Forums

parts researcher

Started by John Larkin September 10, 2021
On 9/12/2021 3:09 PM, Joe Gwinn wrote:
>>> So I think I agree with those who say that part selection can't be >>> delegated, or at least shouldn't be. It's tedious, but I hate to call it >>> boring, because if you took the 2010 catalog back to 1980, >>> nobody would complain that it was "boring." They would be too >>> busy naming their firstborn son after you. We should count our >>> blessings. >> >> Lots of big companies have components engineers. Big mistake? > > Kinda. Their often purpose is to tell people what components they > cannot use.
Or, question why part X has been specified instead of part Y which sure looks similar (or is from a different vendor). "Let's each pick our favorite parts without concern for what parts the organization is already using, has already qualified, has staff trained to troubleshoot, etc."
> There was a big push to standardize components several years ago, > basically by generating and enforcing an approved component list. > (Impendent of the approved supplier list.) At the time, one company > had about 20,000 engineers, and their component engineering group had > about 20 engineers total. Turns out that those component engineers > are not smarter than 1,000 ordinary EEs. And certainly not 1,000 > times faster.
But likely contributed a disproportionate amount more! "Gee, Bob, the XYZ123 that you specified in your design is on back order --- 36 weeks. Bill is using an ABC456 and has no problems getting them by the boatload. How much effort for you to redesign for the ABC456, instead?"
On Sun, 12 Sep 2021 18:09:42 -0400, Joe Gwinn <joegwinn@comcast.net>
wrote:

>On Sun, 12 Sep 2021 14:49:29 -0700, jlarkin@highlandsniptechnology.com >wrote: > >>On Sun, 12 Sep 2021 14:06:44 -0700 (PDT), "John Miles, KE5FX" >><jmiles@gmail.com> wrote: >> >>>On Sunday, September 12, 2021 at 1:10:47 PM UTC-7, jla...@highlandsniptechnology.com wrote: >>>> Parts research is boring. >>> >>>I think there's some truth to every reply in the thread, which is >>>darned rare around here. Yes, parts research is tedious, but it's >>>also what product development *is* these days. I spend as >>>much time looking for parts as I do using them. >> >>Then you could design more if someone did parts research for you. >> >>I need a particular 61-pin circular MS connector with PCB pins, that >>mates with another particular MS connector. Price might be $85 but >>might be $350. It won't enrich my life to hunt around for that one. >> >>> >>>Back in the 1980s-1990s, the DigiKey catalog was about the size of >>>an issue of Electronic Design, maybe 200 pages at most. Now, it's >>>almost 3000 pages long and as thick as a metro-area phone book. >>>The cover says it has 465,000 parts. >> >>Yes. And the search engine keeps getting worse. >> >>> >>>Oh, wait, no, that's the 2010 catalog, the last one I have. They stopped >>>printing them ten years ago in 2011. The web site currently shows >>>13,000,000 parts. You can't actually GET most of them, but that still >>>leaves millions of parts that you can specify. >> >>Not specify: hunt for. >> >>> >>>Meanwhile, Electronic Design is about half as thick and full of not >>>much in particular. I don't know if they even bother to print it anymore. >>> >>>So I think I agree with those who say that part selection can't be >>>delegated, or at least shouldn't be. It's tedious, but I hate to call it >>>boring, because if you took the 2010 catalog back to 1980, >>>nobody would complain that it was "boring." They would be too >>>busy naming their firstborn son after you. We should count our >>>blessings. >> >>Lots of big companies have components engineers. Big mistake? > >Kinda. Their often purpose is to tell people what components they >cannot use.
That's valuable information.
> >There was a big push to standardize components several years ago, >basically by generating and enforcing an approved component list. >(Impendent of the approved supplier list.) At the time, one company >had about 20,000 engineers, and their component engineering group had >about 20 engineers total. Turns out that those component engineers >are not smarter than 1,000 ordinary EEs. And certainly not 1,000 >times faster.
We have about 7000 different parts in stock. That's too many. We certainly don't want to add any new ones if it can be avoided. Sometimes it can't. Imagine if 20,000 engineers started using any parts they could find online. You might have 200K different parts in the stockroom. Or 2 million. -- Father Brown's figure remained quite dark and still; but in that instant he had lost his head. His head was always most valuable when he had lost it.
On Sun, 12 Sep 2021 15:34:53 -0700, jlarkin@highlandsniptechnology.com
wrote:

>On Sun, 12 Sep 2021 18:09:42 -0400, Joe Gwinn <joegwinn@comcast.net> >wrote: > >>On Sun, 12 Sep 2021 14:49:29 -0700, jlarkin@highlandsniptechnology.com >>wrote: >> >>>On Sun, 12 Sep 2021 14:06:44 -0700 (PDT), "John Miles, KE5FX" >>><jmiles@gmail.com> wrote: >>> >>>>On Sunday, September 12, 2021 at 1:10:47 PM UTC-7, jla...@highlandsniptechnology.com wrote: >>>>> Parts research is boring. >>>> >>>>I think there's some truth to every reply in the thread, which is >>>>darned rare around here. Yes, parts research is tedious, but it's >>>>also what product development *is* these days. I spend as >>>>much time looking for parts as I do using them. >>> >>>Then you could design more if someone did parts research for you. >>> >>>I need a particular 61-pin circular MS connector with PCB pins, that >>>mates with another particular MS connector. Price might be $85 but >>>might be $350. It won't enrich my life to hunt around for that one. >>> >>>> >>>>Back in the 1980s-1990s, the DigiKey catalog was about the size of >>>>an issue of Electronic Design, maybe 200 pages at most. Now, it's >>>>almost 3000 pages long and as thick as a metro-area phone book. >>>>The cover says it has 465,000 parts. >>> >>>Yes. And the search engine keeps getting worse. >>> >>>> >>>>Oh, wait, no, that's the 2010 catalog, the last one I have. They stopped >>>>printing them ten years ago in 2011. The web site currently shows >>>>13,000,000 parts. You can't actually GET most of them, but that still >>>>leaves millions of parts that you can specify. >>> >>>Not specify: hunt for. >>> >>>> >>>>Meanwhile, Electronic Design is about half as thick and full of not >>>>much in particular. I don't know if they even bother to print it anymore. >>>> >>>>So I think I agree with those who say that part selection can't be >>>>delegated, or at least shouldn't be. It's tedious, but I hate to call it >>>>boring, because if you took the 2010 catalog back to 1980, >>>>nobody would complain that it was "boring." They would be too >>>>busy naming their firstborn son after you. We should count our >>>>blessings. >>> >>>Lots of big companies have components engineers. Big mistake? >> >>Kinda. Their often purpose is to tell people what components they >>cannot use. > >That's valuable information. > >> >>There was a big push to standardize components several years ago, >>basically by generating and enforcing an approved component list. >>(Impendent of the approved supplier list.) At the time, one company >>had about 20,000 engineers, and their component engineering group had >>about 20 engineers total. Turns out that those component engineers >>are not smarter than 1,000 ordinary EEs. And certainly not 1,000 >>times faster. > >We have about 7000 different parts in stock. That's too many. We >certainly don't want to add any new ones if it can be avoided. >Sometimes it can't. > >Imagine if 20,000 engineers started using any parts they could find >online. You might have 200K different parts in the stockroom. Or 2 >million.
That was exactly the hope and rationale, but the chosen solution had a serious scaling problem. Joe Gwinn
On Sunday, September 12, 2021 at 3:35:02 PM UTC-7, jla...@highlandsniptechnology.com wrote:
> Imagine if 20,000 engineers started using any parts they could find > online. You might have 200K different parts in the stockroom. Or 2 > million.
Ideally you wouldn't even have a stockroom. You'd use the ones in Mansfield, Thief River Falls, and similar places. Why keep all that stuff on YOUR books? (Well, it worked for a while, at least...) -- john, KE5FX
On Sun, 12 Sep 2021 16:51:39 -0700 (PDT), "John Miles, KE5FX"
<jmiles@gmail.com> wrote:

>On Sunday, September 12, 2021 at 3:35:02 PM UTC-7, jla...@highlandsniptechnology.com wrote: >> Imagine if 20,000 engineers started using any parts they could find >> online. You might have 200K different parts in the stockroom. Or 2 >> million. > >Ideally you wouldn't even have a stockroom. You'd use the ones in Mansfield, >Thief River Falls, and similar places. Why keep all that stuff on YOUR books? > >(Well, it worked for a while, at least...) > >-- john, KE5FX
Most of our parts are on reels. We buy a reel and keep it in stock. When we do a production run, we mount reels on the p+p machine, run the lot of boards, count what's left on the reels, and put them back in stock. Ordering the right number of parts from Digikey for every board run would be an expensive nightmare. -- Father Brown's figure remained quite dark and still; but in that instant he had lost his head. His head was always most valuable when he had lost it.
jlarkin@highlandsniptechnology.com wrote:
> On Sun, 12 Sep 2021 16:51:39 -0700 (PDT), "John Miles, KE5FX" > <jmiles@gmail.com> wrote: > >> On Sunday, September 12, 2021 at 3:35:02 PM UTC-7, jla...@highlandsniptechnology.com wrote: >>> Imagine if 20,000 engineers started using any parts they could find >>> online. You might have 200K different parts in the stockroom. Or 2 >>> million. >> >> Ideally you wouldn't even have a stockroom. You'd use the ones in Mansfield, >> Thief River Falls, and similar places. Why keep all that stuff on YOUR books? >> >> (Well, it worked for a while, at least...) >> >> -- john, KE5FX > > Most of our parts are on reels. We buy a reel and keep it in stock. > When we do a production run, we mount reels on the p+p machine, run > the lot of boards, count what's left on the reels, and put them back > in stock. > > Ordering the right number of parts from Digikey for every board run > would be an expensive nightmare. > > >
Plus you not infrequently get some comedian in the warehouse sending you a bag full of short tape strips. (This happens even though we have a 'continuous tape only' note in our distributor accounts.) Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
On Sun, 12 Sep 2021 21:09:26 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>jlarkin@highlandsniptechnology.com wrote: >> On Sun, 12 Sep 2021 16:51:39 -0700 (PDT), "John Miles, KE5FX" >> <jmiles@gmail.com> wrote: >> >>> On Sunday, September 12, 2021 at 3:35:02 PM UTC-7, jla...@highlandsniptechnology.com wrote: >>>> Imagine if 20,000 engineers started using any parts they could find >>>> online. You might have 200K different parts in the stockroom. Or 2 >>>> million. >>> >>> Ideally you wouldn't even have a stockroom. You'd use the ones in Mansfield, >>> Thief River Falls, and similar places. Why keep all that stuff on YOUR books? >>> >>> (Well, it worked for a while, at least...) >>> >>> -- john, KE5FX >> >> Most of our parts are on reels. We buy a reel and keep it in stock. >> When we do a production run, we mount reels on the p+p machine, run >> the lot of boards, count what's left on the reels, and put them back >> in stock. >> >> Ordering the right number of parts from Digikey for every board run >> would be an expensive nightmare. >> >> >> >Plus you not infrequently get some comedian in the warehouse sending you >a bag full of short tape strips. (This happens even though we have a >'continuous tape only' note in our distributor accounts.)
One reason the engineers are (theoretically) not allowed in the stockroom is that it is claimed we mess up the leaders on tape+reel parts. Buying full reels avoids breaks and problems like that. Some parts cost $0.00 each rounded to the nearest cent.
> >Cheers > >Phil Hobbs
-- Father Brown's figure remained quite dark and still; but in that instant he had lost his head. His head was always most valuable when he had lost it.
On Monday, September 13, 2021 at 7:49:39 AM UTC+10, jla...@highlandsniptechnology.com wrote:
> On Sun, 12 Sep 2021 14:06:44 -0700 (PDT), "John Miles, KE5FX" > <jmi...@gmail.com> wrote: > > >On Sunday, September 12, 2021 at 1:10:47 PM UTC-7, jla...@highlandsniptechnology.com wrote: > >> Parts research is boring. > > > >I think there's some truth to every reply in the thread, which is > >darned rare around here. Yes, parts research is tedious, but it's > >also what product development *is* these days. I spend as > >much time looking for parts as I do using them. > > Then you could design more if someone did parts research for you.
John Larkin doesn't seem to do design. If you find a new part that fits you problem you can design a new - and unexpected - solution. An aspect of parts research is finding novel parts that can let you implement a new solution. I got to publish my 1996 millidegree thermostat paper in part because I used a newly available cheap 20-bit A/D converter to converter the voltage output from the thermistor into a digital number with enough bits to let me get to milli-degree accuracy. One of the software guys asked me why I wasn't using a bridge circuit, and I had to point out that the A/D converter used the same voltage reference as was connected to the resistor in series with the thermistor, so it was effectively the other arm of the bridge. Delegating finding new parts doesn't sound like a good idea. Delegating finding second sources and alternatives might work - to some extent - but tricks like double-pinning to accommodate two functional alternatives with different pin-outs might get missed. <snip>
> Lots of big companies have components engineers. Big mistake?
Depends on exactly what they are doing. -- Bill Sloman, Sydney
On a sunny day (Sun, 12 Sep 2021 13:10:38 -0700) it happened
jlarkin@highlandsniptechnology.com wrote in
<ahnsjgl1mqcb9s7fsde3engj9rkkgl9cb3@4ax.com>:

>Parts research is boring.
No, it may give you new ideas
On Mon, 13 Sep 2021 07:44:38 GMT, Jan Panteltje
<pNaonStpealmtje@yahoo.com> wrote:

>On a sunny day (Sun, 12 Sep 2021 13:10:38 -0700) it happened >jlarkin@highlandsniptechnology.com wrote in ><ahnsjgl1mqcb9s7fsde3engj9rkkgl9cb3@4ax.com>: > >>Parts research is boring. > >No, it may give you new ideas
Sometimes. But finding sources for a suitable, reasonably priced, multi-sourced SPDT relay is unlikely to be inspirational. Or a jumper cable. Or a #2 stainless spacer. Or an SMA feedthru with a cap and chain. I do periodically review the "new products" pages on the semi web sites, and check the few, mostly pitiful, ee news sites. Most of my inspiration comes from customer problems. I have to design 8 very demanding boards in the next 2 months, world-class picosecond stuff. And architect two new product lines. Hunting for relays and SMA caps is just a waste of time. -- Father Brown's figure remained quite dark and still; but in that instant he had lost his head. His head was always most valuable when he had lost it.