REF5050

Alexander,

All of our products, including REF5050, undergo a life-time test in a high temperature oven to determine an equivalent shift of various parameters after 10-years (87,600 hours) of constant operation at ambient temperature.  Based on the above test, the expected long-term shift after 10 years of operation is equal or less to the maximum initial accuracy of 0.05% specified in the datasheet (500ppm). 

Even though REF5050 may work well beyond our specified 10-year life, we do not have the data to directly answer your question about the expected long-term shift after 20 years (175,000 hours) but IF the part is still alive, the expected shift should be: 50ppm*√175=661ppm.

For more detailed explenation of the long-term shift, please read my blog under link below:

http://e2e.ti.com/blogs_/b/precisionhub/archive/2014/02/20/ic-long-term-stability-the-only-constant-is-change.aspx

Marek Lis said:

Even though REF5050 may work well beyond our specified 10-year life, we do not have the data to directly answer your question about the expected long-term shift after 20 years (175,000 hours) but IF the part is still alive, the expected shift should be: 50ppm*√175=661ppm.

I've examined many datasheets and application notes over the years and never come across any such statement about a 10-year life; can you provide any references? Why wouldn't anything as important as this not be mentioned in every datasheet?

The way you have worded it implies that TI expect significant numbers to fail by 10 years, but I can't believe you mean that - presumably you are referring solely to the lifetime before parametric shifts double as discussed in that blog? If so it should still be stated in datasheets - especially since you start your blog with:

"Some of the most frequently asked questions we get in the Precision Amplifier Forum here in the E2E Community™ have to do with an IC’s long-term stability of various parameters"

Thanks,

     Tony Holt

Tony,

It is a semicunductor industry-wide standard to life-test all new products for an equivalent 10-year operation at 25C and there are many industry standards like this that are not specifically mentioned in TI's or our competitors' datasheets.

I am not sure where you get that a significant number of units will fail PDS limits after 10-years.  Since most TI's datasheet max/min specifications are based on 4-sigma Normal Guassian distrubution, only 1 out of 15,000 units may be initially close to the max/min limit - see below output voltage initial accuracy distribution for REF50xx (thus 1-sigma = 0.05%/4 = ~0.0125%).

If you consider 2-sigma distribution (+/-0.025%), statistically only 1 out of 22 units may be outside such limit and only such units after 10-years may at most shift by another 2-sigma (+/-0.025%) and thus get beyond the PDS limit of +/-0.05%.  If the initial accuracy and long-term shift were correlated with each other, the probability of this happening would be: P = 1/22*1/22 and therefore 1 out of 484 units could be outside PDS limit after 10 years.

However, since there is no correlation between the initial accuracy and the subsequent long-term shift (meaning worst initial accuracy units typically will NOT shift the most), one must use a vector addition (sum of squares) to calculate the overall long-term shift:  since (0.05%)/sq-rt(2) = 0.0355%, therefore, only the units outside of approximately +/-3-sigma initial accuracy distribution (0.0355/0.0125) AFTER 10 years may shift beyond 0.05% [sq-rt(0.0355^2+0.0355^2)] PDS limit.  Thus probability of finding REF50xx units outside of PDS limit 0.05% after 10 years of constant operation is: P = 1/370*1/370 = 0.0000073 or 0.00073% or  ~ 1 out of 136,900 units - see below Gaussian Distribution table for details.