[volt-nuts] LM399 Long term drift specification

[Tony Holt]

On 11/09/2014 01:50, Mike S wrote:
> On 9/10/2014 7:00 PM, Tony wrote:
>> I've just noticed that TI and Linear's specs for 'Long Term Stability'
>> (typical) are different. TI state 20ppm/1000Hr while Linear state
>> 8ppm/SQRT(kHr). That's  a big difference - is this likely to be a real
>> difference or just specmanship?
>>
>> I note that Linear (in Note 4) also state that "Devices with maximum
>> guaranteed long-term stability of 20ppm/SQRT(kH) are available."
>> Presumably they would be a special order as there doesn't appear to be a
>> unique part no. Would they be likely to be much more expensive?
>
> Isn't 8ppm/SQRT(kHr) better than 20ppm/SQRT(kH)? Why would the latter 
> be more expensive? Or is it the difference between "typical" and 
> "guaranteed?"
>
I'm guessing that typical in this case means the one sigma value so the 
three sigma value would be 24ppm. In any case three sigma still only 
means 93.32% of parts come within that limit, or 6.7% exceed 24ppm, and 
a few could be considerably worse. That compares to a guarantee that all 
all parts meet 20ppm. This link: 
http://www.gellerlabs.com/LM299AH-20_Case_Study.htm provided in 
Andreas's response is very interesting:

"Certified Long Term Drift The National Semiconductor LM199AH-20, 
LM299AH-20, and LM399AH-50 are ultra-stable Zener references specially 
selected from the production runs of LM199AH, LM299AH, LM399AH and 
tested to confirm a long-term stability of 20, 20, or 50 ppm per 1000 
hours, respectively..."

So in this case they really do mean a guarantee. And I doubt that 
individual testing came cheap. I say 'came' because I wonder if they 
still 100% test the 20ppm parts or if they select them using some lower 
costs means?

Tony H