LT1167 LTSPICE noise simulation

Hi Ali,

When I simulate your circuit with noiseless resistors and a gain of ~ 10V/V as you've shown, I get a flatband noise voltage simulation of ~100nV/RtHz at the output which I think is correct. You also have to consider the output section noise voltage density in your estimate:

Vout_noise = RMS (7nV/RtHz * 10V/V , 67nV/RtHz) = 97nV/RtHz

where 7nV/Hz is "eni" and 67nV/RtHz is "eno".

However, I see an issue with the datasheet which may have confused you as well. The datasheet shows "Note 8" for "eno" (output voltage density) and Note 8 states "Note 8: "Referred to input." which is not correct.

The contribution of the output stage noise to the overall noise measured at the output is independent of the In-Amp gain setting (RG value).

Therefore, "Note 8" designation for "eno" should be eliminated since its contribution to output noise is direct (no gain relationship).

I'll work on getting the datasheet updated to eliminate Note 8 for "eno".

Hope this makes sense.

The simulation shows a flatband noise of 100nV/RtHz which is pretty close.

Here is the simulation file for reference:

LT1167 Noise EZ 1_27_21.asc

Regards,

Hooman

Dear Hooman

Thank you so much for the answer. It's very helpful. I am able to match my results.

If you don't mind, I have another question, about the datasheet. I am confused by the attached figure of voltage noise density vs frequency plot. Here at 1k Hz and for a gain of 10, we have a noise density value of around 20 nV/RtHz.  Is it possible to relate this plot and the result we have from the LTSPICE simulation?  My apologies if I am misunderstanding this plot.

Thank you again for your help 

Hi Ali,

You're welcome!

Regarding the voltage noise density plot: If your feedback resistor is changed to a "real" (not noiseless) 5.49k resistor in your LTspice simulation, your output noise density will be around 132nV/RtHz. Reflected back to the input (divide by 10V/V gain), that's 13.2nV/RtHz. The plot you're referring to is showing somewhere below 20nV/RtHz which is the right ballpark.

Regards,

Hooman

Hello Hooman.  Thank you so much. These answers were very useful.

Have a nice day.

Hello Hooman.  Thank you so much. These answers were very useful.

Have a nice day.