Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
Wavesoldering capability study

Wavesoldering capability study

Subscribe to Thread

Im not sure if someone has posted this topic before. Anyway,... - Jul 05, 2007 by Reypal

The variables to study: * Dwell Time * Contact Length ... - Jul 05, 2007 by Dirk Nuendyke

WRONG! Actually I was going to say the same thing but fig... - Jul 05, 2007 by Real Chunks

i defagree ... - Jul 05, 2007 by Peter Gibbons

Here are the basics of determining that your process is in c... - Jul 05, 2007 by davef

Thank You Guys for all your feedback, comments and linkage. ... - Jul 09, 2007 by Reypal

DB Comments on things that should be set and controlled and... - Jul 09, 2007 by davef

Parallelism tends to drift: �h After preventative mainten... - Jul 10, 2007 by Samir Nagheenanajar

Reypal

#50918

Wavesoldering capability study | 5 July, 2007

Im not sure if someone has posted this topic before. Anyway,we plan to make capability study on our wave-soldering machine but we have no experience on how to do it effectively. My question is what we should consider most among the critical parameters? or does someone had experienced and would like to share in here. Thank you in advance for all your replies. / Reypal

reply »

Dirk Nuendyke

#50920

Wavesoldering capability study | 5 July, 2007

The variables to study:

* Dwell Time * Contact Length * Conveyor Speed * Parallelism * Top-side Substrate Temperature Before wave * Solder Pot temperature

All of these variables are inter-related (ie changing one variable will affect another), and you'll need some type of method to measure them.

Search the fine archives - Keywords: wave solder process capability, wave solder SPC, waverider, wave optimizer, malcom dip tester, glass plate

There's been much discussion on this topic.

reply »

Real Chunks

#50921

Wavesoldering capability study | 5 July, 2007

WRONG!

Actually I was going to say the same thing but figured I would have the standard 3 people dissagreeing with me, and or correcting me. I do believe your wave solder temp should remain a constant temp and prolly will not vary at all, so I wouldn't bother doing a Cpk study on it. Also amount of flux is also a variable you may want to measure.

reply »

Peter Gibbons

#50924

Wavesoldering capability study | 5 July, 2007

i defagree

reply »

davef

#50936

Wavesoldering capability study | 5 July, 2007

Here are the basics of determining that your process is in control and capable - http://lorien.ncl.ac.uk/ming/spc/spc9.htm

There are tons of stuff like this on the web.

reply »

Reypal

#50955

Wavesoldering capability study | 9 July, 2007

Thank You Guys for all your feedback, comments and linkage.

reply »

davef

#50957

Wavesoldering capability study | 9 July, 2007

DB Comments on things that should be set and controlled and not adjusted are: * We agree that pot temperature should be excluded. * We'd like you to add parallelism to your list of items to exclude from CK's list. It's given that parallelism is important, but once set, we wouldn't expect it to be changed through the operation of the machine.

reply »

Samir Nagheenanajar

#50973

Wavesoldering capability study | 10 July, 2007

Parallelism tends to drift:

�h After preventative maintenance �h From dross buildup or other obstruction clogging either wave

Parallelism should at least be studied during the process characterization stage, so that you know your machine's process capability with regards to parallelism, but not necessarily tracked as part of your process control or SPC program.

For example, at one of my client's factories, there are 3 wave solder machines - 1 of them is 0.5 seconds "crooked", another is 1 second crooked, and the 3rd is 1 second crooked. When troubleshooting things like solder skips, top-side wetting, etc., it helps to know these things. Now, my client can draw conclusions and adjust the machine and process accordingly, from this data.

The need to measure parallelism also depends on the product mix being run. If you have boards with bottom side 0603's and SOT23's glued on, and they're at the very edge of your assembly, then parallelism is pretty critical.

Samir

reply »