Hey there. First-time post, long-time lurker.
Plenty of discussions on the topic this way as well. I calculate defects 3
ways:
1. Total Jobs with at least 1 defect / Total Jobs Processed
2. Total Parts Processed / Total Parts Defect
3. Total Cost of defect / Total PCB Manufacturing Cost
The first two metrics are no-brainers. Out of every job I start, what are
my chances of it getting through unscathed? Out of every PCB I try to make,
what are my chances of getting a winner?
Item 3 is up for debate. For the total cost of the PCB, I take the sale
price and slice of X%, taking a from-the-hip assumption that I make X%
profit. Then for the cost of the defects, I take the price for all PCBs I
tried to make, take of X%, then take off an additional percentage for where
it was identified in the process, as that is when the cost stopped accruing.
Item 3 is critical, as price can be seen as a constant that is reflective of
turn around and technology. I may have few jobs with defects and minimal
defective PCBs, but if all I do have are 2-day 10-layers, the first two
metrics will not accurately communicate this measure of quality.
The arguments against this are that:
A. This seems inaccurate when the customer wants 1 PCB for $100, but we
build 10 and scrap 5, leaving me with a cost of ($100 * 5) * X%.
I don't disagree... But I also see a direct link when this ratio is higher
than normal and finance chasing operations for proportional increase in
costs of goods sold.
B. The PCBs defective would be thrown out anyways, as they were
manufacturing overs to allow for process deviations.
True, but if I didn't have defects, I wouldn't need to run overs, apart from
panel layout considerations. Besides, defects are measured based on what we
tried to do, not what the customer wanted. We tried and we failed, and I
want to know to what extent so as to apply controls to be better next time,
which is the whole point of my gig.
C. The dollar values do not line up with cost of goods sold.
True, but remember that X% is a from-the-hip s.w.a.g. averaged out over all
PCBs. Take away the dollar symbol, throw a decimal somewhere in the number,
treat it as some mathematical parameter and it becomes less offensive to the
bean counters.
The bottom line is that there needs to be some factor to assess scrap
against technology and turn around. Given that this is one part of my job
and only one aspect of the measurement of the quality system, I am not
prepared to obsess on charts displaying defects against the variables that
go into the cost (defects against (turn around | additional processes |
layer count | track space & trace | customer demographic | geographic region
| etc.)). In addition, the folks above me are not in a position to pour
over charts and graphs for this one aspect of the business.
Where this value is used is in attributing focus and resources on specific
areas of the operation. I can have 10 defects from component legend, but
they are only 1 - 5 PCBs each and the total cost constant may be $1000. I
may have 2 defects due to multilayer pressing, but they may have a cost
constant of $5000. This tells me where to apply resources and keeps things
in perspective, as the component legend may get the focus due to the
frequency by the multilayer will have more impact on profit.
As far as percentage values, my experience in the industry is that overruns
are between 5 to 20%, depending on the technology. Small runs seen in the
US and AU throw this out of whack, as running one extra panel can jack it
over 200% overruns if the customer wants two small 8-layer parts. Based on
jobs, I have seen values between 10% to 25%, PCBs can be between 2% to 8%,
and cost ranges from 8% to 15%.
Figured I'd make my post longer than just a hello. Oh... Hello.
Chad Renando
Quality Manager
Precision Circuits Pty Ltd
Ph: +61 3 9877 3222
Mo: 0417 593 337
Fax: +61 3 9877 5079
www.precisioncircuits.com.au
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Brian Guidi
Sent: Thursday, 20 April 2006 12:39 AM
To: [log in to unmask]
Subject: [TN] How are PCB & Assembly manufacturers measuring scrap?
Good morning folks,
We've had some lively discussions regarding the best ways to measure scrap
within a manufacturing operation. What are some common ways of
characterizing and measuring accurate scrap levels?
Also, I know this next one is a loaded question but I'll fire it off anyway.
When scrap is measured as a percentage of sales, what values are typical
within this industry? One percent? Five percent? Is there a significant
difference between the value one would anticipate for a board fab shop
versus an assembly operation?
Brian Guidi
R&D/Quality Systems Specialist
Teledyne Printed Circuit Technology
Tel: (603) 889-6191 X:310
Fax: (603) 886-2977
E-mail: [log in to unmask]
Visit us @ http://www.tetpct.com
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------
|
