[time-nuts] Totally unrelated, but..

David davidwhess at gmail.com
Thu Dec 8 18:07:25 EST 2016


On Thu, 8 Dec 2016 02:25:14 +0100, you wrote:

>Am 08.12.2016 um 01:20 schrieb Scott Stobbe:
>> Yes, the short hand I like to use is 4 nV*sqrt(R/1000).
>>
>> 2 nV/rthz off a bandgap is pretty darn impressive, that includes a delta
>> vbe gained up ~10x.
>
>Methinks the advantage comes from converting their reference (whatever 
>that may be)
>to a really high impedance current source where a few uF help 
>tremendously in cleaning things up.
>
>regards, Gerhard

The output stage of the LT3042 also operates with unity gain so the
reference noise is not multiplied by the gain set by the divider and
bandwidth of the error amplifier is maximized.  Some regulators
include an extra pin or the voltage divider is external so that point
can be bypassed but their reference is still in series with the error
amplifier so this does not suppress noise from the bandgap reference
itself.


More information about the time-nuts mailing list