From: Fred Bloggs
User-Agent: Mozilla/5.0 (Windows; U; Win 9x 4.90; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0
X-Accept-Language: en-us, en
Subject: Re: accurate fast opamps
References: <email@example.com> <firstname.lastname@example.org>
Date: Wed, 13 Nov 2002 15:17:58 GMT
NNTP-Posting-Date: Wed, 13 Nov 2002 07:17:58 PST
Organization: EarthLink Inc. -- http://www.EarthLink.net
Winfield Hill wrote:
> Tony wrote...
>>In article ,
>> Winfield Hill wrote:
>>> Another nice one, but with a price to pay for the extra
>>> speed: increased bias current. Although they say 15nA
>>> typical they give themselves a 90nA budget. Working
>>> backwards from the current-noise figure, it looks like
>>> the uncancelled input-pair base current is about 2uA.
>>> Still, I wish I had some of those in my inventory now.
>>One thing I like about these TI Excalibur series of
>>opamps is the quality of the data sheets. Ok, the
>>bullet-points are as optimised as ever, but there is
>>a wealth of data in the detailed tables and graphs,
>>even statistical data on some critical parameters.
>>eg, on the input bias-current question; Fig11 shows
>>that 90% of devices tested were between about 4nA
>>and 25nA at 25C.
> Figure 10, yes very nice. Hmm, what is one to do with
> 90% bars? T.I. points out it means that 5% of the 836-
> sample tested lot (42 parts) were above 25nA. What we
> want to know is how many were above say 50nA? I.e., to
> see if we could have safely ignored the 90nA maximum
> spec, at least for that set of parts. :-)
Any grade school child could tell you that this is a simple normality
test on grounds of central limit theorem so that you use the "bars" to
estimate mean and variance which is used to index into a zero mean unit
variance table to tell you the minuscule percentage above those limits.