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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


What is a typical SMT placement defect rate?

Views: 7664

#72392

What is a typical SMT placement defect rate? | 9 July, 2014

I'm looking for a typical SMT placement defect rate from others in the field so I can judge whether or not we are in an acceptable range.

If I give you a board with 200 components (0603, 0805, SOICs, 20mil MSOPs, tantalums, DPAKs and a few 20mil fine pitch QFPs), what percentage of boards coming out of the placement machine would be considered defect free, without having to nudge or fix any components before reflow? In other words, if you run 100 of these boards on your equipment, how many out of 100 would pass through test and inspection without defects?

Assume no odd form hand placed parts. A defect would be a missing part, wrong part, laying on its side, reversed, skewed off the pads or misalignment requiring nudging. Slight misalignment that would self-center during reflow and therefore not cause a defect is not counted.

Simple boards, complex boards...I'd like to hear about YOUR experience. Thank you in advance.

reply »

#72394

What is a typical SMT placement defect rate? | 10 July, 2014

I think what you are asking for is called First Pass Yield and for us that is in the upper 90s. It will depend on the board complexity and quantity - makes big difference if you run 10pcs or a 10000pcs. Regards, Emil

reply »

#72402

What is a typical SMT placement defect rate? | 10 July, 2014

Hi,

An answer to this question is very complex. The amount of components and their type is only one parameter to determine the level of failure. Additional parameters that can influence are: 1. The density components on the card 2. Machine type and accuracy 3. Kind of feeder

By the way you have prevented problems are solved by a software simulation while preparing the machine. Recommendation of our customers, I suggest you read about a tool called QPLAN: http://www.proventustech.com/#!qplan/c1qnl

For more information, you can send me an email: alexeis@proventustech.com.

Best Regards, Alexei

reply »

#72426

What is a typical SMT placement defect rate? | 16 July, 2014

We're a Tier 1 Automotive Supplier building, what I like to call, Peanut Butter and Jelly Boards. (High Volume, Low Complexity) Our first pass yield, using standard 1st Article Inspection and AOI is about 98-99% on any given day.

Last night, for example, we had 28 failed boards at ICT, while testing 2036 PCBs. (98.6%) Granted, I haven't had these failures analyzed yet, so there could be a few that failed due to ICT issues. (Sticky/Dirty pins, etc)

reply »

#72427

What is a typical SMT placement defect rate? | 16 July, 2014

Thank you for your comments so far. I believe I threw some people off by my wording. Clarification:

How feasible is it to run a board straight into the reflow oven without inspecting it first for any placement defects? Is nudging parts and performing a visual inspection before reflow typical or is it reasonable to assume everything on a 300 component board has been placed OK and there is no need to "double check" it before reflow? If you inspect before reflow, what number of placement defects before reflow is considered "acceptable" and do you have to nudge things? Thanks again!

reply »

#72428

What is a typical SMT placement defect rate? | 16 July, 2014

Hi,

If you can not rely on your programmer and the tools he uses then "double checking" can save you only materials but in time you really do not save often extend meaning. There are things you can identify before the stove, but if you use tools like AOI or visual inspection then you should be covered with solder problems. Let's say the issue depends on the product (density components, types of components, etc.) and tools that you have. I suggest that would decompose the matter to topics and see how you can save. For example can you troubleshoot your programmer level, whether manual tools provide you or need automated tools, and the like.

Best Regards, Alexei

reply »

#72433

What is a typical SMT placement defect rate? | 17 July, 2014

> Hi, > > If you can not rely on your programmer and > the tools he uses then "double checking" can save > you only materials but in time you really do not > save often extend meaning. There are things you > can identify before the stove, but if you use > tools like AOI or visual inspection then you > should be covered with solder problems. Let's > say the issue depends on the product (density > components, types of components, etc.) and tools > that you have. I suggest that would decompose > the matter to topics and see how you can save. > For example can you troubleshoot your programmer > level, whether manual tools provide you or need > automated tools, and the like. > > Best > Regards, Alexei

I'd like to add here, you need to decide/determine Consider running 100 boards With 10 parts on each board total number of solder joints might be much higher (SIMILAR for example) to the number of solder aperatures in your stencil for SMT only board (10 14pin SOIC IC's) One NO solder joint stops 1 board in 100 from working. This is one non functioning part in 100 parts placed This is ONE solder defect in 14000 possible solder joints

I've love to know some baseline figures in the context above for high mix, low volume assemblies...... Adding to the discussion to watch

reply »

#72437

What is a typical SMT placement defect rate? | 18 July, 2014

Hi,

If you are working in high mix low volume, your machine time is very expansive. In this case you must rely on the programmer, people experience, and make preparations offline (like Setup etc.), so do not hold the line in vain.

Best Regards, Alexei

reply »

#72450

What is a typical SMT placement defect rate? | 21 July, 2014

As has been pointed out by other well experienced individuals, this is a complex question. But the simple answer is "It is your decision". Choose any goal less than 100% and strive to reach it. There are hundreds of questions to ask before an objective conclusion can be reached: value of the PCB, type of machine, age, and budget. For instance if you have 20 year old poorly maintained equipment you should inspect every board. If you have decent equipment with qualified personnel, easily over 95%

reply »

#72802

What is a typical SMT placement defect rate? | 21 September, 2014

Our company has a process standard that requires a double person check at the beginning of every run (size of run is irrelevent, whether 20 panels or 500 panels). Every changeover that occurs a first piece inspection is made before reflow. If all is well there, we use AOI to determine later in the run if defects are occuring, and can attack the problem then before more defects are made. Usually issues arise from dirty laser alignment lenses or low vacuum from toolbits, something simple and quick to fix but can cause bad placements. Also, we check solder print first piece as well to verify solder is centered on the pads. I would have to say, and this is from my experience, the program itself is usually pretty stable, the machine wear and tear that affects placement is over a long period of time, and if you have a good setup, fiducials, and CAD data, there shouldn't be a real necessity for checking every panel.

reply »

#72818

What is a typical SMT placement defect rate? | 27 September, 2014

In my office we strive for no retouch prior to reflow, it can be done, you just need your operator to be competent enough to know how to set up your machine correctly. I would say 99% of parts that are misaligned are user error, they are either not correctly programmed, or poorly aligned or tolerances are to far out all of which can be adjusted for. I still inspect every board visually but with just a small amount of time and understanding put into programming I only have to adjust 1-2 parts per 1000 placed, and so can you.

reply »

#72845

What is a typical SMT placement defect rate? | 2 October, 2014

the data collected from a standard SMT process is typically very noisy due to the natural.Initially, the most common evaluation method was placement yield.Offset Placement After Solder Screen-printing can reduce or eliminate these defects through corrected placement at the solder printing positions to promote the self-alignment effect.

reply »

Comprehensive Analytical Services and Support

MSD Dry Cabinets