OPA347

Thread Starter

Kingcosmos

Joined Sep 23, 2004
5
I am trying to simulate the OPA347 operational amplifier in TINA-TI. There is a basic test circuit in their examples and I am trying to simulate it in order to understand some of the parameters.

It is nothing more than your basic non-inverting configuration utilizing the OPA347. You can see the attached schematic.

An overview:

I have it set up for a gain of 2, 1k load, single supply +5V. My input is 0V DC to start and I expect (ideally) to get 0V out; however, I get 1.854 mV out. The datasheet specifies 2mV of offset typical. There is not minimum and I am content with this result for the most part. The datasheet also says that the output swing voltage can *typically* be 5mV from the rails again there is no minimum specified so that is why I am fairly content with a 1.854 mV output with 0V in.

Now, when I input 1V I expect to see 2.001854 V. 2V because of the gain plus the "original" offset but this is my actual result: 1.99986 (-140 uV). If I input 1.5v I expect to see 3.0001854 V. Again 3V due to the gain and plus the "original" offset but again, this is my result: 3.000008V (8 uV).

I am aware that if you change the gain, your offset will also increase but even with an increasing gain my results do not show the typical 2 mV of offset. It is all measured in uV!

What am I missing? Why is it that when I increase my input voltage my "offset" is in the uV? Why is it no where near what is specified in my datasheet? Any insight would be great.
 

Ron H

Joined Apr 14, 2005
7,063
The offset at zero is probably caused by the inability to reach the rail, not by offset. I don't understand why you get negative offset with 1V in, and positive offset with 2V in.
You might expect a tiny gain error.
I doubt that TI is modeling nonlinearity, which could account for the discrepancies. I took their model, and ran your configuration (I used 10k resistors) in Linear Technology's SwitcherCAD III, and I got 1.837mV output with 0V in, 2.00004V (40uV error) with 1V in, and 3.00001V (10uV error) with 1.5V in (SwitcherCAD displays 5 significant digits). These numbers can be explained by offset voltage error in combination with gain error. I don't think your numbers are so easily explained.
I just had a flash. If TINA-TI can do it, Plot v(out) - 2 * v(in). The results of my sim are below.
[attachmentid=828]
 
Top