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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


SPC on stencil printer

Sanjay Bhatikar

#12997

SPC on stencil printer | 12 January, 1999

For process management with SPC of the stencil printing process with a fully automatic stencil printer, what process outpur parameter would you recommend for SPC?

It appears to me that there are two options: A: to have a pn chart based upon the % of boards rejected by visual post-stencil inspection. B: to measure the print-height variation by a laser-scanning device (takes less than a minute for a 12in. x 18in. board).

A is inexpensive, but subjective - visual inspection (with optical aids or automated) sounds like a rather tenuous proposition. B sounds more reasonable, but implies additional equipment of an advanced nature.

What is your opinion? I am sure there are other ways of doing post-stencil printing SPC that I have not thought of yet! :) Please feel free to highlight my ignorance.

Good day, all!

reply »

Jason Hall

#12998

Re: SPC on stencil printer | 12 January, 1999

| For process management with SPC of the stencil printing process with a fully automatic stencil printer, what process outpur parameter would you recommend for SPC? | | It appears to me that there are two options: | A: to have a pn chart based upon the % of boards rejected by visual post-stencil inspection. | B: to measure the print-height variation by a laser-scanning device (takes less than a minute for a 12in. x 18in. board). |

A= Depending on your equipment, it might already have the capabilities to complete the task at hand. The company I work for uses MPM, and even though it is capable of doing this we opt to use a cyber sentry machine to do paste measurements. This is due to the ease of operation, both from a programmers stand point and an operators standpoint. The cost difference is relatively small and well worth it when dealing with BGA and Fine-Pitch placement. I am not sure if Fuji GSP's have this capability.

reply »

Earl Moon

#12999

Re: SPC on stencil printer | 12 January, 1999

| | For process management with SPC of the stencil printing process with a fully automatic stencil printer, what process outpur parameter would you recommend for SPC? | | | | It appears to me that there are two options: | | A: to have a pn chart based upon the % of boards rejected by visual post-stencil inspection. | | B: to measure the print-height variation by a laser-scanning device (takes less than a minute for a 12in. x 18in. board). | | | | | A= Depending on your equipment, it might already have the capabilities to complete the task at hand. The company I work for uses MPM, and even though it is capable of doing this we opt to use a cyber sentry machine to do paste measurements. This is due to the ease of operation, both from a programmers stand point and an operators standpoint. The cost difference is relatively small and well worth it when dealing with BGA and Fine-Pitch placement. I am not sure if Fuji GSP's have this capability. | | Sanjay,

You said it all. You're no rookie. The print's the thing - provided it contributes to effecting and assuring acceptable solder joints.

What's that neat unit Steve Gregory talks about? You know, the cheap scope for around $1500 that measures print height through focal adjustment.

Both DEK and MPM have 2D inspection capabilities. MPM has a 3D option that slows things down a bit too much to the point it isn't always useful. Nikon has a pretty nice unit for a lot of money. So do others but, however it's done and all that goes into it, the print's the thing!

Earl Moon

reply »

Jon Medernach

#13000

Re: SPC on stencil printer/ what's important? | 12 January, 1999

What's important in stencil printing (which differs from screen printing) is providing a consistant volume of paste at each interconnect. The process window varies in proportion to the pitch. More than 60% of all defects are related to solder deposition. Examine the defects on a given board (shorts, opens, solder balls) find areas that appears to be problematic. Do an analysis of the printing in that area, ie print and measure results in this are on 5 passes, now clean and make 5 more passes, compare print before and after a clean cycle. Determine the machines ability to provide a consistant volume and adjust appeture geometry to optimize cleaning requirement. Key is to maintain process parameters, Temp, Humidity, pressure, speed,paste parameters. (example; if you remove all the paste from the stencil after 4 hours and put new material you have completely changed the chemistry of the paste you set up to print.

| For process management with SPC of the stencil printing process with a fully automatic stencil printer, what process outpur parameter would you recommend for SPC? | | It appears to me that there are two options: | A: to have a pn chart based upon the % of boards rejected by visual post-stencil inspection. | B: to measure the print-height variation by a laser-scanning device (takes less than a minute for a 12in. x 18in. board). | | A is inexpensive, but subjective - visual inspection (with optical aids or automated) sounds like a rather tenuous proposition. B sounds more reasonable, but implies additional equipment of an advanced nature. | | What is your opinion? I am sure there are other ways of doing post-stencil printing SPC that I have not thought of yet! :) Please feel free to highlight my ignorance. | | Good day, all! |

reply »

Steve Gregory

#13001

Re: SPC on stencil printer | 13 January, 1999

Hello Sanjay!

The cheap deal I use that Earl spoke of is called a SMarT Measure. It's a depth measuring microscope that uses the same principle of measuring optical focus that the units from Vision Engineering do ( if you've ever looked at any of their units). Their specification for accuracy is the same as it is for the laser units +/- .0005.

As far as SPC goes, I'm gonna go out on a limb a bit and say that many times at past employers, the SPC program that was implemented was more of a show for customers, than a tool used to control processes.

You see that at many companies. One thing you can do to see that for yourself, is to look at the charts that some companies have posted all over the place. Many times when you look at the charts you'll see the lines cruising right along down the middle...very rarely do you see them get close to the upper or lower control limits. Is that really reflecting the process? Or is that what they want outsiders to see? If I see an out of control situation on a SPC chart, and then see the line come back within limits, I'll put more stock in the data for that chart than I would one that has the line straight up the middle...

Speaking from experience, I've seen that kind of stuff go on many times before. The day before a survey by a potential customer, there's this flurry of activity. Reposting charts, pencil whipping PM dates, etc., getting ready for "the show"...and that's sad. It also explains why they're my PAST employers. I don't play that game... paperwork and documenting things is a chore that I despise, so if I'm going invest any of my time and effort into it, it better had be for a good reason and be worthwhile accomplishing something other than being a part of "a show".

Which leads me into why I don't do SPC at printing...I don't need it. I found that once I get the printer set-up, and after the first print is measured and verified good, all I'll be doing the rest of the time I'm printing as far as SPC goes, is measuring good prints...the print quality just doesn't go bad unless something changes.

When you do change something, such as maybe adding fresh paste, or maybe changing squeegee speed or something like that, it's only common sense to look and inspect the next several prints after the changes to make sure everything is okay. But then after that, I don't think it's necessary. What's going to change? Alignment variances? If the printer starts wandering while I'm printing, then I'll tell you what, you'd see me on the SMTNET bad-mouthing the manufacturer like there's no tomorrow! Bottom line, it just doesn't happen. Paste Height variances? I suppose that could happen with plastic squeegees if used with a bad set-up. But I use metal ones and don't even think about it now. There's not too much else that can go wrong at printing other than that is there? I suppose you could run out of paste on the stencil, but I would be a little embarrassed to say that I need SPC to make sure I keep paste on the stencil. The stencil could get dirty and need cleaning, but I program a wipe every 5-prints whether it really needs one or not and that'll usually take care of any fuzzy prints that might happen.

What I use where I work now to control my printing process is ALLEN. What ALLEN is, or I should say who Allen is, he's my right-hand SMT man, it's just me and him. Allen came to work with us when we opened the doors here back in July of last year. He's had no experience with the equipment we have. He's worked at Solectron, Avex, companies like that who uses the high-end stuff like MPM's UP3K, FUJI CP6's, etc., and if you know any of those companies, you know they have the kind of deep pockets and staff to support an extensive SPC program. Allen knows how to take SPC measurements (they did teach him that), but what they didn't teach him was the fundamentals of printing...he was a robot at places like that.

Do this first, then do that, and when that green light comes on push this button...in other words; the kind of operators I call "button pushers". Not being paid to think, just push buttons. Well I don't want operators like that. I want people who'll think about what they're doing. I want them to use their minds and not be some pavlovian dog out there that salivates when they see a light come on. The job is more stimulating anyway when you get to exersize the ol' gray matter.

So I spent some time going over what I know about printing with Allen. All it cost me was just to take the time to explain what I was doing whenever I did anything with the printer, or stencils. I explained how the gerber data becomes a stencil, downloaded Lavenir Viewmate for him and put it on his 'pooter, and showed him how he can view gerbers too, how I went about determining what thickness to use, and even took him over to our stencil vendor just before lunch one day so he could see how stencils are made. Of course, I also gave him the operator training for our printer, a high-end MPM SPM-AV...it costs less than some of the options on the high-end printers, but we've done micro-BGA and 15-mil on it here so far.

So I'll get off my soapbox now, and close by saying that a little knowledge goes a hell of a long way! What I invested in Allen dollar-wise, was next to nothing, all it cost me was a little bit of my time, a lunch one day, and a desire to teach him something. I'd rather have my printing process controlled by Allen, than any multi-thousand dollar piece of inspection equipment, or a whole army of process auditors...just my two cents. (actually I think that was more like a $1.99...GRIN)

-Steve Gregory-

reply »

Mike McMonagle

#13002

Re: SPC on stencil printer | 14 January, 1999

Steve, Whoa, take a breath! That was at least a dollar three-eighty answer. I've got to agree that nothing can replace a qualified, concientious operator for assuring quality. When I too worked at a small shop there in SillyCone Valley, my senior operator was as close as you could get to being an engineer without going from hourly to salary (the lucky dog).

I've also too often seen the frenzy of 'dry labbing' charts and logs while the customer sits in the lobby wondering what the tour delay is. If that sort of effort was put forth in smaller increments on a regular basis, the charts and logs would actually be legit.

Back to the SPC issue, though. Many customers (with engineers justifying their existence through making your life difficult), want to see some rudimentary form of data collection on the printing process. Back then we also used the same differential focusing height gage that you have, along with a simple log that we recorded four locations and average height when we did first article off the printer. I have since gotten an Excel spreadsheet that does the same thing, but does the calculations and automatically sets and charts the UCL & LCL based on predetermined subgroups established per QS9000 guidelines. So we have ease of entry, real time data collection on an hourly basis and pretty pictures for that pesky customer engineer. Now I'll shut up before I'm up to a dollar five-ninety seven answer....

| Hello Sanjay! | | The cheap deal I use that Earl spoke of is called a SMarT Measure. It's a depth measuring microscope that uses the same principle of measuring optical focus that the units from Vision Engineering do ( if you've ever looked at any of their units). Their specification for accuracy is the same as it is for the laser units +/- .0005. | | As far as SPC goes, I'm gonna go out on a limb a bit and say that many times at past employers, the SPC program that was implemented was more of a show for customers, than a tool used to control processes. | | You see that at many companies. One thing you can do to see that for yourself, is to look at the charts that some companies have posted all over the place. Many times when you look at the charts you'll see the lines cruising right along down the middle...very rarely do you see them get close to the upper or lower control limits. Is that really reflecting the process? Or is that what they want outsiders to see? If I see an out of control situation on a SPC chart, and then see the line come back within limits, I'll put more stock in the data for that chart than I would one that has the line straight up the middle... | | Speaking from experience, I've seen that kind of stuff go on many times before. The day before a survey by a potential customer, there's this flurry of activity. Reposting charts, pencil whipping PM dates, etc., getting ready for "the show"...and that's sad. It also explains why they're my PAST employers. I don't play that game... paperwork and documenting things is a chore that I despise, so if I'm going invest any of my time and effort into it, it better had be for a good reason and be worthwhile accomplishing something other than being a part of "a show". | | Which leads me into why I don't do SPC at printing...I don't need it. I found that once I get the printer set-up, and after the first print is measured and verified good, all I'll be doing the rest of the time I'm printing as far as SPC goes, is measuring good prints...the print quality just doesn't go bad unless something changes. | | When you do change something, such as maybe adding fresh paste, or maybe changing squeegee speed or something like that, it's only common sense to look and inspect the next several prints after the changes to make sure everything is okay. But then after that, I don't think it's necessary. What's going to change? Alignment variances? If the printer starts wandering while I'm printing, then I'll tell you what, you'd see me on the SMTNET bad-mouthing the manufacturer like there's no tomorrow! Bottom line, it just doesn't happen. Paste Height variances? I suppose that could happen with plastic squeegees if used with a bad set-up. But I use metal ones and don't even think about it now. There's not too much else that can go wrong at printing other than that is there? I suppose you could run out of paste on the stencil, but I would be a little embarrassed to say that I need SPC to make sure I keep paste on the stencil. The stencil could get dirty and need cleaning, but I program a wipe every 5-prints whether it really needs one or not and that'll usually take care of any fuzzy prints that might happen. | | What I use where I work now to control my printing process is ALLEN. What ALLEN is, or I should say who Allen is, he's my right-hand SMT man, it's just me and him. Allen came to work with us when we opened the doors here back in July of last year. He's had no experience with the equipment we have. He's worked at Solectron, Avex, companies like that who uses the high-end stuff like MPM's UP3K, FUJI CP6's, etc., and if you know any of those companies, you know they have the kind of deep pockets and staff to support an extensive SPC program. Allen knows how to take SPC measurements (they did teach him that), but what they didn't teach him was the fundamentals of printing...he was a robot at places like that. | | Do this first, then do that, and when that green light comes on push this button...in other words; the kind of operators I call "button pushers". Not being paid to think, just push buttons. Well I don't want operators like that. I want people who'll think about what they're doing. I want them to use their minds and not be some pavlovian dog out there that salivates when they see a light come on. The job is more stimulating anyway when you get to exersize the ol' gray matter. | | So I spent some time going over what I know about printing with Allen. All it cost me was just to take the time to explain what I was doing whenever I did anything with the printer, or stencils. I explained how the gerber data becomes a stencil, downloaded Lavenir Viewmate for him and put it on his 'pooter, and showed him how he can view gerbers too, how I went about determining what thickness to use, and even took him over to our stencil vendor just before lunch one day so he could see how stencils are made. Of course, I also gave him the operator training for our printer, a high-end MPM SPM-AV...it costs less than some of the options on the high-end printers, but we've done micro-BGA and 15-mil on it here so far. | | So I'll get off my soapbox now, and close by saying that a little knowledge goes a hell of a long way! What I invested in Allen dollar-wise, was next to nothing, all it cost me was a little bit of my time, a lunch one day, and a desire to teach him something. I'd rather have my printing process controlled by Allen, than any multi-thousand dollar piece of inspection equipment, or a whole army of process auditors...just my two cents. (actually I think that was more like a $1.99...GRIN) | | -Steve Gregory- | |

reply »

Earl Moon

#13003

Re: SPC on stencil printer | 14 January, 1999

| Steve, | Whoa, take a breath! That was at least a dollar three-eighty answer. I've got to agree that nothing can replace a qualified, concientious operator for assuring quality. When I too worked at a small shop there in SillyCone Valley, my senior operator was as close as you could get to being an engineer without going from hourly to salary (the lucky dog). | | I've also too often seen the frenzy of 'dry labbing' charts and logs while the customer sits in the lobby wondering what the tour delay is. If that sort of effort was put forth in smaller increments on a regular basis, the charts and logs would actually be legit. | | Back to the SPC issue, though. Many customers (with engineers justifying their existence through making your life difficult), want to see some rudimentary form of data collection on the printing process. Back then we also used the same differential focusing height gage that you have, along with a simple log that we recorded four locations and average height when we did first article off the printer. I have since gotten an Excel spreadsheet that does the same thing, but does the calculations and automatically sets and charts the UCL & LCL based on predetermined subgroups established per QS9000 guidelines. So we have ease of entry, real time data collection on an hourly basis and pretty pictures for that pesky customer engineer. Now I'll shut up before I'm up to a dollar five-ninety seven answer.... | | | Hello Sanjay! | | | | The cheap deal I use that Earl spoke of is called a SMarT Measure. It's a depth measuring microscope that uses the same principle of measuring optical focus that the units from Vision Engineering do ( if you've ever looked at any of their units). Their specification for accuracy is the same as it is for the laser units +/- .0005. | | | | As far as SPC goes, I'm gonna go out on a limb a bit and say that many times at past employers, the SPC program that was implemented was more of a show for customers, than a tool used to control processes. | | | | You see that at many companies. One thing you can do to see that for yourself, is to look at the charts that some companies have posted all over the place. Many times when you look at the charts you'll see the lines cruising right along down the middle...very rarely do you see them get close to the upper or lower control limits. Is that really reflecting the process? Or is that what they want outsiders to see? If I see an out of control situation on a SPC chart, and then see the line come back within limits, I'll put more stock in the data for that chart than I would one that has the line straight up the middle... | | | | Speaking from experience, I've seen that kind of stuff go on many times before. The day before a survey by a potential customer, there's this flurry of activity. Reposting charts, pencil whipping PM dates, etc., getting ready for "the show"...and that's sad. It also explains why they're my PAST employers. I don't play that game... paperwork and documenting things is a chore that I despise, so if I'm going invest any of my time and effort into it, it better had be for a good reason and be worthwhile accomplishing something other than being a part of "a show". | | | | Which leads me into why I don't do SPC at printing...I don't need it. I found that once I get the printer set-up, and after the first print is measured and verified good, all I'll be doing the rest of the time I'm printing as far as SPC goes, is measuring good prints...the print quality just doesn't go bad unless something changes. | | | | When you do change something, such as maybe adding fresh paste, or maybe changing squeegee speed or something like that, it's only common sense to look and inspect the next several prints after the changes to make sure everything is okay. But then after that, I don't think it's necessary. What's going to change? Alignment variances? If the printer starts wandering while I'm printing, then I'll tell you what, you'd see me on the SMTNET bad-mouthing the manufacturer like there's no tomorrow! Bottom line, it just doesn't happen. Paste Height variances? I suppose that could happen with plastic squeegees if used with a bad set-up. But I use metal ones and don't even think about it now. There's not too much else that can go wrong at printing other than that is there? I suppose you could run out of paste on the stencil, but I would be a little embarrassed to say that I need SPC to make sure I keep paste on the stencil. The stencil could get dirty and need cleaning, but I program a wipe every 5-prints whether it really needs one or not and that'll usually take care of any fuzzy prints that might happen. | | | | What I use where I work now to control my printing process is ALLEN. What ALLEN is, or I should say who Allen is, he's my right-hand SMT man, it's just me and him. Allen came to work with us when we opened the doors here back in July of last year. He's had no experience with the equipment we have. He's worked at Solectron, Avex, companies like that who uses the high-end stuff like MPM's UP3K, FUJI CP6's, etc., and if you know any of those companies, you know they have the kind of deep pockets and staff to support an extensive SPC program. Allen knows how to take SPC measurements (they did teach him that), but what they didn't teach him was the fundamentals of printing...he was a robot at places like that. | | | | Do this first, then do that, and when that green light comes on push this button...in other words; the kind of operators I call "button pushers". Not being paid to think, just push buttons. Well I don't want operators like that. I want people who'll think about what they're doing. I want them to use their minds and not be some pavlovian dog out there that salivates when they see a light come on. The job is more stimulating anyway when you get to exersize the ol' gray matter. | | | | So I spent some time going over what I know about printing with Allen. All it cost me was just to take the time to explain what I was doing whenever I did anything with the printer, or stencils. I explained how the gerber data becomes a stencil, downloaded Lavenir Viewmate for him and put it on his 'pooter, and showed him how he can view gerbers too, how I went about determining what thickness to use, and even took him over to our stencil vendor just before lunch one day so he could see how stencils are made. Of course, I also gave him the operator training for our printer, a high-end MPM SPM-AV...it costs less than some of the options on the high-end printers, but we've done micro-BGA and 15-mil on it here so far. | | | | So I'll get off my soapbox now, and close by saying that a little knowledge goes a hell of a long way! What I invested in Allen dollar-wise, was next to nothing, all it cost me was a little bit of my time, a lunch one day, and a desire to teach him something. I'd rather have my printing process controlled by Allen, than any multi-thousand dollar piece of inspection equipment, or a whole army of process auditors...just my two cents. (actually I think that was more like a $1.99...GRIN) | | | | -Steve Gregory- | | | | | | Hell, I can't catch my breath - what there is left of it after changing altitudes and climate.

I hear you Steve. I just use the basics for real time feed back as does Mike. I too use a simple spreadsheet approach and chart variables to assist in immediate corrective action, but more for long term process improvement as in the case of stencils fulfilling their life cycles or whatever.

Sure customers want to see process control, effective process management, and output results before poor quality bites them in the ass. SPC can be an effective tool to do that if properly done - for the right reasons.

The real reason to use SPC is to ensure CPI at the design, fabrication, and assembly levels. Management also reacts kinder when justification is needed for new equipment, tools, facilities, or whatever.

As you all know, after serving time there, the big companies go fast and never stop production. There isn't always a faithful sidekick to work with closely. That's why ISO, CPI, and TQM are applied reasonably well. I mean, each person within an effective quality system is trained and certified to accept individual responsibilities within a "team" atmosphere. This requires real time feedback, beyond constant visual inspection, to make corrections and continuously improve processes for the long run. SPC is but one tool to do that.

Earl Moon

reply »

ICT Total SMT line Provider

SMT spare parts - Qinyi Electronics