Author Topic: EEVblog #528 - Opamp Input Noise Voltage Tutorial  (Read 14481 times)

0 Members and 1 Guest are viewing this topic.

Offline iTist

  • Contributor
  • Posts: 12
  • Country: de
  • iTist
    • coding music diy
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #25 on: October 09, 2013, 08:23:43 pm »
Thanks Dave great explanation.

i build up the Linear Technologies Circuits for the Measurement of  0.1Hz to 10Hz Noise.:

Here some Pictures:

In the Box


OP07 noise 0.1-10Hz  100nV/DIV



LF356 0.1-10Hz 100nV/DIV


Greetz


Visit my Website:
www.itist.de

Offline martinv

  • Contributor
  • Posts: 39
  • Country: us
    • mvforum
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #26 on: December 26, 2019, 03:54:36 pm »
Repeating the experiments in this video, I notice the 1/f noise corner frequency seems to move to a lower and lower frequency as you reduce the span on the instrument.
It is acting more like a spectrum analyzer as you reduce resolution bandwidth the 0Hz peak moves closer and closer to 0.  Using span of 12 Hz moves the corner to 0.2 Hz.  This is repeating the units self noise with a 50 ohm terminator.
Shouldn't the 1/f noise be constant using the PSD (Power Spectral Density)mode?

(side question is if this comment should have been a new topic due to this "Warning" message: 
Quote
Warning: this topic has not been posted in for at least 120 days.
Unless you're sure you want to reply, please consider starting a new topic.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11704
  • Country: us
  • DavidH
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #27 on: December 27, 2019, 12:35:45 am »
What instrument are you referring to?  RBW (resolution bandwidth) changes with FFT bin width and most DSOs do not include normalization to support noise measurements so you have to calculate it yourself.  Also with a DSO FFT, make sure to average *after* the FFT and not before.  Some only allow averaging before making them useless for noise measurements.

A spectrum analyzer with a noise marker function should return consistent results within the limits of its filtering.
 

Offline martinv

  • Contributor
  • Posts: 39
  • Country: us
    • mvforum
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #28 on: December 27, 2019, 01:14:28 am »
HP35660A  just like in the video.  I repeated the same experiment using the 50 ohm terminator and got the same results, but after reduce the span I noticed the cutoff frequency of the 1/f noise also reduced. I understand swept spectrum analyzers much more than this 35660A, but am trying to learn...  It appears the 1/f noise measurement is responding like resolution bandwidth is changing, but isn't the PSD measurement supposed to use a 1Hz bandwidth?  There is no RBW setting on this instrument.  Example of 100Hz and 1.6kHz span below. Note the corner frequency seems to move from about 20 Hz to less than 1 Hz even though no settings changed other than span and re-acquiring an average of 10 samples:
« Last Edit: December 27, 2019, 04:21:40 am by martinv »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11704
  • Country: us
  • DavidH
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #29 on: December 27, 2019, 02:45:09 am »
To me it looks like lack of frequency resolution on the wider span measurement and not enough averages on the lower span measurement are distorting the results.
 

Offline martinv

  • Contributor
  • Posts: 39
  • Country: us
    • mvforum
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #30 on: December 27, 2019, 04:27:22 am »
You may be correct,  I ran a 6Hz span, 20 averages (took a while) and compared the marker value at 1.25 Hz (see additional image I added to previous post).  It is reasonably close.  Basically on this instrument it seems best to not rely on the left 1/3rd of the screen for PSD measurements if your frequency scale is in log mode as there are only about 5 samples. 
« Last Edit: December 27, 2019, 04:29:20 am by martinv »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 11704
  • Country: us
  • DavidH
Re: EEVblog #528 - Opamp Input Noise Voltage Tutorial
« Reply #31 on: December 27, 2019, 04:37:21 pm »
The manual mentions that measurements near DC can be affected by DC offset from the input circuits so it may be necessary to ignore the first few FFT bins when making baseband measurements.  It is not entirely clear from the display how many FFT bins close to DC are not being displayed.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf