[volt-nuts] Fluke 335A

[WarrenS]

"NOISE data" without the upper and lower Bandwidth used, is a near 
meaningless number.
Doing a ADEV plot of the voltage noise is one nice way to include the 
Bandwidth with the data.

ws

****************

I measured a 731 with a scope to see the noise. It was around 25 uVpp. I 
think that is the most easy way. You see how much disconnected and how much 
its connected. I úse a tektronix 547 with E-plugin for this. ( a lot cheaper 
then a 7,5 digit or more multimeter ;-)

Fred PA4TIM
*****************

Op 27 sep. 2011 om 05:04 heeft "gbusg" <gbusg at comcast.net> het volgende 
geschreven:

> Hi Joe,
>
> My calculations agree with yours, as far as Fluke specs for your two
> voltages are concerned.
>
> Fluke spec'd Ripple and Noise (all frequencies) as <20 uV rms on the 10V
> range. So that would be another 2ppm rms adder for 10Vdc setting, 20ppm 
> rms
> adder for 1Vdc setting and 200ppm rms adder for 100mVdc setting. But I'm 
> not
> immediately sure how that rms noise spec might translate to observed
> racking-around on your 3478A (depending on integration time setting on 
> your
> 3478A, etc.).
>
> Fluke's basic 10 uV adder (for 10Vdc range) translates to 1ppm adder at
> 10Vdc setting, 10ppm adder at 1Vdc setting and 100ppm adder at 100mVdc
> setting. The way I look at it, that 10 uV adder accounts for some 
> short-term
> instability (which might also show up as some racking around on your 
> 3478A)
> plus DC zero offset (as observed on your 3478A).
>
> Anyway, obviously the 335D "likes" full (or near full) scale the best.
>
> You could test your 3478A short-term noise (racking around of readings) by
> using it to read a source that's known to be very quiet. Or at least you
> could build a 10 to 1 or 100 to 1 resistive divider between 335D and your
> 3478A to see if that reduces the observed noise.
>
> Best,
> Greg
>