Design Con 2015
Breaking News
Blog

Design Engineers Take On Component Police

NO RATINGS
Page 1 / 2 Next >
View Comments: Threaded | Newest First | Oldest First
DCH0
User Rank
Rookie
Design Engineers need to be responsible for shipping the product
DCH0   2/17/2014 12:08:59 PM
NO RATINGS
You take an adversarial approach.  What you call "Component Police" is actually just using good design practice while noticing that the product has to ship.  "Favored Vendors" are the ones who deliver product to production in time to meet customer requirements.  It also helps to be cost effective so we can pay overheads including ourselves.

I once looked at switching a diode package which would help design for manufacture.  It turned out the desired component which was an industry standard was 26 weeks lead time when we had 250K of the other in stock.  I did not change it and kept my friends in purchasing and logistics still as friends.

Component Injurnear
User Rank
Rookie
COPS Component Police
Component Injurnear   2/17/2014 12:59:57 PM
NO RATINGS
As with most things the Component Engineering function is a question of balance.

Considering the needs of Engineering, Sourcing, Quality, Test, Manufacturing, Field Service, Marketing and Manufacturer/Vendor.

Maintaining appropriate & consistent levels of quality.

Overall cost of component (initial price, long term price trending, quality, defect rate, short and long term availability, toxicity, ease of use, end of life terms, alternate sources that are form, fit & function compliant, agency approvals, industy standard, etc..).

Does the component lifecycle match your product life? Does the manufacturer discontinue almost as many components as they release?

Engineering investments in currently used components such as firmware, software, design expertise, mechanicals, unique functionality, regulatory compliance...

Fostering awareness of component lifecycles for various product areas, industrial, consumer, hi-rel, automotive, wireless, etc..

Monitors industry trends, memory types, decreasing lifecycles, cost trends, supplier shifts....

Online avalability of comprehensive component information, complete part number, data sheet, package dimensions, tape and reel dwg, pricing, inventory, sample ordering, lifecycle, levels of information provided that does not necessitate a call or email to manufactuer that is diverted to voicemail or answered by automated email away message.

My favorite manufacturer is those that meet all of the above and tries to understand the customer products and personalties, provide appropriate product information in a passive, effective manner, does not hound via phone calls, spam, unplanned visits. Does not feel the need to provide cheap pens that require 110 lbs of pressure to expel ink or conversely leak, various sizes of calenders, and other personalized landfill garbage from China.

I'm also available for hire, Componentengineering@yahoo.com

 

anon3887601
User Rank
Rookie
component police
anon3887601   2/17/2014 2:18:20 PM
NO RATINGS
At Northern Telecom everybody hated the Components Engineering department.  But I saw many times where they would assist and educate a young and green engineer who didn't understand how to specify components.

One of my favourites was when the quartz-crystal engineer had to educate some designers that crystal load capacitances were standardized and they couldn't just pick any value they liked.

Another time a colleague had to have a resistor go through the full qualification process, lasting several months.  While that resistor family was widely used in the company that *value* had never been used before.  He was told they couldn't expidite the process.

When Nortel started to collapse, Components Engineering were one of the first groups to be cut.  High management figured that function could be performed by the contract manufacturers then being brought in.  Bad mistake on top of another, and everyone knows the company's end result.

zeeglen
User Rank
Blogger
Re: component police
zeeglen   2/17/2014 3:29:47 PM
NO RATINGS
@anon the quartz-crystal engineer had to educate some designers that crystal load capacitances were standardized

Had a person in the procurement department ask why my VCXO had to be that exact and expensive part number when everybody else was using much cheaper voltage-controlled-crystal-oscillators.  I had to explain the difference between a frequency controlled type of oscillator for a PLL vs the types with a simple on/tristate control voltage.


cookiejar
User Rank
CEO
heck of a compoent list
cookiejar   2/17/2014 2:45:50 PM
NO RATINGS
Back about 1971, one of my co-workers was an incredible self taught analog designer.  He did his sophisticated analog designs using large graph paper on a drafting table, calculating all offsets and temperature drifts graphically.  His low power 16 bit A/D converters were stable to better than 1 LSb through the -55C to 125C temperature range of our Tenney chamber, performance unheard of in the day.
 
When he decided to move on to greener pastures, I volunteered to take over his latest project.  Going over his documentation, I was shocked to see that he had over 150 different resistor values on the board.  While I realized the need for critical values in his analog chain, I cornered him on why he had used different resistor values for every one of his pullups.  It seems he was an equal opportunity designer.  When picking resistors from our supply, he picked only those values that had the most stock in our bins, in other words those poor lonely values which nobody else used.  He had done the same for his selection of transistors.  To my relief he didn't object to my simplification of the parts list to a dozen resistor values and substituting 2N2222 and 2N2907A transistors where applicable.

While our development lab didn't have a component police force to serve and protect us,  I would certainly have welcomed one as a major time saver in picking components.

seaEE
User Rank
CEO
Protect and Serve
seaEE   2/17/2014 11:52:47 PM
NO RATINGS
Wow, a 74F logic device made by Signetics.  Beautiful!  The Signetics 4000 CMOS, HCMOS, and Philips FTTL books use to grace my databook shelf.  And they had a wealth of information in them, not just on parts, but also on topics like ESD and latchup.   It is difficult to do a component engineering job well.  The component engineer is between a rock and a hard place (Design Engineering and Procurement).  They need to ensure a part is chosen that will work in the circuit, is cost competitive, and will not cause supply chain problems.  If they do their job well, they get to focus more on the new design part of the job, and less on the sustaining engineering part. A good component engineer will help ensure that parts are operated at their Recommended Operating Conditions, and not at the Absolute Maximum Ratings where functionality and life may be impaired.  A good component engineer can be the difference between hunting down and finding a difficult to find replacement for an obsolete part in a circuit, and having to redesign the whole circuit because no replacement part can be found.  A good component engineer may know a manufacturer's product line better than the sales person for that line. 

The late Bob Pease addressed the subject in one of his monthly essays:

http://electronicdesign.com/archive/what-s-all-components-engineer-stuff-anyhow

antedeluvian
User Rank
Blogger
Re: Protect and Serve
antedeluvian   2/18/2014 2:58:47 PM
NO RATINGS
seaEE

Wow, a 74F logic device made by Signetics.  Beautiful!  The Signetics 4000 CMOS, HCMOS, and Philips FTTL books use to grace my databook shelf.  And they had a wealth of information in them, not just on parts, but also on topics like ESD and latchup. 

Did you see my blog "Preserving Data Books From Yesteryear"?

 

seaEE
User Rank
CEO
Re: Protect and Serve
seaEE   2/18/2014 11:09:49 PM
NO RATINGS
antedeluvian, I wish I still had those databooks.  Let's just say that my boss was one of those types that preferred a "clean office" to a "comprehensive office", which was understandable as data migrated to the internet. :)  I believe Signetics may have also had an ALS/AS TTL book, which I may also have had, and this  was probably a very rare book.  Hmmm...what is this?  A predecessor to the one I had appears to be on line:

http://ftp.df.lth.se/pub/bitsavers.org/pdf/signetics/_dataBooks/1989_ALS_Data_Manual.pdf

It looks like other Signetics databooks can be accessed by editing the link, and other companies databooks can be accessed by further editing.  It would be nice if some semi company could archive such data in an offical online museum.

It was kind of a badge of honor ro have a full set of books. 

I had TI's huge TTL book, RCA/Harris HC/HCT and AC/ACT logic books, National and Fairchild's set of logic books, and of course Motorola's.  My second level boss had worked for Motorola and designed some of their ECL logic devices.  Every once in awhile he would wander into our area sit down in my or my coworkers' guest chair, kick back, put his feet on our desk, pull out a stick of chewing gum with measured precision, and start lecturing about purple plague and rat-holing in silicon stepping.  He was a knowledgable and fun guy.

 It looks like you are creating quite a scanned in library.  Maybe there is some free cloud drive space you can put them on and make them accessable to the public.

antedeluvian
User Rank
Blogger
Re: Protect and Serve
antedeluvian   2/19/2014 10:18:41 AM
NO RATINGS
seaEE

Maybe there is some free cloud drive space you can put them on and make them accessable to the public.

There is something like that now on Internet Archive. Unfortnately I can't link to a particular comment to take you right there, but if you go back to my blog and look for my comment on 1/30/2014 you will find the links.

 

MeasurementBlues
User Rank
Blogger
Re: Protect and Serve
MeasurementBlues   3/6/2014 1:44:39 PM
NO RATINGS
I found this one in my basement parts bin. Looks like week 37 of 1978.


 The black stuff in the pins comes from the anitstatic foam that's dried out.

DrFPGA
User Rank
Blogger
PALs vs 7400
DrFPGA   2/18/2014 2:16:22 PM
NO RATINGS
One of the reasons I gave at customer trainings for switching to Programmable Array Logic from 7400 TTL was to get around the component police. Get them to add a few PAL devices and you never had to push for a new 7400 component to be added to the approved list again.

In fact, at the end of the training I would drop the TI 7400 data book into the trash can. Always got a good response...

RWatkins
User Rank
Rookie
Component Cops and Reality
RWatkins   2/19/2014 10:56:38 AM
NO RATINGS
In my experience when a company posts personnel as component cops, the persons occupying this position generally stop having the ability to keep up with current technology and trends in component design.  As a result, simple and valuable changes in the availability and cost of certain types of components escape them.  My experience came with a need some years ago to use a very low loss moderate value component in a surface mount application.  At the time, C0G/NP0 ceramics had not yet gotten as dense as today, so the only reasonable choice was a SMT film-foil part.  This initiated a huge fight with the passive component cop.  The saddest part of such thinking is that today, 14 years later, I could easily drop in a replacement 0805 or 1206 NP0 on the same pads if ever the film/foil went obsolete.  More so, the temperature rating of the available film/foil SMT devices then was not good enough for modern lead-free tin-copper-silver solder, so a change would have been required anyway.  What WAS important was using a somewhat standard package footprint that could easily be substituted for by a newer part with better specifications when available.  This component of design, the form-fit-function drop-in replacement path to future components is often greatly discouraged by component cops.

Having on several occasions helped an engineer whose main job was re-designing obsolete circuit boards, I saw how "standard" parts often go obsolete when new technologies replace them.  One classic example was the obsolescence of a "matched pair" of transistors in one of his circuits, where ON Semi had decided to no longer provide the matching service so common since the 1960's, and it was now 2001.  He was in a panic since he could not find similar parts from the several other vendors as well and was looking at a total circuit re-design until I showed him a single-die dual transistor with fully matching (better than prior devices) specifications.  I also showed him how this packaging scheme was replacing the matched pair approach.  He had to modify one footprint on the PCB to continue production.  Note that the original design was fully supported by the component cops, and we had to fight them to get dual transistors into the "system".  Hmmmm...

MeasurementBlues
User Rank
Blogger
Re: Component Cops and Reality
MeasurementBlues   3/6/2014 1:33:24 PM
NO RATINGS
When I worked at GCA, I was asked to find substitute parts for those than had gone obsolete. In some cases, there was no direct replacement, requiring a redesign. In this case, I was kind of the component cop dealing with purchasing. I did have to do some redesign to accomodate the new part, but purchasing and production were happy.

Flash Poll
Like Us on Facebook

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
EE Times on Twitter
EE Times Twitter Feed
Top Comments of the Week