Home › Forums › Old Forums › General › PPM calculation
This topic contains 20 replies, has 14 voices, and was last updated by Jim Shelor 7 years, 9 months ago.
How do I calculate logically PPM’s caused either by design or manufacturing non-conformances? Should it be based on actual RMA’s or the occurence identified on the NCR?
Scenario: There were 42 occurences of defective DN cards due to design in July. An NCR was raised and recall took place. Not all these cards were returned simultaneously due to different customers. No history of how many these cards were sold or out on the field. A PPM report was needed for the month of Dec.
So for Dec 06 our total sales was 7653. Our current formula is:
42/7653 = 0.00548804 x 1,000,000 = 5488
The formula applies only when the NCR is closed i.e. NCR raised in July was closed in Dec. Does this make sense?
I am confused and I need to understand PPM’s better and logically.
Thank you
Hello,
I am a bit confused here, first you saying that:
‘No history of how many these cards were sold or out on the field.’
then later on calculating the ppm you show total sales number for december 06….
If I understood you correctly you are trying to calculate external ppm – defect rate related to customer complaints.
If so, then in my case I am able ton get the production date (down to exact day) of our product (It is being put on the label of the product). and when having also sales numbers I can do a calculation for any period of time.
defective from specific period/sales from specific period x 1000000 = external PPM
Yes, multiplication is very confusing. My 9 year old son can help you.
Steve, you don’t have to be rude with your response. I am confused myself being new to quality plus I did not come up with this formula, this is from someone more qualified than me.
I am looking for a better understanding about ppm to help me in my job.
Dear Leovina,
PPM (more commonly known as DPM (defects per million)) is calculated using the number of failures for a specific group of parts divided by the total number of those specific parts sold multiplied by 1,000,000.
In your case, you want to know the PPM for the cards sold in December 2006. To calculate that number, you would take the number of non-conformances reported for the cards sold in December 2006 and divide by the number of cards sold in December 2006 and multiply by 1,000,000. You would not use non-conformances for cards sold in any other month.
Since you did not list any non-conformances for the month of December, your PPM for December would be 0.
Your PPM for July would be 42/the number of cards sold in July multiplied by 1,000,000 because the non-conformances you have for July would be only on those cards sold in July.
If on the other hand you are being asked for the cumulative PPM for cards sold from July through December, you would take the total number of non-conformances reported from July – December divided by the total number of cards sold from July through December multiplied by 1,000,000.
The non-conformances must be included within the sample (number of cards sold) to calculate a PPM. If I read your post correctly, the 42 non-conformances occurred in July, accordingly, these non-conformances would not be included in a calculation of PPM for the month of December. Only the non-conformances reported on the cards sold in December would be included in the PPM calculations for December.
I hope this helps.
Sincerely,
Jim Shelor
Dear Leovina,
Some people on this site get their kicks and demonstrate their vast knowledge of all things by insulting those or us who try to use this site to learn from each other.
I will be the first to admit that I do not know it all, accordingly I do not use that form of expression to demonstrate my superior intellect.
It is worth neither your time nor your energy to get upset at these superior individuals. Just let it roll of your back, ignore it, and someone who wants to learn and help wll eventually try to answer your question.
Sincerely,
Jim Shelor
let’s summarize the treatise on the calculation of PPM:
as you only have one defect opportunity (card rejected/accepted) simply take the defect rate (proportion of defect) and multiply by a 1,000,000 instead of 100. thus by pure mathematical magic a parts per hundred (percentage) calculation is transformed into a parts per million calculation. i’ll leave it up to steve’s kid to calculate the defect rate … that’s beyond my skill …
Sir,
I want information about ppm calculation. I confused in ppm calculation . I want clear detail and formula for ppm calculation.
Thank you
Excellent Reply.This forum is made to help people not to insult them?
I think your questions is regarding the months right? your thinking it’s Dec. and I’d like to no the PPM for December but all the defects were from months ago right? we’ll that’s the nature of the quality business… it’s not so much using the calculations – you obviously have the equation in hand, it’s thinking about how to get a process in place so that you get real time feedback (or as expedient as possible anyways). I’m in the same category… I’m looking to evaluate R&D quality but there really no effective way to get near instant analysis on how well their designs went without having it go into the field, operate for months or years, and problem reports start coming in. When you find a method let me know.
good luck,
Loc
Dear Steve (the guy who leaves messages about his 9 year old) – your a total tool. You are one of those people that goes around acting like he is smarter than everyone else to mask your deep and rampant insecurities. I hope you get beat up at work.
The the questionee – keep pushing man.
You cannot release a PPM calcualtion without knowing the quantitiy shipped.
The calcualtion for PPM is quite simple but without the denominator it is impossible to give any result
The number returned simply tells you how many people considered it important enough to read your notice and respond accordingly.
A bastardized method I’ve seen is to quantify the total number of things you shipped in a given month and use it as the denominator. The numerator is the total number of confirmed defects you counted that month take the result time 1,000,000.. it isn’t ppm but it as close as you can get without actual data.
PPM is related to the manufaturing in mechanical industries its used to calculate the rejection on total production.
Formula to calculate PPM is
PPM= Total Rejection/Total Production *1000000
Hi,
Particularly in Printed circuit board assembly which of the following method of calculation is right.
1.=(No of defects/total cards produced* No of solder joints per card)*1000000.2.=(No of defects/total cards produced * No of possible type of solder defects)*1000000. Other than this is there any other formula please let me know.Regards,
Basavaraj
In six sigma the important metric is DPMO or defects per million opportunities.
The posting of the printed circuit boards jogged my memory.. In the automotive world PPM was sufficient but in most industries including the automotive DPMO is more accurate as not all defects make an item defective.
Dear Jim Shelor,
I am really new at this and u have been great explaining this DPPM issue. So here it goes.
If I understand it correctly,
Qty of rejects parts divied by the total Qty of good parts X by 1,000,000
Is that correct?
Alan,
Change that to total rejects divided by total produced and you have it.
Jim Shelor
Thank you Jim, you make it so easy. simple.
what can you tell me on first pass yields and whst the difference is with calculating anyother yields?
a lot comes in and you inspect it,
total reject divided by total received gives you the yield correct?
if something at 100% drops 20% what will this equal in % and how you calculate this?
Jim,Good to see someone really trying to help. How refreshing! Just one comment – if the PPM is to be really accurate, the calculation has to be on production from the same time-frame, i.e. rejects returned in December which were made in November should not be counted when calculating the reject PPM for the month… I guess it depends how accurate you want the number to be and how relevant it really is!Rod
Ron,
You are absolutely right. Defects should be counted against only the production run that produced them.
Respects,
Jim Shelor
Good grief! Sorry about the typo in your name Rod.
The forum ‘General’ is closed to new topics and replies.
© Copyright iSixSigma 2000-2017. User Agreement. Any reproduction or other use of content without the express written consent of iSixSigma is prohibited. More »