Searching \ for '[EE]: Board House Quality' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/pcbs.htm?key=board+house
Search entire site for: 'Board House Quality'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Board House Quality'
2005\10\03@123419 by Shawn Yates

flavicon
face
Perhaps this should be under OT instead of EE but...
    Is it me, or is the PCB industry the only industry that seems to
have given itself the right to make a product, but then charge extra to
make sure they did it right?  I am referring to the extra charge for
electrical testing of a bare board.  It just seems wrong.  What if cars
were made this way.  You could buy a car for X, but for X + Y you can
buy a car that was tested and they are sure it will run when you turn
the key.   What's up with that?   Shawn

2005\10\03@124948 by Michael Rigby-Jones

picon face


>-----Original Message-----
>From: spam_OUTpiclist-bouncesTakeThisOuTspammit.edu [.....piclist-bouncesKILLspamspam@spam@mit.edu]
>Sent: 03 October 2005 17:24
>To: Microcontroller discussion list - Public.
>Subject: [EE]: Board House Quality
>
>
>
>Perhaps this should be under OT instead of EE but...
>
>    Is it me, or is the PCB industry the only industry that
>seems to have given itself the right to make a product, but
>then charge extra to make sure they did it right?  I am
>referring to the extra charge for electrical testing of a bare
>board.  It just seems wrong.  What if cars were made this way.
> You could buy a car for X, but for X + Y you can buy a car
>that was tested and they are sure it will run when you turn the key.  
>
>What's up with that?  

Testing the boards is a time consuming operation.  If they made testing mandatory you would be paying a lot more for all your PCB's.  By making it optional, you can choose not to pay for it and get cheaper boards as a result. If you have a simple board with relaxed design rules, chances are your boards will be perfectly ok.

Regards

Mike

=======================================================================
This e-mail is intended for the person it is addressed to only. The
information contained in it may be confidential and/or protected by
law. If you are not the intended recipient of this message, you must
not make any use of this information, or copy or show it to any
person. Please contact us immediately to tell us that you have
received this e-mail, and return the original to us. Any use,
forwarding, printing or copying of this message is strictly prohibited.
No part of this message can be considered a request for goods or
services.
=======================================================================

2005\10\03@131706 by Wouter van Ooijen

face picon face
>     Is it me, or is the PCB industry the only industry that seems to
> have given itself the right to make a product, but then
> charge extra to make sure they did it right?

No, the software industry has embraced this idea and brought it to
perfection (that is, from the vendors point of view).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\03@145407 by Mark Scoville

flavicon
face
Hi Shawn, I have brought up the same issue with our purchasing guy. It just
strikes me as incredibly screwy. We pay a board house money to build a board
to our specifications. When we find boards that have shorts or opens the
purchasing guy "offers" me the option of electrical testing at increased
cost. So let me see if I have this right - you can charge us extra to make
sure that you did your job right? Funny. We can't charge our customers extra
for equipment that is "right" - they paid for the equipment to be right in
the first place. Somehow the PCB manufacturing industry has developed a
mindset (and the consumers have bought into it) where they just run the
boards through the process and you get what you get. If you want them tested
you pay extra.

We typically find about 1% of the boards we populate end up having some sort
of PCB defect (short or open). The really tough ones to find are the ones
where adjacent traces are shorted, but nothing is visible. However, a little
excavating with an X-acto knife between the adjacent traces and the short
disappears. We can't justify paying extra for electrical testing. So we live
with it. We can't change vendors because our parent company has dictated
that we purchase our boards from some outfit in China. Boards are dirt
cheap - and look really nice. But, we do have the 1% stinkers.

Yeah, I know it would add cost to electrically check the boards. Microwave
ovens, TV's, Cars - It adds cost to almost every product on the planet to
check and make sure it is right before shipping - doesn't it? The difference
seems to be that PCB "customers" will "risk" crappy boards to save a few
pennies (or bucks). Another factor is probably the fact that the average PCB
vendor isn't making a zillion of any particular board. If I was making PCBs
I don't think I could burden the cost of testing every board that went
through.

I suppose ultimately the PCB industry was "given the right" to charge extra
for testing because *most* customers (like the company I work for) refuse to
pay for it. If the PCB houses can't spread the cost of electrical testing
between all customers, then the customers that really want it will have to
pay for it. And the customers that don't pay for it - don't get it.

We (you too) do have the option of investing in equipment to do incoming
electrical inspection on the boards received. But that just pushed the cost
up on each board indirectly didn't it? Fact of the matter is you're going to
pay one way or another.

Choices seem to be...

A. Pay for PCB house to do electrical testing.
B. Do electrical testing yourself & reject bad boards before populating
them.
C. Assume boards are OK, and find the bad ones in final test after the
boards are populated.

(Sigh - Gotta switch to caffeine free soda) Sorry, I kinda got carried
away...

Just get used to it - this seems to be the way things work in the PCB world.

Anyway, I'm still trying sell my boss on the idea of paying me extra for me
to make sure I do my job right.

-- Mark


> {Original Message removed}

2005\10\03@151251 by olin piclist

face picon face
Mark Scoville wrote:
> So let me see if I have this right - you can
> charge us extra to make sure that you did your job right?

No, they give you of not forcing you to pay for the guarantee if you don't
want to.

> We can't justify paying extra for electrical testing.
> So we live with it.

But you are ridiculing that this option exists, even though you've already
voted with your money to take advantage of it?

> Boards are dirt cheap - and look really nice. But, we
> do have the 1% stinkers.

You've already shown that dirst cheap and 99% correct is more desirable than
more expensive and 100% correct, and that's what you're getting.  So what's
the problem?  You're getting what you're paying for.

> Yeah, I know it would add cost to electrically check the boards.
> Microwave ovens, TV's, Cars - It adds cost to almost every product on
> the planet to check and make sure it is right before shipping - doesn't
> it? The difference seems to be that PCB "customers" will "risk" crappy
> boards to save a few pennies (or bucks).

Yes, just as you have chosen to do.

> I suppose ultimately the PCB industry was "given the right" to charge
> extra for testing because *most* customers (like the company I work
> for) refuse to pay for it.

Exactly.  I don't understand why any of this is even an issue.  You have an
additional choice that you don't get with most products.  Nobody is taking
anything away.  If you don't like it, don't chose it.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\10\03@163725 by Wouter van Ooijen

face picon face
> We can't justify paying extra for electrical testing.

So why do you complain that you get untested boards? It's your choice!

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\03@203812 by Chen Xiao Fan

face
flavicon
face
For PCBs, lots of the PCB companies are struggling. Recently,
one of our approved PCB vendor moved to China and closed its
Singapore plant and lay off 700 people. Before that they
closed down its US plant one year ago.

Basically you pay what you get. And normally we get 99% good
board without electrical testing for 2-layer boards. For
4-layer boards, we do 100% electric testing since it
is relative expensive to throw away the 4-layer PCBs.

All in all, over competition is the culprit in lots of industries,
including the electronics component industry, the consumer
electronics industry and other sectors. That is why your
Japanese brand electronics gadgets (used to mean high quality)
normally will break slightly longer than the warranty period.
You Dell/HP/Acer/Whatever Brand notebooks have a good chance
of break after 1 year or 3 year depends on warranty period.
I used to believe this is conspiracy theory. However I
recently speak to some insiders and they tell me it is
almost truth --> it is because of the falling margin
and cheap designs will always have reliability problem.
There are so many recalls of power adapters. Why? The
design is to blaim. In the end, the end user is to blaim.

The customers are to blame. We are not willing to pay higher
price to get 100% solid units. We tend to buy those cheap
units with 60% of the price of a solid units and then
complain that it is not as good as the 100% solid units.
Therefore the 100% solid brand will suffer and it will start
to sell 90% solid product at 60% price of the 100% solid
units. Later 90% solid becomes 80% and lower and lower.
Dell is the leader of PC industry because it understands this
earlier than IBM/HP.

Outsourcing is another field where customers should complain.
Often a technical support call bring you to an Indian with
strange accent (some of them are okay though) and who have no
ideas of what you are talking about.

Anyway an Indian with strange accent is are still better
than those forever automatic response (in perfect English)
of "Please press 1 for xxx, 2 for yyy, 3 for zzz,
..., 0 for a real person " and then "our customer service
representatives are all engaged now, please wait" and then
after 30 minutes all so of lousy music, "Sorry I can not help
you here. I need to talk to my manager, please call again
later". That was my experience with Wasington Mutual
which is actually quite an okay bank. It was easy to go
to the bank to solve the problem but the nearest bank
is 300 miles away.

Software industry is another story because of the so called
IP, you do not own the software you purchase with your own
hard cash. Almost all software (free or not free) are
provided "AS IS".

Regards,
Xiaofan

{Original Message removed}

2005\10\03@220150 by William Chops Westfield

face picon face
On Oct 3, 2005, at 11:54 AM, Mark Scoville wrote:

>  We can't justify paying extra for electrical testing.

There's a third alternative of doing bare board testing on your side
of the manufacturing process.

I suspect part of the reason for not providing testing by default is
that customer "specifications" for testing vary a great deal.  There
is no data in the formats accepted by PCB vendors that say which traces
are supposed to be connected to each other (or not), so implementing a
test means transferring a whole new set of data, with room for errors
in the data and now a "you tested it but it still doesn't work; I'm not
paying!" sort of mentality from the customer.  ("oh, I guess I did
forget
to connect the power supply to that one chip.  Never mind.")

The "cheap pcb" forum over on sparkfun.com is interesting reading.
Originally designed to offer hobbyists $2.50/in^2 PCBs on a "you
give us the files, we give you a matching PCB" basis, they ended up
having SO many stupid errors and questions that they had to double
their price just to pay for the time checking designs to make sure
they weren't obviously broken.

One likes to think that a PCB design is done, and you get the same
board matching your cad program from every vendor out there.  But
there's
enough lack of standardization that this just isn't so...

BillW

2005\10\03@222913 by Mark Rages

face picon face
On 10/3/05, William Chops Westfield <westfwspamKILLspammac.com> wrote:
> On Oct 3, 2005, at 11:54 AM, Mark Scoville wrote:
>
> >  We can't justify paying extra for electrical testing.
>
> There's a third alternative of doing bare board testing on your side
> of the manufacturing process.
>
> I suspect part of the reason for not providing testing by default is
> that customer "specifications" for testing vary a great deal.  There
> is no data in the formats accepted by PCB vendors that say which traces
> are supposed to be connected to each other (or not), so implementing a
> test means transferring a whole new set of data, with room for errors
> in the data and now a "you tested it but it still doesn't work; I'm not
> paying!" sort of mentality from the customer.  ("oh, I guess I did
> forget
> to connect the power supply to that one chip.  Never mind.")
>

According to the "Art of Electronics", the circuit boards are tested
against each other, not the design files.  This works in practice
because the kind of problems you are testing PCBs for are random and
rare.

Somebody correct me if this information is no longer true..

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2005\10\03@225851 by Denny Esterline

picon face
I think you got this all backwards. They don't charge extra for testing -
they give you a discount for accepting them untested.

-Denny


You can get a good look at a t-bone steak by sticking your head up a cows
backside, but I'd rather take my butcher's word for it.

2005\10\04@001745 by PicDude

flavicon
face
On Monday 03 October 2005 12:17 pm, Wouter van Ooijen scribbled:
> No, the software industry has embraced this idea and brought it to
> perfection (that is, from the vendors point of view).


Ahhhh --- my sentiments exactly!  Why is it that the software industry can
deliver something that does not work as advertised, and not be forced to fix
it by law.  They'll only fix it if there is a threat of losing market-share,
but mostly they'll require you to pay more for an "upgrade".

Cheers,
-Neil.


>
> Wouter van Ooijen
>
> -- -------------------------------------------
> Van Ooijen Technische Informatica: http://www.voti.nl
> consultancy, development, PICmicro products
> docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\04@011609 by Chen Xiao Fan

face
flavicon
face
Because you do not own the software you buy. Buying
a software allows you to use the software (the rights
to use the software as a licensed user) but the software
itself is still belong to the vendor.

*******From Microsoft Windows EULA:****************
"3. RESERVATION OF RIGHTS AND OWNERSHIP.
 Microsoft reserves all rights not expressly granted to you
 in this EULA.  The Software is protected by copyright and
 other intellectual property laws and treaties. Microsoft
 or its suppliers own the title, copyright, and other
 intellectual property rights in the Software.  The
 Software is licensed, not sold."
***************************************************

Firmware is a bit different. It is part of the hardware and
the developers need to ensure the quality of the firmware
especially when it is not possible to upgrade the firmware
in an easy way. Lately PC peripheral vendors also learn
from software vendors so that drivers are without any
warranties as well. So as long as the hardware is
without major problems, the hardware does not need to
guarantee to work when used with the driver together.

Can we consider Windows kind of firmware for a PC preloaded
with Windows? Then we can return the PCs with serious bugs.
PC vendors only offer hardware warranty but not software
warranty.

It is reported in C/Net (news.com) that People in some wealthy
countries have thrown out Windows PC (keep the LCD) plagued
with virus/spyware and buy a new PC instead since it is cheaper
to get a new low end PC (US$300-500) than to get someone
to fix it. I am not so sure whether it is true or not.
Maybe it is better to donate them to some charity and install
Linux on it.

Anyway almost all the free and non-free software are without
warranty. So the companies can charge money for supporting the
not-working-as-advertised software and they call this "service".

Regards,
Xiaofan



{Original Message removed}

2005\10\04@012816 by PicDude

flavicon
face
Isn't this only a recent change (past few years), and this is MS -- not across
the board for all software companies.

Cheers,
-Neil.



On Tuesday 04 October 2005 12:16 am, Chen Xiao Fan scribbled:
{Quote hidden}

> {Original Message removed}

2005\10\04@020221 by Wouter van Ooijen

face picon face
> All in all, over competition is the culprit in lots of industries,
> including the electronics component industry,

I don't agree

> In the end, the end user is to blaim.

That I do agree! If the end user prefers a laptop that is somewhat
chaper but works only 3 years, that's what he will get (and it is what
he deserves).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\04@020724 by Chen Xiao Fan

face
flavicon
face
Here are two examples. Therefore it is better not to use
MPLAB C18 for anything critical. ;-)

Linux core and lots of open source software are licensed
under GPL version 2.

So for software, you can not guarantee to get what you want
no matter you pay or not pay.

Regards,
Xiaofan

*******************************************************
Microchip MPLAB C18 license (not free):

9.        NO WARRANTY. TO THE MAXIMUM EXTENT PERMITTED BY
LAW, COMPANY AND ITS LICENSORS PROVIDE SOFTWARE "AS IS"
AND EXPRESSLY DISCLAIM ANY WARRANTY OF ANY KIND, WHETHER
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE, OR NON-INFRINGEMENT. YOU ASSUME THE
ENTIRE RISK ARISING OUT OF USE OR PERFORMANCE OF SOFTWARE,
AS WELL AS ANY DERIVATIVES OF THE SOFTWARE MADE FOR YOU
OR ON YOUR BEHALF. COMPANY AND ITS LICENSORS ASSUME NO
RESPONSIBILITY FOR THE ACCURACY OR ERRORS OR OMISSIONS
OF SOFTWARE AND DO NOT WARRANT THE FOLLOWING: (A) THE
FUNCTIONS CONTAINED IN SOFTWARE WILL MEET YOUR
REQUIREMENTS, (B) THE OPERATION OF SOFTWARE WILL BE
UNINTERRUPTED OR ERROR-FREE, OR (C) ANY DEFECTS IN
SOFTWARE WILL BE CORRECTED.

*******************************************************
GNU GPL version 2 (open source)

NO WARRANTY

 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU.  SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.

 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.

*******************************************************

-----Original Message-----
From: PicDude [.....picdude2KILLspamspam.....narwani.org]
Sent: Tuesday, October 04, 2005 1:29 PM
To: Microcontroller discussion list - Public.
Subject: Re: [EE]: Board House Quality


Isn't this only a recent change (past few years), and this is MS -- not
across
the board for all software companies.

Cheers,
-Neil.

2005\10\04@045319 by Alan B. Pearce

face picon face
>I suspect part of the reason for not providing testing
>by default is that customer "specifications" for testing
>vary a great deal.  There is no data in the formats
>accepted by PCB vendors that say which traces
>are supposed to be connected to each other (or not),

I think you will find that the test matrix is derived from the netlist for
the PCB, so any errors in that (the net list does not show that pin
connected to anything) would be your responsibility, not the PCB house.

2005\10\04@073656 by olin piclist

face picon face
William Chops Westfield wrote:
> There
> is no data in the formats accepted by PCB vendors that say which traces
> are supposed to be connected to each other (or not),

Actually board houses routinely infer this from the Gerber files.

*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\10\04@074939 by olin piclist

face picon face
> Ahhhh --- my sentiments exactly!  Why is it that the software industry
> can deliver something that does not work as advertised, and not be
> forced to fix it by law.  They'll only fix it if there is a threat of
> losing market-share, but mostly they'll require you to pay more for an
> "upgrade".

It's called the "free market".  If you don't like the deal offered, don't
buy it.  If you do, then obviously you think the tradeoff is still to your
advantage in the end.  In either case its pointless complaining about it
here (and tiring for the rest of us to keep hearing it).  Complain to the
vendors.  Since it is a free market, they will change their ways if they
think they can make more profit by doing so.  If you're the only voice out
of a million customers then too bad, it's not worth serving your needs.
That's the free market too.  Get over it already.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

2005\10\04@084419 by alan smith

picon face
I believe they electricaly test against the gerbers, not a board



               
---------------------------------
Yahoo! for Good
Click here to donate to the Hurricane Katrina relief effort.

2005\10\04@101117 by Timothy J. Weber

face
flavicon
face
Chen Xiao Fan wrote:
> It is reported in C/Net (news.com) that People in some wealthy
> countries have thrown out Windows PC (keep the LCD) plagued
> with virus/spyware and buy a new PC instead since it is cheaper
> to get a new low end PC (US$300-500) than to get someone
> to fix it. I am not so sure whether it is true or not.

It's certainly true here in New York.  Two factors come to mind: (1) the
going rate for professional computer repair work seems to be >=
$60/hour, and (2) the average person is still pretty uncertain what the
difference is between hardware and software, much less what exactly a
virus does.  So a lot of people think if their computer is "infected,"
that's that.  They don't really understand that if you reinstall the OS
it will be just the same as it was new.

> Maybe it is better to donate them to some charity and install
> Linux on it.

I would say so.  On the other hand, even the charities here want to set
up machines to run Windows, because that's what most their recipients
know if they know anything.

To get back on topic, my take on this is that it simply costs less to
diagnose and fix or replace things that are observed to break, at a high
level, than it does to do rigorous testing to find low-level problems in
all constituent parts.

Here's a little numeric model off the top of my head:

- fraction of boards that have a fault = f

- cost of the bare board = b

- cost of testing a board before assembly = t

- cost of everything else that goes into a unit, assembled to the point
where it can be functionally tested = e

This assumes that we're always going to functionally test units once
they're assembled.

If your strategy is S1 = "pre-test all boards, and use only boards that
work" then you will incur cost (b + t + e) on all good boards, but just
(b + t) on bad boards.  So average unit cost is:

       f(b + t) + (1 - f)(b + t + e)

If your strategy is S2 = "functionally test all units after assembly,
and throw away the ones that fail," you will incur cost (b + e) on good
boards, and also (b + e) on bad boards.  So average unit cost is:

       b + e

In order for strategy S1 to be cheaper than S2, we must have

       f(b + t) + (1 - f)(b + t + e) < b + e

Simplifying:

       fb + ft + b + t + e - fb - ft - fe < b + e

       b + t + e - fe < b + e

       t - fe < 0

       t < fe        

So let's say f, the board failure rate, is 1%.  Then pre-testing makes
sense if, for instance, your parts and assembly together cost > $100 and
pre-testing adds < $1 to your board cost.  This seems fairly unusual
given what I'm hearing about testing costs.

And note that if you decide that repair is cheaper than discarding
defective units, or if you allow for failure in components other than
the board, pre-testing gets even less attractive.

I better get back to work.
--
Timothy J. Weber
http://www.lightlink.com/tjweber

2005\10\04@111140 by Spehro Pefhany

picon face
At 10:11 AM 10/4/2005 -0400, you wrote:

>And note that if you decide that repair is cheaper than discarding
>defective units, or if you allow for failure in components other than the
>board, pre-testing gets even less attractive.

Another issue is that if you don't find the fault in a defective unit, you
obviously don't know for sure what the cause was. That means there could be
issues in other parts, in the assembly process or whatever. If you don't
understand it, there's no way to be sure it won't be 0.5% today and 5%
tomorrow, or a tolerable 0.5% in house and a disastrous 0.5% in the field.

Thus, I prefer to diagnose 100% of faults as quickly as possible, even if
the boards are likely destroyed in the process. The ones that are due to
microscopic whisker shorts under component bodies can really take an
inordinate amount of time from a skilled technician, so having the boards
pre-tested can be worthwhile even if the cost of discarded boards and
other materials and assembly labor is minimal.

The disadvantage of pre-tested boards is that you don't know how crummy
their yield was. That could mask some pretty awful boards that will later
fail in the field. That's where dealing consistently with reliable suppliers
comes in- hopefully if they have a problem with something like plated hole
edge cracks (essentially invisible to casual inspection) they'll junk the
entire batch.

If you do start seeing board faults in pre-tested boards, it's a real
warning sign. Shorts would probably indicate the testing was not done
thoroughly, and opens could be advance warning of an impending disaster.

>Best regards

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
EraseMEspeffspam_OUTspamTakeThisOuTinterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff


2005\10\04@122140 by Timothy J. Weber

face
flavicon
face
Spehro Pefhany wrote:
> Another issue is that if you don't find the fault in a defective unit, you
> obviously don't know for sure what the cause was. That means there could be
> issues in other parts, in the assembly process or whatever. If you don't
> understand it, there's no way to be sure it won't be 0.5% today and 5%
> tomorrow, or a tolerable 0.5% in house and a disastrous 0.5% in the field.

I like that logic.  And that phrase says a lot right there - "a
disastrous 0.5% in the field."  Obviously the cost of a failure in the
field is high for you, but that really should be taken into account as a
variable in a more complete model - some may feel that 0.5% failure rate
is acceptable versus the cost of reducing it further.
--
Timothy J. Weber
http://www.lightlink.com/tjweber

2005\10\04@124631 by gacrowell

flavicon
face

> [piclist-bouncesspamspam_OUTmit.edu] On Behalf Of Alan B. Pearce
> Sent: Tuesday, October 04, 2005 2:53 AM
> Subject: Re: [EE]: Board House Quality
>
> >I suspect part of the reason for not providing testing
> >by default is that customer "specifications" for testing
> >vary a great deal.  There is no data in the formats
> >accepted by PCB vendors that say which traces
> >are supposed to be connected to each other (or not),
>
> I think you will find that the test matrix is derived from
> the netlist for
> the PCB, so any errors in that (the net list does not show that pin
> connected to anything) would be your responsibility, not the
> PCB house.

And

> [@spam@piclist-bouncesKILLspamspammit.edu] On Behalf Of Olin Lathrop
> Sent: Tuesday, October 04, 2005 5:38 AM
> Subject: Re: [EE]: Board House Quality
>
> Actually board houses routinely infer this from the Gerber files.


Correct.

We seldom deal with the low-end board houses.  But most others require
an IPC-371 (IPC-371A preferred) netlist to be provided with the board
Gerbers.  There are several ways to obtain this netlist.  Many CAM
programs will generate it from the Gerber files.  For that matter, you
can instruct the board house to generate it themselves from the Gerber
files.  (Note however, that the Gerber files themselves have no inherent
connectivity information - if two Gerber-defined shapes 'touch' they are
assumed to be correctly connected [though sometimes there can be
additional conditions imposed, depending on the generating software].)

This all assumes that the CAD software that generated your Gerber files
indeed generated what you wanted.  What you see on your CAD screen, may
not be what the Gerbers look like after they are generated (see the
current thread relating to software quality).  That's why it's a good
idea to look at and error check the CAD-generated Gerber data with a CAM
program (CAM350, GerbTool, et al.).  

The *best* place to generate an IPC-371 netlist is from the original CAD
data.  That's the source that has the original connectivity data that
you designed.  Some CAD software has this ability built-in, some have it
available from third party vendors, and some just don't.  A good check
is to compare the CAD-generated '371 netlist to the CAM-generated '371
netlist (belt and suspenders).

If you send your CAD '371 netlist to the board house, they will (should)
routinely check it against their own CAM generated '371 netlist and stop
the fab if there are any discrepancies.  (Hence those late-night calls:
"Your board is on hold ...")

There is another aspect of bare-board testing that affects the cost.  If
you are doing a low volume proto order, most likely it will be tested on
a flying-probe tester.  This is a robot that literally touches a
ohmmeter to the pads, net-by-net, checking for connectivity and shorts
just like a person would do (albeit faster).  On a complex board, this
can be a slow process.  But, after generating the flying-probe program
from the netlist and CAM data, the process is largely automatic.

The alternative testing process is to build a test fixture.  This is a
costly process, so its usually done only on high-volume boards.  The
fixture touches a probe to all test points on the pcb simultaneously, so
a full test takes only seconds per board.  The trade-off decision
between flying-probe or test fixture is a function of board
complexity/test-time/volume.  The downside of test-fixtureing is that
any change to subsequent revs of the board tosses out a multi-hundred$$
fixture.  However if you go back to the same board house for additional
purchases of the same (unchanged) board, you shouldn't have to pay the
NRE of a fixture again.

Note that you can usually visually tell if a bare board has been tested;
the probes will leave a mark on the pad at each probe point.

Note that there is an IPC-371B spec out now, but I haven't seen it in
common use yet.

Note that the ODB++ data format (originated by Cadence) is sometimes
preferred by board houses.  It replaces Gerber data, and *does* include
netlist information.  Some common CAD systems can generate ODB data, but
there still seems to be some incompatibilities with data generated by
non-Cadence systems.  It's a good idea to send a set of test data to a
board house first, before your time-critical board order, to make sure
the data exchange is working.  Some board houses give discounts for
orders supplied with ODB data, because it reduces their up-front CAM
time.

Note that there is a new data format being prepared by IPC, called
Offspring.  It has great promise of incorporating all pcb data into a
single common format; we'll see.

Final notes (I promise).  You should always be sending RS-274X format
Gerber data, which imbeds the aperture information.  Otherwise, with
RS-274D (or just 'RS-274') you allow a great risk of an aperture error.
Some CAD/CAM systems do have the ability to imbed netlist information in
the Gerber file, but it's done as comments I believe, and is totally
non-standard and non-exchangeable between systems.

Gary Crowell, CID
Micron Technology

2005\10\04@131059 by Bob Blick

face picon face
Gary writes:

> We seldom deal with the low-end board houses.  But most others require
> an IPC-371 (IPC-371A preferred) netlist to be provided with the board
> Gerbers.

Did you mean IPC-356A? I've made those, but I've never heard of IPC-371A.

-Bob



2005\10\04@131542 by Spehro Pefhany

picon face
At 10:46 AM 10/4/2005 -0600, you wrote:


>We seldom deal with the low-end board houses.  But most others require
>an IPC-371  <snip>

Do you mean IPC-D-350/356?

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
KILLspamspeffKILLspamspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




2005\10\04@132057 by gacrowell

flavicon
face
Dain-Bramage, of course, -356.

GC

{Quote hidden}

> --

2005\10\04@151609 by gacrowell

flavicon
face


> [TakeThisOuTpiclist-bouncesEraseMEspamspam_OUTmit.edu] On Behalf Of Spehro Pefhany
> Sent: Tuesday, October 04, 2005 11:24 AM
> To: Microcontroller discussion list - Public.
> Subject: RE: [EE]: Board House Quality
>
> At 10:46 AM 10/4/2005 -0600, you wrote:
>
>
> >We seldom deal with the low-end board houses.  But most
> others require
> >an IPC-371  <snip>
>
> Do you mean IPC-D-350/356?

Yes, I've been dealing with IPC-7351 lately and the those are the digits
that stuck in my mind.  Senior Moment; in fact, its my birthday.

GC

2005\10\04@155152 by Howard Winter

face
flavicon
picon face
Wouter,

On Tue, 4 Oct 2005 08:02:14 +0200, Wouter van Ooijen wrote:

>...
> > In the end, the end user is to blaim.
>
> That I do agree! If the end user prefers a laptop that is somewhat
> chaper but works only 3 years, that's what he will get (and it is what
> he deserves).

That idea falls down because it doesn't say on the box "Will fall apart after 3 years", does it?  When a house
is built in Britain it is "Guaranteed" by the National House Building Council for 10 years - that doesn't mean
it will fall down after 3651 days, it means that it was built to a design and with the right materials and
skill that they are prepared to fix anything that fails due to a lack of those things for 10 years, so the
buyer doesn't have to worry about it.  The expectaion is that as long as the house is maintained properly, it
will last "for ever".

As a consumer there is no way to tell if something is made from inferior or short-lived materials, apart from
the reputation of the manufacturer, if there is any.  The old saying "You get what you pay for" is flawed -
you don't usually get *more* than you paid for, but you can easily get less.  Paying more is no guarantee that
you will get something better than if you pay less, because manufacturers charge as much as "the market will
stand" and if you have a good name you can charge more without actually selling something better.

So, back to laptops:  Which manufacturer would you buy from (and pay extra) to get something that lasts longer
than average?

Cheers,





Howard Winter
St.Albans, England


2005\10\04@160734 by Mark Rages

face picon face
>
> So, back to laptops:  Which manufacturer would you buy from (and pay extra) to get something that lasts longer
> than average?
>

I don't know how this plays into your thesis, but I bought an IBM
laptop after seeing my friend's 486-based IBM laptop in daily use and
working perfectly. My laptop still works fine, but it still has one
year left until the fateful three-year birthday.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2005\10\04@160939 by Wouter van Ooijen

face picon face
> That idea falls down because it doesn't say on the box "Will
> fall apart after 3 years", does it?

Does it state the opposite?

And can't "Will fall apart after 3 years" be considered common
knowledge?

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> because
> manufacturers charge as much as "the market will
> stand" and if you have a good name you can charge more
> without actually selling something better.

not for long, if customers communicate their experiences

> So, back to laptops:  Which manufacturer would you buy from
> (and pay extra) to get something that lasts longer
> than average?

Dell, IBM, Grid used to have good names, but at the moment I would not
know. I would probably just buy 3.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\04@162635 by PicDude

flavicon
face
On Tuesday 04 October 2005 03:07 pm, Mark Rages scribbled:
> > So, back to laptops:  Which manufacturer would you buy from (and pay
> > extra) to get something that lasts longer than average?
>
> I don't know how this plays into your thesis, but I bought an IBM
> laptop after seeing my friend's 486-based IBM laptop in daily use and
> working perfectly. My laptop still works fine, but it still has one
> year left until the fateful three-year birthday.


Same here -- I have used many laptops and have 3 here currently.  The ones I
hate are Compaq and Toshiba.  One here is a Dell, which has been "averagely"
used, but is falling apart -- mechanically and certain components dying.  My
IBM Thinkpad 600x (which I'm on now), has travelled with me extensively, and
in on almost always, has been a real workhorse and running smooth for many
years now (about 4 yrs old).  And I've used other IBM's as well with great
results.  Only real problem I've had is hard-drive failures, which are not
IBM-specific drives.  Everytime I look into a new laptop, I look at IBM
first.

Cheers,
-Neil.



2005\10\04@163056 by Spehro Pefhany

picon face
At 02:07 PM 10/4/2005 -0600, you wrote:
> >
> > So, back to laptops:  Which manufacturer would you buy from (and pay
> extra) to get something that lasts longer
> > than average?
> >
>
>I don't know how this plays into your thesis, but I bought an IBM
>laptop after seeing my friend's 486-based IBM laptop in daily use and
>working perfectly. My laptop still works fine, but it still has one
>year left until the fateful three-year birthday.

Seems to me that most portable consumer electronics (cell phones, MP3
players, notebook computers) typically has a rather limited lifetime (1-3
years) if
you actually haul it around all over the place. If it gets to sit on the
desktop and has an easy life, it should do a lot better.

There are currently some adverts for an especially robust notebook.. forget
who it is (also don't know how much fact is behind it).

Of course IBM's notebook division is now owned by the large Chinese computer
company Lenovo.

>Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
RemoveMEspeffspamTakeThisOuTinterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff


2005\10\04@170339 by Richard Prosser

picon face
Interesting.
I just had a desktop motherboard fail - at almost exactly 3 years!
(Mind the same PC already had a video card replaced under warranty and
a hard drive fail just one month past its 2 year warranty period).


RP

On 05/10/05, Spehro Pefhany <speffEraseMEspam.....interlog.com> wrote:
{Quote hidden}

> -

2005\10\04@180247 by Howard Winter

face
flavicon
picon face
Wouter,

On Tue, 4 Oct 2005 22:09:38 +0200, Wouter van Ooijen wrote:

> > That idea falls down because it doesn't say on the box "Will
> > fall apart after 3 years", does it?
>
> Does it state the opposite?

In the UK, the "Sale of Goods Act", 1893, amended in 1975, states that a consumer can expect goods to be of
"Merchantable quality and fit for the purpose".  I think "consumer durable" items such as we are discussing
would not be seen to be of merchantable quality if they all fell apart after three years.

> And can't "Will fall apart after 3 years" be considered common knowledge?

Absolutely not!  I've had my television for over 10 years, my washing machine for longer than that, I've had
laptops that lasted for 6+ years, a laser printer that's 5 years old and a colour laser that's probably twice
that.  Are you really saying that you expect to replace all your posessions every three years?  I've never
been paid enough for that!

Cheers,

Howard Winter
St.Albans, England


2005\10\04@181356 by Howard Winter

face
flavicon
picon face
Wouter,

On Mon, 3 Oct 2005 19:17:04 +0200, Wouter van Ooijen wrote:

> >     Is it me, or is the PCB industry the only industry that seems to
> > have given itself the right to make a product, but then
> > charge extra to make sure they did it right?
>
> No, the software industry has embraced this idea and brought it to
> perfection (that is, from the vendors point of view).

I resent that!  *Microsoft* has embraced and somehow made everyone expect it (one of their most heinous
crimes, IMHO) but it doesn't apply to the whole industry.  When I was developing bespoke software the quality
control was part of the job, and however much a client wanted to chop it out of the project, the firm I worked
for was adamant that they couldn't do that (we lost the odd order from real cheapskates, but that's the sort
of business you don't want).  Once it was delivered, already tested by us, they had a fixed (agreed) period of
time to test it themselves and report any shortcomings, which were fixed for free.  After that they were
deemed to have accepted it, because you can't have an open-ended liability - the accountants won't allow it!  
:-)  That's about as fair as you can get, I reckon!

Cheers,


Howard Winter
St.Albans, England


2005\10\05@001939 by Dmitriy Kiryashov

picon face
Hi Folks.

Discussing quality issues and control. There is approach called sigma
six which is quite good for keeping company's good reputation in event
of mass production counted in millions of units. Anybody heard of Dell,
IBM, Sony or others implementing that in their mass production ?


WBR Dmitry.

PS. I'm not so reach to buy cheap things ( well known old saying )



Wouter van Ooijen wrote:
{Quote hidden}

> -

2005\10\05@015649 by William Chops Westfield

face picon face
>> Which manufacturer would you buy from (and pay extra) to get
>> something that lasts longer than average?
>>
Why should a manufacturer, or for that matter a customer, apply extra
effort and cost to make/acquire a computer that will last more than 3
years when it will be pretty much obsolete in 2 years anyway (not
necessarily un-useful, but definately getting a bit dodgy...)?

cisco currently leases employee laptops for 30 months.  My old laptop
is coming up for renewal pretty soon now.  The replacement will have a
CPU that is more than twice as fast (1.7GHz P-M vs 900MHz P III), the
disk will be twice as big, it'll have a CD-RW/DVD combo drive, and it'll
run WXP instead of W2k.  I don't NEED all that, but that's a lot of
new stuff to have had happen in 30 months!  (and things have SLOWed
since the .com bubble burst; computers used to go obsolete even faster.)

BillW

2005\10\05@020411 by Wouter van Ooijen

face picon face
> In the UK, the "Sale of Goods Act", 1893, amended in 1975,
> states that a consumer can expect goods to be of
> "Merchantable quality and fit for the purpose".

The roblem with such acts (we have roughly the same) is that their
reference point is "what is common", so it does not help to slip the
industry as a whole towards lower lifetimes.

> I think
> "consumer durable" items such as we are discussing
> would not be seen to be of merchantable quality if they all
> fell apart after three years.

If they all do they effectively set the definition of "consumer
durable".

> Are you really saying that you expect to replace all
> your posessions every three years?  I've never
> been paid enough for that!

No, but I do expect a laptop to fail a lot sooner than my washing
machine (as it indeed does).

Don't take me wrong: I don't like this situation, but I don't think
there is much an individual can do about it, except voting with his
money.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\10\05@035424 by Chen Xiao Fan

face
flavicon
face
CISCO is a *BIG WEALTHY* company.

The Dell desktop I am using at work is more than 3 years old
(a OptiPlex GX150 Pentium III 1G with lousy built-in graphics).
It is quite okay actually for normal use (PSpice simulation, MPLAB,
schematics drawings, Outlook, Word, Excel, Customized Oracle
based applications, ...). The only complain from me is the
new web service based ERP program (host on IBM AS400 mid-range
servers). Before the switch, we use terminal emulation and
the interface sucks but it is really fast. Now the interface
(using Internet Explorer) is quite nice but the speed sucks.
However the speed bottleneck is not the client PC but the
network bandwidth. It seems to me I need to continue using
this PC for a while. The switch to XP from 98SE early this
year is quite smooth and I get an update of double the
SDRAM to 512MB. The mechanical developers are using much
faster PCs due to the use of ProEngineers which seems to
be always slow no matter using any PCs.

Of course I will prefer a fast PC and indeed I have much
faster PCs at home. I think the current P3 1GHz Dell
PC is still serviceable and I will prefer some better
equipment than a faster PC for my work. A faster scope,
for example, is much more important for an electronics
engineer.

Regards,
Xiaofan

{Original Message removed}

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...