Simulation for Ro shows around 6.4 Ohms (Ignore V scale in simulation chart)

Simulation files shows the above result is correct and does not agree with data sheet chart.

Simulation for Ro shows around 6.4 Ohms (Ignore V scale in simulation chart)

Simulation files shows the above result is correct and does not agree with data sheet chart.

Hi Jinol,

That's interesting, but can you tell me why the simulation models you

provided are correct and the ones I've been using are not? The simulation

circuits are ones I've used while in my analog design classes and I also

see them repeated in articles online. For example, here is a circuit from a

TI article slyt677 "Modeling the output impedance of a op-amp for

stability analysis" which is the one I used for my simulation. I would

really like to be educated on the differences in your methods.

Also for the closed loop gain the simulation results show very little gain

error versus frequency. If I use the charts in the data sheet for open loop

gain at my frequency of interest I get the following results when I use a

gain of 2.

I'm not very comfortable with these differences and would really like to

understand the reasons why. Thank you very much for the help.

Regards,

Bruce

Hi Bruce,

That's the latest version of the SPICE model which I also used. As for the closed-loop output impedance, I believe your test circuit if for Open-Loop output impedance which sadly was not modeled. For closed-loop output impedance, use the test circuit shown on the previous reply. As for your concern regarding the variation with our test circuit and yours, see the link for a document about SPICE Model testing. We are using the "universal" test circuits, same with TI.

Pspice Universal Test Circuits

Regards,

JinoHi Jino,

I now can get my simulations to line up with the data sheet, but the closed

loop output impedance that you showed only works if the AC source on the

non-inverting input is set to 0 volts. If I set it to 1 as show in the

document you provided the lowest value of output impedance is 1 Ohm. So I'm

still puzzled as to how to determine the output resistance for, in my case,

a gain of 2. Do I set the input voltage for the simulation to 0 volts as

you did or 1 as shown in the spice model document?

Regards,

Bruce

Hi Bruce,

Setting the input to 1 at gain of 1, gives you an output of 1. With output current set to 1, then we are forcing the output impedance to 1ohm from DC to certain frequency where the AC components are starting to take effect. On figure 49 of the datasheet, that was measured at AC 0.

Regards,

JinoHi Jinol,

Ok, but why does the simulation article you provided use 1 volt for the

input? Would that mean the information they provided is incorrect? Can I

determine what output impedance I would have for various gains using the

gain of 1 chart? If not what should I use for resistor values and input

voltage to see the output impedance at a gain of 2?

Sorry for all the questions, but I just want to make sure I understand

this.

Regards,

Bruce

Hi Bruce,

Both test circuits are correct. It only depends on the methodology we use. We can also use the same method with TI. It just happened that we used a different one. I used my test circuit (AC 0) to see if I can get the same graph/response as the figure 49 of AD8045 datasheet. And to further help you picture out the variations on the methodology of testing, see figure 34 of LMV791. You will notice that it goes down to 0.02 as compared to the SPICE document. Because theoritically, the Output Impedance of an ideal op-amp is 0 ohms.

That would depend on your practical application. AD8045 suggested RF is 499 ohms. Just remember that the higher the resistor values are, the faster would the bandwidth roll-off. Just go with 499. Also extend the horizontal scale for the closed-loop output impedance curve, you will notice that it will flatten out to a certain frequency and that frequency is the dominant pole.

Regards.

Hi,

Could you provide your test circuits please?

Thanks and Regards,

Jino