Searching \ for '[EE] BOM management and contract manufacturing' 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/index.htm?key=bom+management+contract
Search entire site for: 'BOM management and contract manufacturing'.

Exact match. Not showing close matches.
PICList Thread
'[EE] BOM management and contract manufacturing'
2010\06\02@145910 by Mike Hord

picon face
Informal poll:

How does your organization handle the bill of material for your circuit
board assemblies?

Hypothetical methods:
1.  The assembly has a part number in your system.
2.  Each component (resistor, IC, etc) has a part number in your
system.  Alternates are controlled under this list.
3.  The BOM calls out the internal part numbers.

OR
1.  The assembly has a part number in your system.
2.  The CM has a list of approved parts for each component
on the BOM.  You don't worry about it.

Other options?

Mike H.

2010\06\02@154023 by Vitaliy

face
flavicon
face
Mike Hord wrote:
> Informal poll:
>
> How does your organization handle the bill of material for your circuit
> board assemblies?
>
> Hypothetical methods:
> 1.  The assembly has a part number in your system.
> 2.  Each component (resistor, IC, etc) has a part number in your
> system.  Alternates are controlled under this list.
> 3.  The BOM calls out the internal part numbers.
>
> OR
> 1.  The assembly has a part number in your system.
> 2.  The CM has a list of approved parts for each component
> on the BOM.  You don't worry about it.
>
> Other options?

Whenever possible, we let the CM source the parts. The BOM (a simple
spreadsheet) lists the parts, their values and tolerances. We also specify
whether substitutes are acceptable. By the way, we recently automated the
process of creating the BOM using component attributes (this is based on
Olin's suggestion -- a big thank you!).

When we provide the parts, we enter them into our inventory management
system (FishBowl) like we would any of the parts we sell or use in-house.

A lot of times we use a hybrid: we let the CM source most of the parts, and
we provide the rest (programmed PICs, or parts that we can get cheaper than
the CM).

Vitaliy

2010\06\02@160320 by Olin Lathrop

face picon face
Vitaliy wrote:
> By the way, we recently
> automated the process of creating the BOM using component attributes
> (this is based on Olin's suggestion -- a big thank you!).

In case others are interested too, this system uses Eagle attributes and
includes ULPs and host programs to automatically generate a Excel
spreadsheet from the information.  This is all available for free in the
Eagle Tools release at http://www.embedinc.com/pic/dload.htm.  After
installation, look at the EAGLE_BOM documentation file in the DOC directory
and follow the cookie crumbs from there.

This system is getting better and better for us as more library parts get
touched.  Of course new parts added to the libraries are now made with the
BOM information up front.  If used properly, pretty much the whole BOM
except for the price column can be generated automatically.  This used to be
a much more tedius process.  I still find a few parts every design that
don't have the attributes, but BOM generation is down to a few 10s of
minutes instead of hours now.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2010\06\02@174218 by Jesse Lackey

flavicon
face
Having not used Olin's system, and instead still making it by hand in
Excel, I can say with certainty to use Olin's system.  By hand is
tedious at best and often error-prone, including recently forgetting to
fix a mistake in the BOM (a 0.1uF and 22pF caps were switched) in a qty
1000 manufacturing run.  I caught it after about 1/2 the boards were
made, my contact at the CM literally jogged into the production area and
stopped the pick&place and changed the programming mid-panel.  I had to
swap them with a hot-air gun, pretty fast but there were ~500 to do...

So invest a bit of time up-front in this if you (the OP) use Eagle.

J



Olin Lathrop wrote:
{Quote hidden}

2010\06\02@193021 by Vitaliy

face
flavicon
face
Jesse,

Do you have anyone to help you proof the BOM? For some projects, we conduct
at least a dozen reviews. Review the schematic when it's "done", review the
layout when it's "done", review the BOM..  we discover lots of mistakes this
way. Not to say that some of them still make it to production, but this
"buddy system" definitely cuts down on the number of prototypes we have to
go through, and the severity of mistakes.

We used to check the "handcrafted" BOM against the EAGLE-generated BOM. It's
fastest when you have two people w/ highlighters, one calling out the part
names and the corresponding value/package/etc and one verifying. When you're
done, you should have all the line items highlighted.

Vitaliy


{Original Message removed}

2010\06\02@203001 by Jesse Lackey

flavicon
face
Hi - all a good idea, the 'proof' is when I (a) order the parts for the
protos and (b) assemble the protos (qty several - by hand).  I have a
very part time assistant now, but he's really new to all this, and it
will be awhile before he's able to do more than very basic checking.

I often catch errors in the proto assembly phase, in both BOM and just
other dumb physical problems, and this happens often enough that I don't
see how there can be a division between "design" and "proto build" in
other companies, or between "design" and "ordering" at least at the
prototype stage; maybe other EEs make fewer mistakes.  :)  Or between
schematic and layout for that matter.  In nearly every design I find
errors when I get to the detailed layout, dumb stuff, but real.

But even with proto assembly verification phase there are still mistakes
that get into production, and it is almost always last minute changes
done in too much of a hurry.  Fortunately my contact at the CM
(meritronics in san jose) is on IM and is a go-getter, can-do kind of
guy, and doesn't hesitate to ask me about anything odd.

Cheers
J



Vitaliy wrote:
{Quote hidden}

> {Original Message removed}

2010\06\03@045248 by Alan B Pearce
2010\06\04@144650 by Vitaliy

face
flavicon
face
Jesse Lackey wrote:
> Hi - all a good idea, the 'proof' is when I (a) order the parts for the
> protos and (b) assemble the protos (qty several - by hand).  I have a
> very part time assistant now, but he's really new to all this, and it
> will be awhile before he's able to do more than very basic checking.

He doesn't need to understand anything to help you check the BOM. During
crunch times when other engineers were not available, I enlisted the help of
our administrative staff. "Just read me each line." I think it's even better
that they don't understand what the stuff means, since they can't make
assumptions and "see" what is not there.


> I often catch errors in the proto assembly phase, in both BOM and just
> other dumb physical problems, and this happens often enough that I don't
> see how there can be a division between "design" and "proto build" in
> other companies, or between "design" and "ordering" at least at the
> prototype stage; maybe other EEs make fewer mistakes.  :)  Or between
> schematic and layout for that matter.  In nearly every design I find
> errors when I get to the detailed layout, dumb stuff, but real.

I didn't mean to make it sound like we follow the waterfall model. Of course
there is a lot of going back and forth. However it usually makes sense to
finalize the schematic as much as possible, before moving on to the layout
(since schematic changes are likely to impact the layout).

It's definitely easier when you have someone to do the checking with. It's
even better when you can involve a third person to do the final look-over,
because by the time you two think you're done, your eyes develop blisters
and you can miss even the obvious errors.


> But even with proto assembly verification phase there are still mistakes
> that get into production, and it is almost always last minute changes
> done in too much of a hurry.

It rarely happens to us, despite the fact that design complexity and part
counts seem to be increasing exponentially. Although we almost never start
production with Rev A boards, we almost always make changes to the
prototypes.


> Fortunately my contact at the CM
> (meritronics in san jose) is on IM and is a go-getter, can-do kind of
> guy, and doesn't hesitate to ask me about anything odd.

This definitely makes it easier. Our CM often drag their feet and
communication is often a big challenge. But, they're cheap. :)

Vitaliy

2010\06\04@152650 by Mike Hord

picon face
On Fri, Jun 4, 2010 at 1:45 PM, Vitaliy <spam_OUTpiclistTakeThisOuTspammaksimov.org> wrote:

> Jesse Lackey wrote:
> > Hi - all a good idea, the 'proof' is when I (a) order the parts for the
> > protos and (b) assemble the protos (qty several - by hand).  I have a
> > very part time assistant now, but he's really new to all this, and it
> > will be awhile before he's able to do more than very basic checking.
>
> He doesn't need to understand anything to help you check the BOM. During
> crunch times when other engineers were not available, I enlisted the help
> of
> our administrative staff. "Just read me each line." I think it's even
> better
> that they don't understand what the stuff means, since they can't make
> assumptions and "see" what is not there.
>

Huge benefit- like explaining something to your dog when you have a
problem.

As for the rest of the input, it sounds like most of you are in the same
frame of mind I was- I have one part (the assembled board) that I buy
from my CM.  They manage a list of parts that they are allowed to use
(for instance, multiple manufacturers of *3904 transistors) to fill a given
order.

We used to do it in a far more complex manner, involving an individual
part number under ECO control for every component (down to resistors)
on every circuit board (of course, parts may be used on multiple boards).
Having to do an ECO to add a new vendor to the list of approved 1%
0402 1k resistors seems like a REALLY big waste of money to me, and
fortunately we've moved away from that.

Mike H.

2010\06\04@174132 by Vitaliy

face
flavicon
face
Mike Hord wrote:
> As for the rest of the input, it sounds like most of you are in the same
> frame of mind I was- I have one part (the assembled board) that I buy
> from my CM.  They manage a list of parts that they are allowed to use
> (for instance, multiple manufacturers of *3904 transistors) to fill a
> given
> order.

We allow our CM even more freedom: we don't give them a list of parts that
they are allowed to use, we only tell them which parts they must use as
specified. Most of the time it's OK to assume that a jellybean 3904 will
behave like a 3904, and that a generic 1k 0402 will have a resistance in the
950-1050 ohm range.


> We used to do it in a far more complex manner, involving an individual
> part number under ECO control for every component (down to resistors)
> on every circuit board (of course, parts may be used on multiple boards).
> Having to do an ECO to add a new vendor to the list of approved 1%
> 0402 1k resistors seems like a REALLY big waste of money to me, and
> fortunately we've moved away from that.

That does sound like overkill. The notes at the bottom of our BOM read:

- OK to substitute equivalent parts, unless marked with "No" in Sub column.
- Resistors:
   - If power dissipation (W) is not specified, assume it's not critical.
   - If tolerance (%) is not specified, assume 5%. OK to substitute 1%
parts as long as value remains within range.
- Capacitors:
   - If minimum voltage rating (V) is not specified, assume 6V.
   - If tolerance (%) is not specified, assume 20%.

We also get FAs so worst case, if a component does not behave as expected,
we can tell the CM to substitute it before starting production.

Vitaliy

2010\06\05@133804 by Jesse Lackey

flavicon
face
Hi All - these are good, sensible tips easily added to my BOMs.  Thanks.
J


Vitaliy wrote:
{Quote hidden}

2010\06\05@135201 by RussellMc

face picon face
>> As for the rest of the input, it sounds like most of you are in the same
>> frame of mind I was- I have one part (the assembled board) that I buy
>> from my CM.  They manage a list of parts that they are allowed to use
>> (for instance, multiple manufacturers of *3904 transistors) to fill a
>> given order.

> We allow our CM even more freedom: we don't give them a list of parts that
> they are allowed to use, we only tell them which parts they must use as
> specified. Most of the time it's OK to assume that a jellybean 3904 will
> behave like a 3904, and that a generic 1k 0402 will have a resistance in the
> 950-1050 ohm range.

Where is your CM located?
Where / who do they buy from or are allowed to buy from.
I'd be "scared" of doing that in China OR am scared of doing it in
China depending on the factory :-).

Carbon / metal.
Known good vendor eg resistor quality can be very variable.
Known non-supplier of out of spec /  reject parts.
Guaranteed performance PV wafers (not an issue for you).
Guaranteed ROHS parts (discovery only at or after point of foreign
im[port can be vvvv costly)
ESR guaranteed on caps.
Temperature / hours at specd temp on ecaps.
Height of caps (some may vary).
Voltage of ecaps too high (depolarising).
Ceramic cap composition.
LED brands utterly crucial (maybe not as indicators only).
Battery brands.
Bipolar beta.

More.

Any disagreement with any of the above?
If so, may indicate relative trust you can put in market or CM.


         Russell





{Quote hidden}

>

2010\06\07@155535 by Vitaliy

face
flavicon
face
RussellMc wrote, using the "Quoted-Printable" format (tsk, tsk!):
>> As for the rest of the input, it sounds like most of you are in the same
>> frame of mind I was- I have one part (the assembled board) that I buy
>> from my CM. They manage a list of parts that they are allowed to use
>> (for instance, multiple manufacturers of *3904 transistors) to fill a
>> given order.

> We allow our CM even more freedom: we don't give them a list of parts that
> they are allowed to use, we only tell them which parts they must use as
> specified. Most of the time it's OK to assume that a jellybean 3904 will
> behave like a 3904, and that a generic 1k 0402 will have a resistance in
> the
> 950-1050 ohm range.

= Where is your CM located?

Shenzhen, China.


= Where / who do they buy from or are allowed to buy from.

For the most part, we don't know, and don't care. :) As I mentioned, we
specify the critical "do not substitute" components and let them use
jellybean parts for the rest.


= I'd be "scared" of doing that in China OR am scared of doing it in
= China depending on the factory :-).

Our first attempt at outsourcing failed. A year later, we sent them a
custom-built test rig that verified every part of the DUT's operation, and
used the same rig for verification testing on our end. Without the test
procedure in place, I would be scared too.


=> Carbon / metal.
Known good vendor eg resistor quality can be very variable.
Known non-supplier of out of spec /  reject parts.
Guaranteed performance PV wafers (not an issue for you).
Guaranteed ROHS parts (discovery only at or after point of foreign
im[port can be vvvv costly)
ESR guaranteed on caps.
Temperature / hours at specd temp on ecaps.
Height of caps (some may vary).
Voltage of ecaps too high (depolarising).
Ceramic cap composition.
LED brands utterly crucial (maybe not as indicators only).
Battery brands.
Bipolar beta.<=

We run into some of these issues from time to time. However, we always
request an FA for inspection. Also, the agreement makes the CM responsible
for anything that is out of spec. This clause had been invoked a couple of
times already, and they corrected the problems at no charge.


= Any disagreement with any of the above?
= If so, may indicate relative trust you can put in market or CM.

We've been working with the same CM since 2003, and even though we don't
always get a perfect FA on the first attempt, and communication can be
frustrating at times, we've learned to deal with it. The alternative is to
pay x2 per unit, which given our target market and quantities, is
unacceptable.

Customers are happy with the product, and we offer a 3 year warranty (most
competitors offer a 1 year warranty).

Best regards,

Vitaliy Maksimov
ScanTool.net, LLC
+1 623 582-2366
http://www.scantool.net

More... (looser matching)
- Last day of these posts
- In 2010 , 2011 only
- Today
- New search...