I don't see why this is not open loop?



And the .asc files. This should be right but the plots don't look right to me...

 

Your last two "closed loop" simulations have no negative feedback then they are open loop instead.
With an open loop gain of almost 1 million times then with the 0.5V peak input the output is clipping like crazy causing the curves to be wrong.

Even if the opamp had a gain of 101 times then with the 0.5V peak input, the output will still be clipping like crazy trying to produce 0.5V x 101=50.5V peak output and its output curve will be wrong..

 

Let's look at this...
at TA = 25°C, VS = 40 V ( ±20 V), RLOAD = 10 kΩ connected to VS / 2


And my corrected unity gain...

The Vout should be ((R1+R2)R1/)Vin = 2V = A

That's it for tonight, too tired to do any more now.

 

My previous schematic had an error (the opamp output was shorted to ground) that I fixed.

Your new opamp still has no negative feedback so it is open loop with a voltage gain of nearly one million times.
Negative feedback has a resistor or a wire from the output to the inverting input that is missing in your circuit. Maybe with a resistor from the inverting input to ground forming a voltage divider that sets the amount of voltage gain.

Maybe your simulation shows an attenuation instead of a gain because its output is clipping like crazy.
Here is the correct non-inverting closed loop unity gain opamp;

 

In post #103, a non-inverting configuration, you have positioned the resistors for a gain of 2
The gain of 2, using 20 Log (2) results in 6 dB in the passband

 

Let's look at this...
at TA = 25°C, VS = 40 V ( ±20 V), RLOAD = 10 kΩ connected to VS / 2

View attachment 258018
And my corrected unity gain...

The Vout should be ((R1+R2)R1/)Vin = 2V = A

That's it for tonight, too tired to do any more now.

Your circuit still is not correct.
AGA has given you the correct circuit in post #104.
PB has point out in post #105 that your circuit in #103 has a gain of 2.
I have already given you a unity gain example in the zip file.

Look carefully at graph figure 6-9.
G=1 (black trace) is the unity gain measurement.
Unity gain is a gain of 1 and is 0db on the graph.

The example in the zip file is posted again below. I've circled the Unity Gain circuit.

 

Ok, the title is wrong. It is in fact 2 gain. When I did a true unity feedback...

Which is odd because yours obviously worked. When I looked at the output plots in Volts, I see that the dB conversion is correct. Here is what I have now and in Volts it looks good.
2 gain

15 gain

Yes, it is clipping...

Open loop

And yes, it is slamming the rails. So the models are working even though the plots in dB looked odd to me because I'm not used to working in dB.

 

Here's a simple op amp open-loop AC gain simulation to give the Bode plot:
The resistor provides DC feedback so the op amp DC output equals the DC input (in this case 0V), while the 1kF capacitor shunts the AC signal feedback to ground so the AC gain is below the open-loop to a very low frequency.

 

Thanks Carl! I think that was the last piece of the puzzle that I needed. Now to do some breadboarding and looking at it on the scope with the sweep mode of my sig gen.


Rolls off pretty quick @ max gain. And for giggles 0.001V@1kHz input.

 

hi Sam,
This LTS link is worth a few minutes of your viewing time.
E
https://www.google.com/search?clien...pa+amp+test#kpvalbx=_xN_mYfH0ENHGgQaTnq6wDw22

 

I haven't gotten to the phase part yet but watched it and marked for later also. Thx Eric!

 

Now to do some breadboarding and looking at it on the scope with the sweep mode of my sig gen.

You will need a more complex circuit for a real device open-loop gain as the RC time-constant for the simple circuit in the simulation is too large to be practical.
Figure 5 in this reference, is one such circuit.

 

Good article. Not quite as simple as I thought it would be. I'll read it in more detail later instead of just scanning it. Also, for the sweep I'll need to figure out how to look at it on my DSO. Kinda new area for me to explore as almost all has been in real time so far. Thx Carl!

 

Hadn't considered this although I've heard it from AG many times concerning breadboards. Anyway what I am really looking at now is the effect of peripheral components on gain so I'll have to play within the limits of my sandbox. I shall see...

 

Also, for the sweep I'll need to figure out how to look at it on my DSO

You would synchronize the DSO sweep to the sweep speed of the generator.
But to see the open-loop plot similar to the simulation, you would need a log amp and precision rectifier for the output to convert the AC output to a log DC value.
They log amp can be built using diodes, or use an IC log amp such as Analog Devices builds.

Regarding the caveat about frequencies above 100kHz for the Analog Devices test circuit , I would think reducing the values of R1, R2, R3, and R9 each by a factor of 10 or 100, would increase the upper frequency limit.

 

Why do you simulate the frequency response of an open-loop opamp when its graph is shown on its datasheet??

 

Why do you simulate the frequency response of an open-loop opamp when its graph is shown on its datasheet??

So you always believe everything the data sheet states?

 

Why do you simulate the frequency response of an open-loop opamp when its graph is shown on its datasheet??

hi agu,
It's a simple way of cross-checking your own simulations, whilst learning LTspice.
E

 

So you always believe everything the data sheet states?

I do not buy cheap no-name-brand Ch***** junk.
The Simulator also does not believe the datasheet, showing no problems when the opamp max allowed supply voltage is exceeded.

 

Why do you simulate the frequency response of an open-loop opamp when its graph is shown on its datasheet??

To locate the low frequency pole, which may or may not be specified in the datasheet.