# Reliability Analysis for board

#### Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

•  Subject
• Author
• Posted on
hi all

i have to do reliability analysis for a board design. i have seen some
documents in which some calculations has be shown. but i m not getting
it. can somebody guide me how to do it? what are these terms lambda,
quality
factor of parts n all..

where i will get the correct procedure to do it? is there any good
document
for it?

shridhar.

Re: Reliability Analysis for board

MIL-HDBK-217F provides some information.  For tools, one choice to look at
is Item Software.  You enter all your parts and out pops a failure rate for
the whole thing.

Re: Reliability Analysis for board

Time to get some books on probability and statistics, or go back
to school.  You do not have the background for that job.  BTW,
your shift key is broken, as is the apostrophe, which makes your

--
Chuck F ( snipped-for-privacy@yahoo.com) ( snipped-for-privacy@worldnet.att.net)
Available for consulting/temporary embedded and systems.
We've slightly trimmed the long signature. Click to see the full one.
Re: Reliability Analysis for board

Calculating reliabilty can be a complex and difficult procedure.  Strictly
speaking you need to know the operating environment of every component, its
electrical stress, its reliablity and and manufactured quality.  Some
software has been written to do this but is is expensive and its value and
the accuracy of its results is questionable.

To get an idea of the likely reliability at the design stage, the parts
count method can be used.  This assumes common environment and stress
levels on all components and uses a simple multiplying factor for different
qualities of components.  For each type of component you work out the
failure rate from the base falure rate for that component multiplied by the
environment multiplied by the quality factor multiplied by the number of
components.  Then sum these valuse for all components to get the overall
failure rate.  take the inverse to get the mean time between failures.

I first came across the parts count method in the old US MIL STD Reliabilty
Handbook which may still be available.  The only problem with this is its
values tend to be rather pessimistic especially with the advances in
component quality caused by the consumer electronics explosion.  At that
time I created a spreadsheet which saved a huge amount of effort in
calculating reliability.  You just entered the number of each type of
component and out popped nine answers covering each of three different
environments and three different quality levels.  Some years later I
created a DOS version of this which I released as shareware.  I got just
one registration.  I still have it lying around somewhere and you are
welcome to a copy if you widh.

Ian

Re: Reliability Analysis for board

hi Ian,

first of thank you all for your kind replies. i really got some good info.
now what i want to know is to calculate failure rate for individual part
from where i can get this quality factor (PiQ) and enviornment factor (PiE).

also how can i calculate base failure rate? is there any formula for this.? is
same i.e. "Lamda G"?

if i m right with this information, then i just have to calculate
part failure rate = quantity x base failure rate x PiQ x PiE

thanking you,
shridhar.

Re: Reliability Analysis for board
hi

i just gone through MIL-HDBK-217F document n found useful information. What i
think now that for base failure rate (lambda B) and all other factors like
enviornmental factor (Pi E) , quality factor (Pi Q), Application factor (Pi A)
etc.. i can look in to this document.

just find out the suitable vaules depending on ur operating enviornment( ground
benign, ground fixed or airborne applicaiton etc), quality designator(screen
level) and also with ur junction temperature from this document (MIL-HDBK-217F)
and them calculate part failure rate, which then multiplied with the no. of
components (quantity) .

am i correct with this assumption.? waiting for ur reply.

thanking you,
shridhar.

Re: Reliability Analysis for board

That is correct.  These are the values I used in my program.

That is basicaly it.  However, you should be aware that the MIL STD Handbook
takes a rather pessimistic view of the quality of commercial (that is non
MIL SPEC) components.  IIRC PiQ is about 35 for commercial grade components
which IMHO is far to high.

Ian