What rate is considered world class when it comes to pick performance for SMT placement machines? Can you define pick performance? Board Talk
Board Talk is presented by Phil Zarrow and Jim Hall of ITM Consulting.
Process Troubleshooting, Failure Analysis, Process Audits, Process Set-up CEM Selection/Qualification, SMT Training/Seminars, Legal Disputes
Phil Zarrow
With over 50 years experience in PCB assembly, Phil is one of the leading experts in SMT process failure analysis.
He has vast experience in SMT equipment, materials and processes.
Jim Hall
A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics.
He is a pioneer in the science of reflow.
Transcript
Phil
Welcome to Board Talk. This is Phil Zarrow with Jim Hall. We are the Assembly Brothers, Pick and Place. And we're here to talk about SMT processes and materials, equipment.
Jim
We're going to have to figure out pretty quick, because this question is about pick performance, it's not pick and place performance. And it comes from R.W.
The question is "What rate (in percentage) is considered world class when it comes to pick performance for SMT placement machines? Also, can you define pick performance?
Phil
I've got to tell you, in the world specifications, I don't know that I've ever seen it broken that way. And going back to your and my respective pick and place days, we used to do things like cycle time, and we say the pick from the same place.
I think in terms of performance, you think old R.W. was alluding to accuracy as opposed to speed, but the actual accuracy, and picking the component out of its respective pocket?
Jim
We're only concerned about the performance of the machine in picking parts out of the pockets of tapes and feeders. And thus the percentage, so what percentage are you successful when you go to pick up a part?
And why is this now becoming a question and not one that was asked 30 years or 40 years ago when you and I were working with the prehistoric placement machines?
Phil
Let's see if I can answer that.
1. We don't have enough specs to worry about, and the salesmen need something else to talk about.
2. R.W. is dropping a lot of components.
3. R.W. is not getting components out of his feeders.
Jim
I'm guessing, Phil, that the reason is that R.W. is having to deal with very small components - 0201s, maybe01005s. Having worked with 01005s, you know that picking them up becomes a big issue, because they're so small.
Phil
Seeing them then becomes an issue, let alone picking them up. Plus we all know the other issues attributed to picking them up, getting good release from the pick-up tooling.
Jim
Yes, but this is confusing. I'm shooting from the hip here, ladies and gentlemen, because pick performance can be affected by the machine. It can also be affected by the accuracy of the part within the pocket of the tape.
And again, if we're talking about these real small chips, you can get bill boarding flipped upside down parts. If the machine fails to pick it up because it's bill boarded in the pocket before it even attempted to pick it up, is that the machine's fault?
Is that be included in the percentage of failures?
Phil
My brother's made a very, very good point. It doesn't all aim down to the equipment itself.
We're equipment guys, and we always had a bone to pick with component guys. How good is the actual tape and the stuff, the way it's being fed? But to answer the question directly, I don't know of any metric per se.
Certainly, you suggested some good ones, Jim, but I don't know of anybody - we're probably giving the Pick and Place guys some really "Hot damn, here's a new one we can out spec 'em on." We know how that works.
Jim
I don't know. I just know that I've seen it in many people - have talked about the issues about picking these really small parts with perchance the need to use vision to find the part before you pick it up in some extreme cases.
What are world class? I honestly do not know. I would assume you'd have to rate it for a given size and style of part. I know a 0201 or an 01005 - I would assume that your success rate, if we say it's the success rate of the pick-up tools to be able to pick parts up from a feeder pocket, then I would assume it would be lower for an 01005, because it's simply harder.
You're bound to get more mistakes. But in terms of world class, I don't have any numbers.
Phil
But I'll tell you what, putting on my six sigma curmudgeon hat here, this is definitely an opportunity for a specmanship on the part of the Pick and Place guys.
Jim
And everybody understands the difference between placement accuracy at three sigma and placement accuracy at four sigma. We will not go there today. We won't go there.
Phil
We'll save that for another time. Until that time, this is Phil Zarrow and Jim Hall, Pick and Place.
Jim
Whatever you do, whether you're picking placing or handling both hats, don't solder like my brother.
Phil
Don't solder like my brother.
Comments
So many comments here and no hard numbers, so as yet no one actually answered the question just lots of comments (IMHO) From my decades of experience mostly with one of the top three placement machine vendors (But I would suggest not much of a gap between them) the actual component size has little effect as the machines were designed to place them in the billions reliably (Can't speak for all vendors though, but I guess this is why you ask for a number!)
While consulting to customers I always placed a target of 99.98% pick up reliability as world class (Acceptance spec would be less than this though) This number would be for the whole line (All feeders and components) actually the odd shaped components would be the ones to encounter issues but these would be dwarfed statistically as they were placed in lower volumes (But I did encourage the customer to drive all feeders towards this target) Phil and Jim are correct when they say it is not always the machines issue, but there are an enormous amount of engineering put into the machines to overcome some of those tape problems.
But also simple process engineering issues can be fine tuned by the user to drive performance up such as how the component shape is defined and tolerances set, how illumination settings are adjusted, even down to feeder acceleration profiles and many more things on high end equipment. In my experience it is all about attention to detail, and a drive for perfectionism and learning at the start of the learning curve.
Once world class performance levels are achieved it actually takes a lot less effort to sustain it. (However if you use low quality components and tape, and don't follow recommended feeder and equipment maintenance, you will never achieve such performance levels or more importantly remove such noise to have the chance to refine your process engineering expertise to fine tune the parameters mentioned above!)
Robert Gray
Actually, it is not how fast nor how accurately a machine can pick, or place, or both! There are a lot of other factors that come into play regarding overall performance. The biggest is set-up time. If the machine can pick and place 1 jillion parts per hour with 100% accuracy, but after the 5-minute run is finished the machine won't be available again until sometime tomorrow afternoon because every single feeder location requires re-teaching, etc., well, that's not really so fast, is it?
Richard Stadem, General Dynamics Mission Systems
Basically, it would seem that this is a comparison between a drum style head that rotates and very rapidly picks, typically a single value. Then, after loading the picks, moves along placing the parts onto the assembly. The drum type of machine really only out-performs a dual-head in the case where a lot of a single value and simple configuration part is being processed, like LED's, to be placed into a very large array.
As well stated and hinted at with the previous answers, the real objective is to make a fault-free assembly, quickly. The ultra fast picker has distinct limits of versatility and too often sacrifices place precision. Deciding 'World-Class' really cannot be done without the second portion, the place accuracy and speed thereof.
Jaye Waas, Renkus-Heinz
Guys, This is beneath you, as entertaining as your replies are it is cringe worthy to treat this as a valid question. R.W. needs to be revisited are queried about what he/she has been asked to investigate. Many machines can pick 200,000 pph (some can do 100,000 ppm). To pick and have the part properly positioned on the tip, and stay on the tip during transit to the placement, and placed correctly is the desired measurement. Six Sigma doesn’t care how many picks a process can do it cares how many the process can do correctly and not cause an error (OR waste)! Kudos Mr. Jeglum!
Ike Sedberry, Sedberry Sales Inc
You guys have successfully skirted the issue. Pick and place with out error is the issue. In the SMT assembly business it is not how fast you assemble but how you assemble fast.
