Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


SPC and Wave

Any one using SPC to monitor their wave solder process? My ... - Jul 14, 2005 by URL  

How about profiling? ... - Jul 14, 2005 by


URL

#35512

SPC and Wave | 14 July, 2005

Any one using SPC to monitor their wave solder process? My point is collecting defects and using a scatter plot is useful but only after a problem occurs. SPC's true function is to warn of impending problems that are about to occur. Can SPC really be used on wave?

reply »

#35513

SPC and Wave | 14 July, 2005

It can be used as a sales tool. Choose a CPK you want, then "adjust" your definitions and limits untill you achieve that. Just be prepared to explain that someone who is unavailable for explainations set up the SPC system, and you are just following it. But chances are anyone asking won't truely understand SPC anyways.

I wonder if all other industries has as many naked emperors.

Stephen

reply »

#35514

SPC and Wave | 14 July, 2005

Here are links to the web sites for three instruments specifcally designed to monitor wave solder process. Some have SPC logging capabilities:

http://www.ecd.com/emfg/instruments/waverider/

http://www.swpc.co.il/wso.htm

http://www.solderstar.com/waveshuttle_home.html

reply »

slthomas

#35515

SPC and Wave | 14 July, 2005

How about profiling?

reply »

slthomas

#35516

SPC and Wave | 14 July, 2005

I guess I took too much time editing my response. :P

The bottom line is that unless you're measurements are at least as repeatable and accurate as your machine's closed loop systems, it's a waste of paper *BUT* makes nice wall art for tours. I like using profiling (referenced in PeteC's response) for SPC, though.

reply »

dougs

#35517

SPC and Wave | 14 July, 2005

yes SPC can be used on a wave solder machine, you should use a c-chart to monitor the amount of defects per board or group of boards, this will let you know if your process is stable, as you improve your process you should re-calculate your UCL, hopefully to a new lower figure and monitor again. it's ok profiling your wave but that won't tell you about the number of defects the process is seeing.

reply »

#35519

SPC and Wave | 14 July, 2005

Sure you can use SPC charting on the wave setup parameters to warn of impending problems. Examples of useful points of analysis are: * Dwell * Temperature * Pot analysis

Certainly there's a range of other process controls that could be monitored. Others are: * Preheat temperature * Conveyor speed and angle. * Height of the solder wave * Flux density. * Flux height with foam fluxers. * Depth of PWA in wave

[DB: Did you keep any of the money? Or was it lost when you parachuted out?]

reply »

slthomas

#35523

SPC and Wave | 14 July, 2005

Sorry to nitpick, but counting and responding to the number of defects isn't process control.

It's done all the time, but it doesn't really fit the definition and that's what I think DB is trying to investigate.

reply »


RDR

#35535

SPC and Wave | 14 July, 2005

Good answer slthomas! This has got to be the biggest misconception out there. To perform SPC on a wave process you must first perfrom DOE to determine the critical parameters and or control limits. For example you will need temp reading devices that read the board temp just as it enters the wave (Xbar R)OR something that monitors and adjusts your preheat temps (the machine already does that) flux specific gravity will need monitoring (usually an option for machines)at set intervals and charted, wave height must be measured at set intervals, etc...

So as someone else here stated the machine pretty much keeps your process in control as it is. The biggest bang for the buck on a wave is proper setup and verification of such. there are "wave riders" out there that can be used to verify a setup once it has been optimized to be repeatable every time. For example - If you have a board that is now running perfectly you would take this device and let it record your preheat temp, immersion depth, parallelism, dwell time, contact patch, etc... These recorded values need to be documented for that product, and when you run that board again, run the machine through your wave, compare the data and adjust accordingly.

SPC is to PREVENT process variation and is not intended to be a after the fact report that something is/was wrong.

reply »

#35540

SPC and Wave | 14 July, 2005

I guess DB's not commenting on the money.

Jerry

reply »

dougs

#35548

SPC and Wave | 15 July, 2005

thomas,

why isn't it process control, you cant just think that because you've set up your machine properly ( pot temp, wave height, pre-heat etc etc )that everything is going to be ok, there are other factors that you wont find with wave riders or profiling equipment. c-charts and the other attribute control charts are SPC tools, i don't think you should be thinking of SPC as a defect prevention tool, it is what it says, a process control tool.

reply »

#35552

SPC and Wave | 15 July, 2005

And why do you want to control the process? To prevent defects, I would think. Actually orginally the way I first saw SPC was to reduce inspection. If you are using SPC for plating thickness, you don't have to measure every piece to be confident they are all within acceptable parameters.

SPC reminds me of people going to a plastic surgeon and asking for "laser treatment" They don't know what they want it for they just know they want it. And they don't know how exactly it works or what alternatives there are.

I'm talking about management types not the people on here.

reply »

dougs

#35557

SPC and Wave | 15 July, 2005

HHmmmm!! looks like i'm getting it way wrong by wanting to chart the quality of product coming from my wave solder machine. could some of you let me know whats the best way to monitor quality on this process.

reply »

#35558

SPC and Wave | 15 July, 2005

It depends on what you want to accomplish. I believe DB Cooper wants a better product with less wasted effort. If so then Russ and DaveF are the ones to listen to.

If you want to impress managers, customers, and others, then you are doing it right.

I read a little story in reader's digest that shows how a lot of people operate. A woman was cooking her first thanksgiving dinner for her mother and of course wanted to do everything right. She had the turkey inthe sink to defrost. Her mother asked why she had the dish drainer tray over top of it. The daughter answered but you always did it that way. The mother responded "but you don't have a cat".

I find a lot of people do things in a certain way without knowing why.

I'm hoping I'm going to be alittle less cantankerous after we get AC or this heat wave is over. Stephen

reply »

dougs

#35559

SPC and Wave | 15 July, 2005

how do you do it stephen?

i'm interested to know what methods can be used to monitor a process such as this. is it best to set-up the wave machine parameters as russ and davef were talking about, what then, how do we know what effect a change in say pre-heat temp, or any of the machine parameters, will have on product quality.

reply »

pr

#35561

SPC and Wave | 15 July, 2005

As Russ said, you do a DOE beforehand. BUT Maybe we are missing something, we do SPC after the wave (1/5) to catch defect trends and fix them with a minimum of inspection. If you are running a 2,000 piece pull, are you that confident that your original set up will be fine at #1,899? By inspecting 1/5 (and charting UCL) the inspector can get us involved while the problem is happening, this seems to help prevent defects.

PR

reply »


RDR

#35564

SPC and Wave | 15 July, 2005

You are not wrong in monitoring the Product (your really not monitoring the process). This methodology is valuable if your process is unstable (SPC is used to stabilize process). There is value in "real time" inspection of a process's product as long as the feedback is immediate and not published weekly or the next day or whatever. But, if the process is stable and the critical parameters are/have been identified and controlled there should be no change in the product after this process. Use your C charts to determine what your defect level is, then perform a DOE to determine the cause/process window of these defects, then measure the critical parameters with SPC tools/charts and determine if your process is capable of meeting your control limits that were established in the DOE.

reply »

Brian

#35566

SPC and Wave | 15 July, 2005

A process consists of much more than just thee machine. When developing a cause and effect diagram ffor the process, there are many parameters that fall outside the machine itself. There are Man, Machine, Methods, and Materials. Just monitoring the machine and getting the parameters set up correctly for thaat board does not account for variations in the other categories. Case in point: You have ddone all your set-ups of the machine correctly, including nice DOE's. As you run the product, all the machine parameters show no variation from your set points. The you start getting solder defects. InN our case it was pinholes. Because we monitored prouct coming out of the wave as well as the parameters, we were able to shut the line down before building a lot of defects and determine it was a problem with the bare boards (and no, baking didn't solve the problem). The manufacturer had poor control over their processes and had a unsolderaable material get on some connector pads. In my experience, a good SPC program monitors both; process parameters AND defect levels. SPC can and should be done on both attribute and variable data. I have used ppm charts (with control limits) and pareto charts. The result has been wave sodler and SMT processes that have conssistently run under 50ppm.

Brian

reply »


URL

#35609

SPC and Wave | 18 July, 2005

Thanks to all for the input. It has helped change our standard SPC program around.

(To davef: Dad handled most of it, but Vegas got the rest)

reply »

#35610

SPC and Wave | 18 July, 2005

Hello,

According with quality standards TS16949 ,QS 9000 etc. improvement can be perform when the process is stable and capable. So, if DOE is method for improvement it is a good idea, first to run SPC for investigation process stability and capability. SPC (by variables and attributes) also is monitoring and preventive tool for quality. But before do any actions ask yourself: Why?, What of that?

Good luck!

reply »

slthomas

#35613

SPC and Wave | 18 July, 2005

"thomas,

why isn't it process control,"

Minute detail, really. Defects aren't a measurement of process parameters, they're the result of them (and a few other things that have been mentioned). SPC measures process parameters. It is preemptive, and it's goal is to eliminate defects resulting from process fluctuation.

Executing countermeasures based on defect rates is reactive even if you're using control limits.

That's not to say that tracking defects is a bad thing...we do it too. It's a quality tool that has it's place, but it's not SPC.

reply »


Rob

#35618

SPC and Wave | 19 July, 2005

Solder Paste Inspection

Lead Free Wave Solder - 1 Click SMT