PRISM Forum - Message Replies
Topic: PRISM Questions and Answers
Topic Posted by: SRC
Date Posted: Wed Jan 12 8:33:33 US/Eastern 2000
Topic Description: Welcome to the PRISM forum! Please feel free to post your questions and comments about the PRISM assessment software here.
Posted by: George Pyryt
Date posted: Tue Apr 11 12:14:22 US/Eastern 2000
Subject: Storage reliability
I would like to calculate the dormant storage reliability of a system using PRISM. I set duty cycle and cycling rate to zero and also did not input any stress parameters for parts. The result gives me a failure rate which includes solder joint and electrical overstress failure rate components for each part in addition to the part nonoperational failure rate. I also applied Process Grade Factor. Is this failure rate comparable to previous approach of using 1% conversion factor for dormant storage? If not, what data should be used to get storage reliability.
Subject: Dormant Storage Failure Rate
Reply Posted by: Norman B. Fuqua
Date Posted: Wed Apr 12 11:02:38 US/Eastern 2000
The dormant storage failure rate calculated using PRISM should be more accurate than the previous methodology of using 1% of the operating failure rate. This is because the PRISM models actually address the fact that the failure mechanisms involved in dormant storage are different than those in normal operation. That previous methodology, which you referred to, assumed that the failure mechanisms were the same, but that they just occurred at a lessor rate.
To improve its accuracy under all operating and non-operating conditions the PRISM RACRates models actually look separately at five main contributions to the total part failure rate: a) the operating contribution, b) the non-operating contribution, c) the temperature cycling contribution, d) solder joint contribution and e) the EOS contribution.
In general the method that you described, that of setting the duty cycle and cycling rate to zero with no input stresses, is correct except for one very important element. To get the correct solder joint contribution to the dormant storage failure rate it will be necessary to individually reduce the internal part temperature rise (delta T) to zero for each and every diode, IC, resistor, thyristor and transistor in the system. This is necessary because the PRISM RACRates models normally assign a nonzero default value for the internal part delta T.
Also, to get the most accurate cycling contribution you may need to reevaluate your decision regarding temperature cycling occurring in your storage environment. There may actually be some e.g., daily or seasonal temperature cycling and then set your upper (operating) and lower (dormant) temperatures and your annual cycling rate accordingly even though the actual duty cycle is zero.