SDS-2104X+ Bode Plotting

Thread Starter

SamR

Joined Mar 19, 2019
5,031
@tautech et al
Installed all updates and plotting this
1643385644378.png
Set up the Bode Plot analysis with the AWG set for 0.01Vpp. Ran several plots using 3-20 inputs per decade and even @20/decade am getting this. 1) Does it do any averageing with multiple scans for input to the data? 2)Why the heck is the graph so jittery?

1643385903903.png
Doing it by hand in Vpp to Gain I get this.
1643386521576.png
Is there any correlation? I am not used to working in dB. Also, is there any way to invert the screen color like the CML does. Black is hard on the printers.
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,159
You are working with Voltage Gain

\( Vgain(dB)\;=\;20\times log_{10}\left(\cfrac{Vo}{Vi}\right) \)

Gain​
Vgain(dB)​
1.9​
5.58​
14.4​
23.17​
104​
40.34​
182​
45.20​
92​
39.28​
11.8​
21.44​
1.2​
1.58​
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
I do know the Gain to dB conversions, just not used to the Bode plot in dBs so don't have a feel for it yet. What I am hoping is that I can get better automated Bode plots from the scope. What I got does not resemble what I obtained doing plotting the hard way.
 

Papabravo

Joined Feb 24, 2006
21,159
I do know the Gain to dB conversions, just not used to the Bode plot in dBs so don't have a feel for it yet. What I am hoping is that I can get better automated Bode plots from the scope. What I got does not resemble what I obtained doing plotting the hard way.
Right, so,
  1. log(1) = 0 dB, so the zero dB line show where the input equals the output
  2. log(1/√2) = -3.01 dB is where the corner frequency of a filter is
  3. log(x) > 1 means there is gain happening
  4. log(x) < 1 means there is attenuation happening
At -86 dB (= 50μV) you are down in the mud and EVERYTHING looks like noise. It is often referred to as the noise floor. In LTspice you usually set the AC input source to an amplitude of 1 Volt so that whatever the output is you can take the log of it and get dB directly. In theory you could do the same thing with you generator and scope, but you know those instruments better than I do.

Do remember that when talking about power, the multiplier changes from 20 to 10.
 
Last edited:

Thread Starter

SamR

Joined Mar 19, 2019
5,031
K, I took the auto set parameter off (it stays on autoset) and started the analysis again with the grid on 0 and at first nothing. Then the grid jumped down to -87dB and again, at best, -47dB. I am using my 50Ω Palomar BNC to minigrabber lead and set the scope up for 50Ω load on the lead. I also upped the input to 0.1Vpp. I did get a bit better plot. When it finishes a sweep it automatically starts again and I'm not sure if it is doing any averaging of the gain data or maybe that is wishful thinking. I think I forgot to mention that this scope has the AWG built in and links internally to the bode plot menu or I can set up a LAN connection to an external AWG.
1643395667064.png
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
You mentioned that I had run the LTS @1Vpp input so I ran this. Actually the LTS is @0.1Vpp in. Here the "dip" is even larger and I assume that is due to it hitting the rails? Will have to investigate to confirm.
1643397329624.png
And @ 0.1Vpp input using 10:1 probe. That would be HI-Z load on the scope. Does not ask for ratio.
1643398279332.png
That didn't work... Same probe @ 1:1 HI-Z
1643398440537.png
A little better.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,281
You mentioned that I had run the LTS @1Vpp input so I ran this
That's only because the output is then normalized to 0dB for a gain of 1.
The AC simulator uses linear approximations so does not saturate and you can use any input voltage.

For the actual measurement you need to reduce the voltage to keep the output from saturating.
Tor the circuit is post #1, which has a maximum gain of 200, that would be about 5-10mv.
0.1Vpp will give a maximum output of 20Vpp, which may slew-rate limit at the higher frequencies.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
To paraphrase my buddy Pogo from the swamp down here "I have found the problem and it is I". Apparently I screwed up some channel assignments in the Bode Plot config. Got it all straightened out and...
1643416416099.png
I didn't quite go to crossover at 1Mhz so I added a decade. Now I can sleep well...
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
It's actually a rewording by Walt Kelly of a Commodore Perry battle report "We have met the enemy and they are ours".
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
And by adding a parallel capacitor across Rout it gets even better at reducing noise but sharper/narrower at max gain. It did shift the center freq a bit but would make a nice tight notch filter.
1643427724474.png
Back to the handbook...
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
@crutschow re: your bode plot on #6. I guess I was too frazzled to pay attention, but we were talking apples to oranges. Both your plot and mine were the fruit of a Bode plot but the scope Bode plot is Gain only versus your Output voltage plot. Worked a bit more with the scope and LTS and apparently the only Bode plotting the scope does is of gain. It has a ton of waveform analysis functions but, so far, I have not found a way to sweep and plot the output only. As to LTS, I can add a V(out)/V(in) gain trace but is there any way to build that into the .acs file so it doesn't have to be hand added each time the file is opened or edit a net name node for gain?
 

crutschow

Joined Mar 14, 2008
34,281
we were talking apples to oranges
Don't think so.
the scope Bode plot is Gain only versus your Output voltage plot.
As to LTS, I can add a V(out)/V(in) gain trace but is there any way to build that into the .acs file so it doesn't have to be hand added each time the file is opened or edit a net name node for gain?
The LTS plot is gain, V(out)/V(in), versus frequency in dB (20*log(Vout/Vin).
You can click on the left vertical dB values to change the display to linear or logarithmic voltage display.
Do you want something else?
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
I did find out how to get Vout instead of gain on the scope!

On this I added node points for IN and OUT but for gain I have to edit a trace for V(out)/V(in) to show gain. Yes, the V(out) show the gain over the freq but not the Vo/Vin gain? Due to the reactive nature of the RC input the Vin changes so is the "real" gain from the Vout or Vout/Vin? Obviously Vout is all you are going to get out of the op amp.
1643485202572.png
 

eetech00

Joined Jun 8, 2013
3,858
I did find out how to get Vout instead of gain on the scope!

On this I added node points for IN and OUT but for gain I have to edit a trace for V(out)/V(in) to show gain. Yes, the V(out) show the gain over the freq but not the Vo/Vin gain? Due to the reactive nature of the RC input the Vin changes so is the "real" gain from the Vout or Vout/Vin? Obviously Vout is all you are going to get out of the op amp.
View attachment 259109
Hi

I'd like to help but I'm having trouble understanding what your describing.

Are you using the word "scope" to describe something on your HW scope?
Or are you using "scope " to describe something in LTS?
The graph in LTS is called the Waveform Viewer.

LTS can really display just about anything you want displayed, but I don't have a clear picture (no pun) of what that is...
 
Top