I wonder

I wonder to knock this site with this question? Are there some body have applied 6 SIGMA TOOL to improve the smt process or etc. especilly smt process is better. I want to hear the difficulties to apply this tool. and experiences. I dont't care good or bad experience. i think it is good theory. i want to hear true experience you did from you guys. sorry for silly question

thank in advance.

reply »

| I wonder to knock this site with this question? | Are there some body have applied 6 SIGMA TOOL to improve | the smt process or etc. especilly smt process is better. | I want to hear the difficulties to apply this tool. | and experiences. I dont't care good or bad experience. | i think it is good theory. i want to hear true experience you did from you guys. | | sorry for silly question | | thank in advance. | | | Kyung: It's not a silly question, let's see how many silly responses you get.

For many processes the cost of obtaining 6 sig quality could be very expensive and with the rate that products turn the effort may not pay-off. So, setting a very difficult target may be inspiration to reach 4 or may be 5 sig levels.

Very few machines have process capability of over �3 sig.

Good luck

Dave F

reply »

| | I wonder to knock this site with this question? | | Are there some body have applied 6 SIGMA TOOL to improve | | the smt process or etc. especilly smt process is better. | | I want to hear the difficulties to apply this tool. | | and experiences. I dont't care good or bad experience. | | i think it is good theory. i want to hear true experience you did from you guys. | | | | sorry for silly question | | | | thank in advance. | | | | | | | Kyung: It's not a silly question, let's see how many silly responses you get. | | For many processes the cost of obtaining 6 sig quality could be very expensive and with the rate that products turn the effort may not pay-off. So, setting a very difficult target may be inspiration to reach 4 or may be 5 sig levels. | | Very few machines have process capability of over �3 sig. | | Good luck | | Dave F | I have used 6 Sigma tools to improve the process and product quality. The 6 Sigma tools are very useful in that they give one the ability to identify all the variables in a process, define them, and then make real changes to improve our processes. Too ften, change is made by someoen saying: "I thik this is the cause...". Six Sigma requires facts, not guesses. As the old saying goes: "In God we trust, all others bring data". Many of the tools used are the everyday tools we have been using for a long time(for example, Cause and Effect diagrams). By defining the processes and identifying the factors, we can make the process/procedural/systemic changes in all phases of business operations to make real change. As more and more variation is reduced, it usually becomes more costly to fix issues. I like the thought process behind 6 Sigma, and use it all the time.

Brian

reply »

| I wonder to knock this site with this question? | Are there some body have applied 6 SIGMA TOOL to improve | the smt process or etc. especilly smt process is better. | I want to hear the difficulties to apply this tool. | and experiences. I dont't care good or bad experience. | i think it is good theory. i want to hear true experience you did from you guys. | | sorry for silly question We have great success using Design of Experiments(DOE)to tune our SMT and wave solder processes. In most cases a simple screening experiment will bring a process with a Cpk 2.0. If your time and resources are limited then I recommend starting with the printing process. Getting that process under control with a Cpk > 2 will have enormous impact on every other process. | thank in advance. | | |

reply »

| | I wonder to knock this site with this question? | | Are there some body have applied 6 SIGMA TOOL to improve | | the smt process or etc. especilly smt process is better. | | I want to hear the difficulties to apply this tool. | | and experiences. I dont't care good or bad experience. | | i think it is good theory. i want to hear true experience you did from you guys. | | | | sorry for silly question | We have great success using Design of Experiments(DOE)to tune our SMT and wave solder processes. In most cases a simple screening experiment will bring a process with a Cpk 2.0. If your time and resources are limited then I recommend starting with the printing process. Getting that process under control with a Cpk > 2 will have enormous impact on every other process. | | thank in advance.

scott. I appreciate your sincere answer. would tell me the hint about printer if i's possible To tune our smt process we have to know what is the major X's for defect. To know major X's with DOE we need a good measurement equipment. for example we want to know what is the major x's for printer. but if we cannot measure the variances when the paramter is varied. DOE is useless tool. In smt process i think that It's also smt technology to make stuff is invisible visible. I tried to do DOE before but I dropped. We have got a LSM2(CYBER OPTICS)to check solder volume. Resolution of LSM2 isn't enough to detect the variance of solder volume on pad associated with parameter. do i have wrong concept in process.

we have got equipment like below -wetting tester -micro section m/c -viscosity tester -magnifier -LSM2 (To check aperture,land size,soler volume,thickness of silk .etc) -vc( visual inspect machine) -x- ray inspecter what kind of equipments do you have not direct m/c but indirect m/c for production. i do not mean your company just know about general U.S.A companies. thank you

reply »

| | | I wonder to knock this site with this question? | | | Are there some body have applied 6 SIGMA TOOL to improve | | | the smt process or etc. especilly smt process is better. | | | I want to hear the difficulties to apply this tool. | | | and experiences. I dont't care good or bad experience. | | | i think it is good theory. i want to hear true experience you did from you guys. | | | | | | sorry for silly question | | We have great success using Design of Experiments(DOE)to tune our SMT and wave solder processes. In most cases a simple screening experiment will bring a process with a Cpk 2.0. If your time and resources are limited then I recommend starting with the printing process. Getting that process under control with a Cpk > 2 will have enormous impact on every other process. | | | thank in advance. | | | scott. | I appreciate your sincere answer. | would tell me the hint about printer if i's possible | To tune our smt process we have to know what is the major X's | for defect. To know major X's with DOE we need a good measurement equipment. for example we want to know what is the | major x's for printer. but if we cannot measure the variances | when the paramter is varied. DOE is useless tool. | In smt process i think that It's also smt technology to make stuff is invisible visible. | I tried to do DOE before but I dropped. | We have got a LSM2(CYBER OPTICS)to check solder volume. Resolution of LSM2 isn't enough to detect the variance of solder volume on pad associated with parameter. do i have wrong concept in process. | | we have got equipment like below | -wetting tester | -micro section m/c | -viscosity tester | -magnifier | -LSM2 (To check aperture,land size,soler volume,thickness of silk .etc) | -vc( visual inspect machine) | -x- ray inspecter | what kind of equipments do you have not direct m/c but indirect m/c for production. | i do not mean your company just know about general U.S.A companies. | | | thank you | | | Kyung: You have some cool toys.

Just for laughs, do a guage R&R study on your LSM2. It will help explain things to you about that device. It is very difficult to use.

My2�

Dave F

reply »

| scott. | I appreciate your sincere answer. | would tell me the hint about printer if i's possible | To tune our smt process we have to know what is the major X's | for defect. To know major X's with DOE we need a good measurement equipment. for example we want to know what is the | major x's for printer. but if we cannot measure the variances | when the paramter is varied. DOE is useless tool. | In smt process i think that It's also smt technology to make stuff is invisible visible. | I tried to do DOE before but I dropped. | We have got a LSM2(CYBER OPTICS)to check solder volume. Resolution of LSM2 isn't enough to detect the variance of solder volume on pad associated with parameter. do i have wrong concept in process. | | we have got equipment like below | -wetting tester | -micro section m/c | -viscosity tester | -magnifier | -LSM2 (To check aperture,land size,soler volume,thickness of silk .etc) | -vc( visual inspect machine) | -x- ray inspecter | what kind of equipments do you have not direct m/c but indirect m/c for production. | i do not mean your company just know about general U.S.A companies. | | | thank you In a previous job I had the luxury of the following process: 1. Purchase any equipment regardless of cost 2. Set it in a lab for a year with a dedicated operator 3. Buy any test equipment you want 4. Create a Process specification backed with data that can never be change without a DOE to support a change. My primary responsibility was screen printing, paste, etc. I did this job for 7 years ( I left for a real world job 2.5 years ago). During that time many DOE's were run on printers. Here are a few pointers. We purchased a couple of tools from CyberOptics - Laser Section Microscope and a automated tool for measuring volume (I forget the model #). In a lab environment both measurement tools had a Gauge r&r of less than 15%. I believe both tools were more than adequete to get the resolution you need to define the printer process. I also recommend that you never use the measurement tool provided by the printer manufacturer for serious experiments. Many test were consistent in showing that the following variables were the most important: 1: Downstop 2: SnapOff 3: Print Speed 4: Squeegee Blade Type (metal vs diamond) 5. Plating thickness vs. soldermask thickness (tricky) Depending on the printer type other variables came into play. One critical measurement decision is deciding what to use as the reference plane. Our tools had trouble with soldermask so we used the land for the reference plane for measuring. This is usually OK because the apertures are reduced and the land is the highest point on the board so the stencil is resting on the land not the mask. We did have some boards with vias tented with dry film and then covered with wet film. Subsequently the vias were higher than the pads causing all kinds of failures. Once you decide how you are to measure you must figure out what the specification. After measuring both volume and solder paste height we ended up using solder paste height because it was easier to test in production than volume (and faster).We always used Target +/- .002". This number was backed up with extensive life testing using thermocycling and random vibration stress testing. So I still use this specification and have never had a problem. With todays equipment this is a huge range. A long answer to a simple question. I would use the LSM with measuring solder paste height. Use the pad as a reference. If you are having trouble measuring the paste height on fine pitch parts only measure the 50 mil pitch parts (the brick should be much better defined). Run a screening experiment with at least 8 variables. Forget about the paste viscosity (I used both Brookfield, Malcolm - the fancy one and handheld, that is another long discussion), instead require the supplier to send a certification on the past viscosity (Malcolm preferred). To measure end results I recommend cross sections before and after thermal cycling. Sorry for the ramble I got started and couldn't stop.

reply »

thank you very much Mr. Scott.S Snider Your experience gotta be big helps to me.

reply »