Dr. Lasky’s Blog
Recent Blog Postings
Folks,
When IBM sold most of its Endicott, NY facility to Endicott Interconnects (EI) in 2002, I was among the doubters that EI would fare well. Having as its core business printed wiring board fabrication and electronics assembly, two of the most competitive fields, it seemed only a matter of time before EI would fade away. This thought was particularly painful to me as I spent 26 years working at this IBM site and was also an “IBM brat,” as both of my parents met (in the 1930s) and worked there.
So, my expectations were not too high when I worked with some friends at EI to organize a tour of EI’s PWB fab facility for some of my colleagues from Indium Corporation. I have to admit I was blown away with what I saw. Our tour started with an overview by EI President “Jay” McNamara. He told the story of how EI used the experience, drive and creativity of the former IBMers to develop a business focused on large 26x36 inch (66 X91.5 cm), thick 0.400 inches (1.02 cm) PWB fab and assembly. The PWBs they design and manufacture are so challenging to make, that for 60% of their business they have no competition. The Department of Defense considers them a “mission critical” site.
In a time when much of electronics manufacturing is going off shore and there is a sense that the US is losing its competitive edge, it is comforting to learn of EI’s great success. Best wishes to them in the future.
Cheers,
Dr. Ron
Tags:
Folks,
I continue to get much interest in the solder alloy density calculator I developed some years ago. It is now online at http://www.indium.com/densityCalc/. It assumes no chemical interaction between the metals and no formation of interstitials. It works well for solder alloys.
Many people have an incorrect idea of how to perform this calculation. The most common incorrect concept is to multiply the % by weight of each alloy times its density and add them together . The correct derivation is below:
We want to find the density of an alloy composed of 3 metals. Assume the mass of the alloy is M. Metal A has a mass ma and a density da, Metal B has a mass mb and a density db and Metal C has a mass m c and a density dc. The total volume, V, of the 3 metals is va + vb+ vc, however since v = m/d, the total volume can be expressed:
1) V = ma/da + mb/db +mc/dc The density of the resulting alloy is D = M/V, hence 1/D = V/M, therefore:
2) 1/D = V/M = (ma/M)/da + (mb/M)/db +(mc/M)/dc
Now ma/M is the mass fraction of a, which we will call Xa, and similarly Xb and Xc for metals B and C.
Equation 2 then becomes:
1/D = Xa/da + Xb/db +Xc/dc which is our solution.
This principle can be applied to alloys of more than 3 metals.
Cheers,
Dr. Ron
Tags:
Folks:
A while ago I posed the following as a typical type of question an SMTA Certified Process Engineer should be able to answer:
A 7 zone oven has a heated tunnel length (TL) of 250 cm. A 20 cm PCB (PCB Length =PL) needs to be in the heated part of the oven for 4 minutes. What is the maximum throughput (TP) in PCBs per minute that this oven can support?
The answer:
First we need to establish the belt speed, if the PCB needs to be in the 250 cm heated part of the oven for 4 minutes (t), the the belt speed BS = TL/t or 250 cm/4 minutes or 62.5 cm/min. So every minute the belt presents 62.5 cm new space for PCBs. So
TP = BS/PL = (62.5cm/min)/(20 cm/PCB) = 3.125 PCBs/ min, since a little spacing would be desired between PCBs, about 3 per minute is a practical maximum.
Golf Update: As I had mentioned early in the year, for the first time in many years I really worked on my golf game, playing about twice a week, and exercising and stretching. My typical scores were in the high 70s to low 80s, compared to mid to high 70s in my youth. Perhaps the most interesting thing is that with the modern clubs and balls I hit the ball about 30 to 50 yards father than when I was 20. I don't score as well because my shot game is not as good now (due to less practice I think).
As a result of this extra distance and interesting comparison can be made. From the time I was 10 years old until I was 60 (last year) a played about 1000 rounds, yet I scored only 4 eagles (2 under par on a hole). Form August 2007 through August 2008 I scored 3 eagles.........the difference? I reach many par 5s in two or drive a par 4 today, whereas when I was young it was a rare occurrence.
Cheers,
Dr. Ron
Tags:
Folks,
The essense of WEEE is to one day approach 100% re-cycling of electronics. Around 2002 the world was at something like 10% re-cycling. With the high price of metals and the value of recycled components* re-cycling is not only the right thing to do, it is also cost effective. It is a critical part of "Sustainable Design," which many consider a new wave in the evolution of technology.
Therefore, one of the first questions to ask of a new technology is "does it easily support re-cycling." Thus easy dis-assembly is a requirement. So, when you see new electronics manufacturing technologies that make grand claims, check and see if it supports easy re-cycling.
The photo is from the January 2008 issue of National Geographic. It is easy to see from this photo why recycling is needed.
Cheers,
Dr. Ron
* Hopefully legitimate re-cycle components. Re-cycled components have a value to play in very inexpensive electronics and what pollution they save the environment when these components don't have to be made "again". The point is that anyone buying recycled components should know that they are just that.
Tags:
Folks,
It seems that there is several year lag before many of "our"* concerns become mainstream. As an example, in December of 2005, I posted a note Re discussions I had with PC Magazine's famous "Cranky Geek" John Dvorak. As of that late date, a well informed chap like he, was not aware of RoHS. Now PC Magazine gives a green rating to all products Re their RoHS compliance and energy usage.
Also in late 2005 I discussed counterfeit components. The October 13, 2008 issue of Business Week has an article "Dangerous Fakes" . This article discusses the growing threat to the United States' military hardware from counterfeit components. Perhaps the most unsettling point in the article, was how easy it is for an inexperienced person to obtain a government supplier code and supply the pentagon with parts obtained through searches on the internet. The supplier is often unaware that they are occasionally supplying counterfeit components.
Many have worried about reliability of RoHS compliant electronics products. I believe that the increasing numbers of counterfeit components are creating a reliability risk that will dwarf any concerns about RoHS reliability.
Cheers,
Dr. Ron
* Our = those in the electronic assembly industry
The image is of a stealth fighter from http://timm84.files.wordpress.com/2008/06/stealth-bomber.jpg. By using this image, I'm not suggesting that this fighter has any counterfeit components in it.
Tags:
Folks,
Last week Indium Corp held its annual Technology Summit. There were many excellent technical presentations. Chris Anglin's brought out some information I would like to share with you.
In stencil printing of fine features, the area ratio (AR) of a stencil aperture is an important metric. AR is defined by the area of the stencil opening divided by the area of the side walls. For a circular aperture this is:
AR = Pi * r2 / (2Pi * r * t ) = r /2t or D / 4t, where r is the aperture radius and t the stencil thickness (D = 2r).
Experience has shown that for successful printing, AR should be >0.66. Chris pointed out that successful printing is more dependent on consistancy (small standard deviation) among the printed solder paste deposits than a transfer efficiency (TE) close to one. Transfer efficiency is the ratio of the volume of the printed deposit divided by the volume of the aperture. In other words, you can have successful printed with a TE of say 0.5 as long as the variation among the deposits is small.
What was most surprising to me was the AR of common objects. Chris used this approach to convey the concept in terms of objects that we can relate to. The can of Bumble Bee tuna in the photo has an AR of only 0.49! Intuitively, I think most of us would have thought that an aperture with the dimension ratios of a tuna can would be good for printing. Chris points out that it is low by almost 20%.
AR shows the extreme challenge of 01005 printing. To print for these small passives, the aperture will need to be about 6 mils. To have an AR of = 0.66, the stencil thickness must be t = D/ (4*0.66) = 6/(4*0.66)= 2.3 mils. Since most other devices would might need a 4 mil or thicker stencil to obtain enough solder to make a good joint, a step stencil or two printing steps may be needed for successful 01005 assembly.
And I'll bet people are working on a developing an 0050025 passive. Yikes!
Cheers,
Dr. Ron
Tags:
Folks,
The following question is one that students in SMTA Processes certification struggle with. Jim Hall and I are always a little surprised, as it is taught in the refresher class quite thoroughly and seems straight forward. See how you do. I will post the solution in a few days.
BTW, I am teaching a "Pre-Certification Class" to review the concepts required to pass the SMTA certification test on September 16, 2008 at the Concourse Hotel at the Columbus, OH airport. Come and join us if you are in the area.
The question follows:
A 7 zone oven has a heated tunnel length of 250 cm. A 20 cm PCB needs to be in the heated part of the oven for 4 minutes. What is the maximum throughput in PCBs per minute that this oven can support?
The photo is of several Speedline ovens and wavesolder machines.
Cheers,
Dr. Ron
Tags:
Folks,
Since 2002 SMTA has had an SMT Process Engineer Certification Program that was developed by Jim Hall, Phil Zarrow and me. Some folks are a little nervous about taking the workshop and exam. In light of this concern, I have been asked by the Ohio SMTA chapter to give a 5 hour workshop on the types of problems one would be expected to solve. I have proposed a list of problems to be covered in the workshop. Here is a typical one that I might use. See if you can do it.
Component placement is the limiting process in an SMT line for throughput. Throughput (cycle time) is now one board every 50 seconds, when the line is running. There is one chipshooter (CS) and one flexible placer (FP). The bill of material (BOM) is 300 passives, 24 SICs (simple ICs), 8 CICs (complex ICs). The chipshooter is taking 50 seconds and the flex placer 20 seconds. All passives are being placed by the CS. The FP places CICs in one second. It places passives and SICs at the same rate.
1. Maximum throughput will be obtained when the CS and FP take the same amount of time. This can be accomplished by moving passives from the CS to the FP for placement.
a. What is the minimum cycle time if the line is balanced?
b. How many passives were moved to the FP to achieve this cycle time?
Note: We assume all other processes can keep up with the new and improved cycle time.
I will publish the solution in a few days. I expect to be giving this workshop quite frequently in the future.
Solution:
The FP takes 20 seconds to place the SICs and the CICs. The 8 CICs take 8 seconds, so the 24 SICs must take 12 seconds. Hence, the placement rate for SICs is 2 per second. The CS places 300 passives in 50 seconds, so it places 6 passives a second. Since the CS takes 50 seconds, to balance the line we must move passives to the FP. Our goal is that the time spent by the CS is the same as the FP. Let us assume that the number of passives we have to move to the CS is x, then time balancing can be expressed by the equation:
tcs = (300 -x)/ 6 sec = tfp = 20 secs + x/2, solving for x: 300 - x = 120 +3x => 4x = 180 => x = 45
So 45 passives are moved from the CS to the FP, hence the cycle time is now: (300 -45)/6 = 42.5 sec.
The solution was quickly verified by LineSimulator(TM) software I developed. If anyone would like a copy of this software tool, send me a note at
RLasky@indium.com.
Cheers,
Dr. Ron
Tags:
Folks,
Well, I arrived at SMTAI a few days early to take my wife, daughter and two granddaughters to Disney World. We averaged 7 miles of walking a day, Friday through Sunday. My 2 year old and 4 year old granddaughters had a great time. However, we learned that the 3D movie, "Honey I Shrunk the Audience" is a little too scary for a 2 - 4 year old.
It ends up that mid August is a good time to visit DW, as many families have stopped coming as they prepare for their youngsters to attend school.
The weather was also mild (i.e not extremely hot and humid) as a prelude to tropical storm Fay. So my family had 4 good days of touring, before the strong rains came.
The Show? The tech sessions and workshops seemed quite well attended on Sunday-Tues. Hot topics: the need for recycling driving "design for disassembly, counterfeit components, and conversion of lead-free alloys to a few standards seemed of interest to many.
Sadly, it looks Fay has reduced traffic on the show floor as I write this note on Tues afternoon.
If you are here and can make it, join us at our "Meet the Bloggers" meeting on Wednesday, August 20 at 11AM at Indium's booth # 517. It should be fun. Five of Indium's bloggers will be there.
Cheers,
Dr. Ron
Tags:
Folks,
Over the past few years I have been surprised at the interest in the Excel based alloy density calculator that I developed. Thanks for your interest!
In light of this interest, team Indium suggested that it be made available on the web for all to use. Find it here.
I have written some other software that I will be discussing in future posts. One even perfroms the calculations to time balance the component placement machines on a line. Another matches the solder paste spec to the reflow profile. Stay tuned.
Cheers,
Dr. Ron
Tags:
Want to read more? Browse the archive of past entries.
