz wrote: >>>> On Sun, 06 Jun 2010 14:02:47 -0700, Joerg <invalid(a)invalid.invalid> wrote: >>>> >>>>> krw(a)att.bizzzzzzzzzzzz wrote: >>> [...] >>> >>>>>> Several manufacturers are apparently dumping these lines (CS10?). All but a >>>>>> few crystals (14.7456MHz, and such) are custom ordered. This one happens to >>>>>> be 12.0MHz. In reality it can be anything from 4-40MHz. We also use a >>>>>> 12.5MHz crystal in the product. I could easily use it, but with 12.7K in >>>>>> inventory there isn't any point. >>>>>> >>>>>>> When I did that with an inductor for a client a long time ago we looked >>>>>>> at it and it was "Wow, custom from Taiwan costs less than catalog from >>>>>>> the US". So they are still using that Taiwanese inductor. >>>>>> Yes, these crystals are from China, too. >>>>> If a certain type seems to get dropped you might want to ask around. In >>>>> China there always seems to be someone who keeps making older stuff. For >>>>> example, through-hole parts and single-sided phenolic are kicking and >>>>> alive over there. When you ask a board house about phenolic in the US >>>>> that can result in a blank stare ;-) >>>> Yeah, we have no trouble finding someone else. It's just some work showing >>>> that the new parts are really equivalent. It also takes 12 weeks (or more) to >>>> get the evaluation parts before useful quantities can be ordered (then another >>>> 12 weeks). >>> >>> For a crystal? It's been a while but I could always get them in two >>> weeks if needed. In a real crunch within the same week. >> >> At one time, sure. No longer. >> >>> But yeah, China >>> tends to take longer and often the willingness to expedite drops down >>> when the supplier grows and takes on more important (a.k.a. bigger) >>> customers. Been there :-( >> >> Apparently these parts are only sourced in China, now. Even crystal >> oscillator lead times have slipped out that far and the disty can, under >> normal circumstances, program them. > > >Hmm, putting my capitalist hat on (which I usually only take off at >night ...) all this opens numerous opportunities: > >a. A "good-old-days" type of service for crystals. The problem here is that there isn't enough money in crystals anymore to make them worthwhile for anyone but the Chinese. >b. A futures trading system for components. With hedge contracts and the >whole nine yards. Isn't that what disties do? They have the added advantage of being fronted by some manufacturers. >c. Parts obsolescence insurance plans with a leadtime rider, with an >(expensive) option to also insure against excessive leadtimes. Quick, call AIG. >d. Hunter businesses, where there are numerous very smart engineers who >no longer participate in the productive life of a circuit designer but >canvas the market all day long. In search of parts that shoot up in >leadtime but where they find out that these are crucial to some >products. Then buy the whole stock in one fell swoop, and re-sell it to >companies that failed to move away from JIT. e. Index Arbitrage. Buy in the US and sell in Europe. Buy in China and sell in the US.
From: Joerg on 6 Jun 2010 20:19
Robert Baer wrote: > Joerg wrote: >> Robert Baer wrote: >>> Joerg wrote: >>>> Yzordderrex wrote: >>>>> On Jun 5, 10:06 am, Winfield Hill <Winfield_mem...(a)newsguy.com> >>>>> wrote: >>>>>> Yzordderrex wrote... >>>>>> >>>>>>> I heard that Vishay is cancelling open orders for fets and shutting >>>>>>> lines down. I suspect my intelligence sources though. Has anyone >>>>>>> heard anything that might confirm this? >>>>>> Can you get more specific information. With part numbers? >>>>>> For example, Vishay bought a raft of old MOSFET products >>>>>> from IRF a few years ago, parts that IRF was likely ready >>>>>> to retire then, and Vishay may finally be retiring now. >>>>>> >>>>>> -- >>>>>> Thanks, >>>>>> - Win >>>>> >>>>> Sorry Win, I don't have much to go on. I was told that they are >>>>> canceling open orders. I think this is probably an embellishment of >>>>> the truth. I do notice quite a few fets are out of stock at digikey, >>>>> so I expect there is a lot of panic buying going on. Not just >>>>> Vishay, but other sources as well. A different intel source told me >>>>> that they are trying to slow down this environment where purchasing >>>>> groups are helter skelter ordering 2x what they need in order to get >>>>> what they need. ... >>>> >>>> >>>> That actually _is_ what I recommend to client for some parts. Better >>>> to have $3k of extra inventory sitting on the shelf than being hit >>>> with a hard 20wk leadtime and next month you have to furlough half >>>> the production staff because of a hardcore line stop. I know that >>>> accountants and CFOs really don't like this but man's gotta do what >>>> man's gotta do. >>>> >>>> The only explanation I could imagine is that some mfgs must have >>>> slammed the brakes end of 2008. >>>> >>>> [...] >>>> >>> 20wk lead time sounds like they have to open new mines for raw >>> materials, then build plants, then... >> >> >> Well, I see that sort of time frame a lot lately. Essentially it means >> "No, we don't have anything we could test, package and bond. Nada, >> zip, zilch". So one has to react by hoarding stuff. >> > Is it not a bit difficult to get what does not exist due to the long > timeframe? That's why it is customary to run extra wafers and put those on the shelf in case of increased future demand. That seems not to be the case anymore with some products. And yeah, when you have to schedule a whole new wafer run you'll see those long leadtimes, plus it only makes sense if the total orders have reached a substantial level. Here we have the chicken-and-egg phenomenon: I as a design engineer will generally move on when I see a 22wk leadtime, trying to find a solution with parts that are more available. Or I go it discrete instead of using a chip. That means my clients who are unlikely to touch successful designs again without a compelling reason will never even order such stuff. That fuels the downward spiral for such parts. Old American sales wisdom: If you don't take care of your customers, someone else will. I don't have a good explanation for this. Either manufacturers lost the art of planning to some extent, are hardcore cash-starved or really peed into their pants after the downturn that set in around 2008. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |