SATURDAY, AUGUST 23, 2014
Font Size
Topic Help on choosing the correct type of control chart..

Help on choosing the correct type of control chart..

Home Forums Old Forums General Help on choosing the correct type of control chart..

This topic contains 17 replies, has 1 voice, and was last updated by Avatar of  Anonymous 9 years, 5 months ago.

Viewing 18 posts - 1 through 18 (of 18 total)
  • Author
    Posts
  • #79487

    Hi all, i am 3rd yr student and am currently creating a control chart for a piece of work but am having problems identifying which type of control chart to use, i think it is an NP chart but if i get this wrong then it will really be detrimental. The scenario is this:”MSCI (Manufacturer of Small Circuits – Inexpensive) makes small inexpensive printed electrical circuits and most of the soldered connections are performed by hand. MSCI wants to develop a process control chart to see if the operation is in control. Samples of size 100 are withdrawn over a 48-hour period and inspected.”I have to create this control chart. Any help would be appreciated.Regards
    Ben

    #79490

    Stevo,
    Here it is about 24 hours later. How hard is it to figure out?
    The most ironic part is believing that a hand solder operation will ever be in control.
    Regards,
    Mike

    #79499

    um i am just not sure mate.
    good point about the soldering though.
    is it an NP chart?

    #79501

    What do you think it is and why?
    What do you think it isn’t and why?
    On a good day hand soldering will produce pretty but latently defective product. I am not sure if Handbook 217 is still available but it provides you with the numbers for the difference in reliability between hand and machine soldering. Basically you want pretty product and reliability doesn’t matter?
    If this were a machine solder operation you wouldn’t be wasting your time on this.

    #79504

    mike, this is a piece of coursework i am doing toward my degree, and i agree with you about the machine soldering, i know very little about the soldering itself, just the info we are given. I wish they had had a machine then i wouldnt have had to do the coursework.
    as for the chart i think it is a NP chart i should be using because the data given (48 samples and the number of defective circuits in the batches of 100) seem to be to attribute data, and then the fact that the sample size is the same leads be to think it is an NP chart!?

    #79509

    Ben bfc,
    It is attribute data. The sample size appears to be fixed (48 samples of 100?). Why not a p chart or a c chart?

    #79518

    sample size is fixed yeah, 100 samples every hour for 2 days.
    i dont think its a P chart because that is for non even samples.
    not a c chart because number of defects present are calcuable

    #79520
    Avatar of beverly daniels
    beverly daniels
    Participant
    Reputation - 0
    Rank - Aluminum

    c chart.  (assuming you are plotting the solder joint defects)  p and np are for defective units (defective board = one or more defects on the board).  c and u chart are for defects.  (defect = a bad solder joint)
    if you are combining different boards types together, you should use a u chart with the unit of opportunity = number of joints.  (of course if the count of joints for the boards are fairly similar you can default to a c chart…) 

    #79521

    beverly thanks for your help: it transpires that i have to do 2 control charts and they r slightly different.Number 1 – The title given above the data is “Number of defective circuits” Number 2 – Simply “Defectives”does this mean number 1 is a C chart and number too is a NP chart???Thanks Ben

    #79522

    beverly, thanks for your help: in fact i have to do 2 charts in total and they are slightly different by name:For the data in question 1 – the title is “Number of defective circuits” so i would assume that is a c chart.For question 2 it says “defectives” which i assume is a NP chart??RegardsBen

    #79523
    Avatar of beverly daniels
    beverly daniels
    Participant
    Reputation - 0
    Rank - Aluminum

    as long as your definitions are correct. 
    c or u chart for DEFECTS.  you may have sevaral defects on a single  board.
    np or p chart for DEFECTIVES.  a board is dfective if it has one or more defects on it. 
    2 cold solder joints on a board = 2 defects and 1 defective.

    #79524

    beverly thanks, yeah they are the definitions.cooli will get on and do them!

    #79544

    Ben,
    You should be cautious .. one of the reasons systematic problems are difficult to solve is because most people consider them random problems.
    Shewhart charts assume independence .. two dry joints are not necessarily independent. If you start plotting dependent defects the statistical control limits will become tighter, and you many falsely detect some special causes of variation! Unless of course you just want to blame the process engineer, who seem to be leaving industry in droves, leaving blackbelts to run the process!!!
    Regards,
    Andy

    #79554

    Beverly, you ae a nice lady giveing answer to ben.  i have to take myexam for black belt and need ansers to 3 questons.  tank you for your hlp.  Maybe uthers like Mike, phil, stan, Vinni and Darth can help to.
    1.  when duing a x bar and r chart shuld I plot against the mean, median or target?
    2.  in a DOE output what valu of p will tell me weather the varible is significant?
    3.  why is not the value of sigma the same as value i find in z table?
    Pleze hurry with anser since I have to turn in exam monday.  thank you nice lady four helping all students do tests.  much easier than us having to studi on our own.
    3. 

    #79556

    Vern has cataloged this classic isixsigma educational moment using the search keyword pleze and/or uthers for future reference. Thanks!

    #79557

    Andy,
    Lack of independence is really not the issue when you are in a hand solder process with the exception of a board type defect (PCB). If anything there is to much independence. Every joint, every board, etc is a completely independent event in terms of the solder joint. There is the solderability of the board and the component and it is complicated by operator variation. Probably the worst independent variable in the whole thing is the inspector (measurement system).
    This is a process (particularly with the volume) where control charts are going to drive you nuts.
    Just my opinion.

    #79558

    Beverly,
    Good advice but if you want this guy to learn he needed to figure out more on his own. Check out Jonas in The Goal – a better learning model.
    Just my opinion.
    Good luck

    #79565

    Mike,
    Most people these days use solder paste and wave soldering. One of the common problems with this type of process is board layout; especially when the ‘design’ allows components ‘suck out’ too much heat locally, as this can be a main cause of dry solder joints – movement is another. These types of failure are not independent.
    I’ve never worked in a facility that uses hand-soldering; the closest I’ve come to that is soldering by robots – on microwave components – so I can’t really comment on a hand soldering process.
    Regards,
    Andy
     

Viewing 18 posts - 1 through 18 (of 18 total)

The forum ‘General’ is closed to new topics and replies.

Login Form