Hi everyone,

I'm trying to simulate the behavior of TI's OPA2674ID on LTSpice, here's the link to the datasheet for your convenience:

https://www.ti.com/lit/gpn/opa2674

The following circuit diagram shown below is given in the datasheet and I have modeled my simulation's schematic after it.



Here's a cropped version of my simulation's schematic:



RL, VS, and RF were chosen based on the information below:



RG was obtained from the formula given in the first attached image while CG was calculated using the cutoff frequency formula (f = 1/(2*pi*CG*RG).

According to the datasheet, I should be obtaining a differential gain of +9 but when I ran my simulation and checked for the voltages across RLoad and across the non-inverting input, the voltage output is lesser than 5V. Shouldn't the output voltage be scaled up instead of down? How would I scale up the voltage and the current as well to obtain a power of about 17W across the load resistor? The source voltage is meant to represent the AD9834 chip by the way.

Here's a snip of the simulation's results:



I'd appreciate any help and criticism I could get regarding my question. Thank you in advance!

 

Hi lei,
Welcome to AAC.
Please post your LTS asc file.
E

 

Hello,

What is the input voltage?
When it is 5 Volts, it will be to high.
Try 0.5 Volts and see what happens.

Bertus

 

Thanks for your replies Eric and Bertus!

When I set the input to 0.5 Volts, here's the results:



That's quite weird seeing how the voltage across non-inverting input is about 2.75 Volts when it should be 0.5 Volts.

I have also attached the LTS asc file.

 

hi lei,
You have set the GBW to 10megaHz and are using a 13.56megaHz signal.

Where did you get the GBW value?

E


Update: increased GBW to 100megaHz

 

Hi Eric,

The GBW value was a pre-set value in the universal Op-Amp model in LTSpice if I'm not mistaken. I did not even know about it. Should I set it to match 13.56 MHz?

L

 

hi lei,
You have set the GBW to 10megaHz and are using a 13.56megaHz signal.

Where did you get the GBW value?

E


Update: increased GBW to 100megaHz
View attachment 299089View attachment 299090

Hey Eric,

Thanks a bunch! How do you determine the optimal value for the GBW?

L

 

hi,
You should set the GBW as shown in the datasheet to suit the OPA's operating range.

E

 

The input signal source is shown as floating.
It should have a DC path to ground for the input bias current.
Try 10kΩ to start.

 

Interestingly, the circuit in the post #4 asc file gives the same output voltage across Rload with or without the Universal opamp power supply terminals being connected to anything and with or without the DC path referred to in post #9.

 

Interestingly, the circuit in the post #4 asc file gives the same output voltage across Rload with or without the Universal opamp power supply terminals being connected to anything and with or without the DC path referred to in post #9.

Yes.
It would appear the op amp model is allowing the input to go more or less directly to the output, even when it is not powered.
Seems like a glitch in the model.

 

hi,
You should set the GBW as shown in the datasheet to suit the OPA's operating range.

E
View attachment 299091

This is the output I obtained when I set the GBW of the OPA to 100 MHz, why is the voltage across the non-inverting input not begin at 0 V and peak at 0.5 V? And why is the lower peak of the voltage across the load resistor not at ~-4.4 V?

 

The input signal source is shown as floating.
It should have a DC path to ground for the input bias current.
Try 10kΩ to start.

Sorry, do you mean to set the parasitic properties of my source's series resistance to 10 kΩ? I have done that and the results appear to be more or less the same. Apologies if I'm missing something.

 

hi lei,
Using current path resistors, as per this diagram, is a common way.
E