Re: Noise: *Megavolts* in for picovolts out??

Helmut Sennewald wrote:

"Robert Baer" <robertbaer@xxxxxxxxxxxxx> schrieb im Newsbeitrag news:wJG5e.2947$An2.181@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Helmut Sennewald wrote:


"Robert Baer" <robertbaer@xxxxxxxxxxxxx> schrieb im Newsbeitrag news:erX4e.1846$An2.1439@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx


???WHAT??? 


Hello Robert,

You shouldn't cry "wolf" if there is no wolf.

Your circuit has wrong DC-operating points for many
of your amplifiers. They have saturated outputs.
So it's no wonder that the results look strange.
Before one tries .AC or .NOISE, it's mandatory to run
a .OP or .DC to see that no output is limited.

After I have fixed your design, the results are as expected.

Best Regards,
Helmut



------------ SNIPped for brevity ---------
Sorry about that; the bridge was connected incorrectly.
I fixed that and get about 230mV in for 300uV out; a lot better but still rather puzzling.
Here is the fix:
Version 4
*** 1 1220 708
WIRE -256 128 -256 96 



Hello Robert,

Why is there no analysis command in your file?

What are the 230mV?   Did you want type 230uV?

By the way, the noise density at the ouput V(N016) is about
230uV at 1/(2pi)Hz.

Your circuit looks really useless after your latest change.
I talk about the connection of resistor RTAA to the input source.
Now the circuit amplifies only a very small fraction of the input.

Still LTspice simulates it correctly. I never believe that
you simulate with LTspice. You would get correct results
with LTspice as I have got.

What simulator do you use?

Best Regards,
Helmut


1) The file is from LTspice, and Simulate/Run gives a list of points in a window. Selecting inoise and/or onoise then results in a graph covering the frequency range specified in the .noise card.
2) Hope you got the whole file; i see only a few lines above.
3) 40 percent of the bridge drive is not a "very small fraction".
4) If you look carefully, you will note that the bridge is rather close to balance; a large input signal would thus be attenuated by the bridge. However, the SENSitivity is at max, near a gain of 21 and the last stage has a gain of about 5 for an approximate total gain of 105.
I presume that you have guessed that TA (RtaA & RtaB) is a pot with an adjustable wiper. I did not find any way to put in a Bourns pot of a given description (one would like a PCB layout to conform to parts).
**
In simulations for DC op points, i have been able to run the bridge balanced from ground to the 1.91V max and get satisfying results.
Granted that at (or near) ground, the opamps act a little fussy, but that is to be expected.
.

Quantcast