iSixSigma

What Do You Consider a “Part” in an MSA?

Six Sigma – iSixSigma Forums General Forums Methodology What Do You Consider a “Part” in an MSA?

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #54205

    Alex Pamatat
    Participant

    I’ll start with the questions, read below for more detail:
    1) What is considered a “part” in an MSA?
    2) Should the model be a three factor nested model?
    3) Is each wafer a different “part” in my MSA?

    I’m designing an attribute MSA for a new piece of metrology equipment used to measure defects for a gold ball jetting process for semiconductor wafer processing. The output of the inspection will only be bins (i.e. 1,2,3,4,5,6,7), each bin corresponds to a certain failure mode.

    As I’m working through the MSA design and trying to identify the factors I’m trying to understand how to classify the wafer component. What exactly is considered a “part” in an MSA? Would each wafer be considered a “part”? What we call parts in our fabrication facility would represent a different product (i.e. different layout, maybe different number of die per wafer, maybe require slightly different process, etc.) and I’m not quite sure if this is what is meant by “part” in an MSA.

    Here is how I envisioned the MSA:
    -There are three wafers, each wafer is a duplicate of the other (i.e. same layout, process conditions).
    -There are 86 die on each wafer which will be measured.
    -Each wafer will be loaded, measured and unloaded three times per day.
    -Each time a wafer is loaded I will run the inspection 3 times without unloading/loading (this would be the repeatibility measurements).

    I assumed this would be a 3 factor nested model which would make the design: Wafer + Day(Cycle) + Cycle(Day,Wafer) + Repeats(Cycle,Day,Wafer). Does everyone agree with that?

    • This topic was modified 8 months, 1 week ago by Katie Barry.
    0
    #194226

    Chris Seider
    Participant

    You’ve partly answered your question. Each die is considered one part. Can you reload a “die” into your measuring device to get repeat measurements? Is there only one device that measures the attribute defect–if not, just treat them as “operators” as discussed in standard MSA checks.

    0
    #194231

    Alex Pamatat
    Participant

    The only way I could reload a die is by reloading the entire wafer….and in that case, yes, I could get repeat measurements. I’ve never heard of a die being considered a “part”. Why wouldn’t each wafer be considered a “part”?

    Yes, there is only one module on this one piece of equipment that measures the attribute defect.

    I planned on using the die (and wafer-to-wafer variaton) to measure total process variation for the %R&R calculation. Now that I think of it, how is total process variation calculated for an attribute MSA? Note, there is no historical run data for this process. Do you quantify the attribute data? For example, do I use the std dev of the # of defects per die, or perhaps use the std dev of the number of bin flips per die (i.e. a measurement failed in cycle 1 for bin7, during the second/repeat measurement the die failed for bin2)?

    This is the first attribute MSA I’ve run and there doesn’t seem to be anyone at my company that can advise on this type of MSA. Thanks for the help.

    0
    #194236

    Chris Seider
    Participant

    If you treat the multiple sites as individual parts in a wafer, you can use the attribute agreement analysis. You have two options if using Minitab. You can consider the site as defective or not (good or bad) or if you differentiate the type of defect, you can give it various states as “good, defect1, defect2” and see if the repeat measurements are the same. The caution is reloading the wafer, you’d have to make sure the orientation stayed the same within the device for the 2nd measurement.

    0
    #194239

    Alex Pamatat
    Participant

    Thanks for the continued followup Chris.

    I was up all night working on this. During my research I discovered that I should be running an attribute MSA, not a continuous variable MSA which is what was intended in my original post. Here’s my update based on what I learned last night….am I getting closer?

    IΒ’m trying to design an attribute MSA for a new piece of metrology equipment used to measure defects for a gold ball deposition process for semiconductor wafer processing. Each gold ball that’s deposited is analyzed for certain criteria, based on the results of the analysis the ball either passes or fails. The output of the inspection will only be bins (i.e. 1,2,3,4,5,6,7) where each bin corresponds to a certain failure mode however to reduce complexity I will simply assign any gold ball that’s binned a “fail” and any that are not as “passing”.

    I will have three wafers at my disposal. On each wafer is 86 die, on each die is 244 gold balls, each ball is inspected during the inspection step. The equipment uses automated inspection: all loading, alignment, inspection and binning is done by the equipment automatically.

    Will someone please tell me if my proposal for designing and running the attribute MSA is correct:

    1) Run automated inspection on one wafer to generate a list of die which the equipment believes to be passing and failing.
    2) Select die which have representative defective (failing) gold ball deposition and passing gold ball deposition. Try to select die with a 1/1 ratio of good to bad parts is recommended (I believe this will be highly unlikely, a 1:20 ratio of bad:good is more likely and that’s in the extreme case). I propose selecting 10 die on the wafer which are spread across the entire wafer surface.
    3) Have an expert appraiser categorize each gold ball on these 10 die to confirm passing balls are indeed passing and failing balls should have failed.
    4) Run wafer through the inspection 10 times. This sequence would be load, align, measure, unload, repeat.
    5) Generate list for each cycle to record whether a unique gold ball passes or fails for each measurement run.
    6) Use the methods described at the link below to analyze the data. In my case I’m assuming that there is only one operator (the equipment). Note, this is a first of a kind tool for a process with no process record. There is no historical or data or data from another tool to use for comparison. We’re using this MSA data to accept or reject the inspection equipment from the supplier.

    http://www.isixsigma.com/tools-templates/measurement-systems-analysis-msa-gage-rr/making-sense-attribute-gage-rr-calculations/

    Do my methods seem sound? Any recommendations on how this could be improved?

    Individual repeatibility (step 5 at link) and individual effectiveness (step 6 at link) seem straightforward to calculate since there is only one operator however this only provides repeatibility, how do I calculate reproducibility with only one equipment (equipment = operator)? Can I use the load/measure/unload sequence for the reproducibility and simply measure the wafer 3 times without load/unload sequence to gauge repeatibility?

    Thanks,
    Alex

    0
    #194240

    Chris Seider
    Participant

    Sounds like you got it! Grats…..I’d suggest, with my limited knowledge of your specific process specifics, to only use one wafer unless time/cost is minimal. I’d suggest only running through 3 times maximum to save time.

    Each of the gold balls is a unit being tested.

    0
    #194242

    Alex Pamatat
    Participant

    Thanks. Took me until 2:30AM but I thought I got close. I do plan on only using one wafer, between trying to force the entire spectrum of defect modes, inspecting all the gold balls and running the wafers it will take a long time and my time at the vendor site is limited.

    My questions still remain though: can I use the load/measure/unload sequence for the reproducibility and simply measure the wafer 3 times without load/unload sequence to gauge repeatibility?

    0
    #194243

    Chris Seider
    Participant

    You are using one gage that doesn’t (presumably) have operator impact on measurements so you are checking repeatability.

    If you are using Minitab, go to Stat>Quality Tools>Attribute Agreement Analysis and click on help. You will see the output you get. Pay most attention Each Appraiser vs Standard.

    0
    #194244

    Alex Pamatat
    Participant

    Ok, I see what you’re saying. Since the definition of reproducibility is that different assessors would come to the same conclusion…..and we don’t have different assessors (equipment/gage) then we cannot measure reproducibility. Did I sum that up correctly?

    We don’t use Minitab here, we only use JMP. Company tried to convert to Minitab years ago to save money and evreyone revolted. We’re all back on JMP now. JMP does have some attribute gage features, I’ll look into these.

    Thanks again.

    0
    #194249

    Alex Pamatat
    Participant

    I’ve summarized my proposed model tree, could you please have a look at it and let me know if it’s correct.

    0
Viewing 10 posts - 1 through 10 (of 10 total)

You must be logged in to reply to this topic.