OP-amp TLV2781

Farshad,

In order to answer your question, we need your actual schematic (there are three schematics on the page 6 of the link you referred to), supply voltage, input signal (magnitude and frequency), output load (resistive and/or capacitive) as well as the values of all components used.  It would be also extremely helpful you actually attached a scope picture of your "distorted' signal (picture is worth thousand words) - you may attached it by going to Options at the top of E2E forum page.

Without the above information, I can only guess that if your output becomes triangular instead of sinusoidal, you might be seeing a slew-rate induced distortion (also called Full-Power Bandwidth or Maximum Peak-to-Peak Output Voltage vs Frequency) - see graph below.

Hi Marek:

Thanks for your reply, you're right. I should have added some more details.

Now here are the details,

I used two designs, and I will explain both:

1- The design shown below is used to amplify the signal 10 times:

Input: 5.5 Tone-burst signal at 100 KHz frequency. The input signal is provided by a Signal Generator.

Output: is measured with 50 Ohm coupling of an oscilloscope (I have to measure with 50 Ohm, since the signal is going to be input to a component with 50 Ohm coupling).

Power: 2 V, supplied by a DC source.

The output gets distorted when the peak-to-peak amplitude of the input signal goes higher than 80 mVpp. For example, the output signal is measured and shown here:

2- Second design is used to amplify the signal about 3 times and is shown here:

The circuit before the amplifier is used to bias the input signal, and is working perfect.

input: Similar to previous case, is 5.5 Tone-burst signal at 100 KHz frequency provided by a signal generator.

Output: Similar to previous case, is measured with 50 Ohm coupling of an oscilloscope

Power: Similar to previous case, 2 V, supplied by a DC source

The measured output is shown here:

The output gets distorted when the peak-to-peak amplitude of the input signal goes higher than 100 mVpp

Just to mention:

The input to the op-amp is measured with 1 MOhm coupling and is perfect. So, the distortion is for sure is coming from the amplifier.

The output is not distorted when it is measured with 1 MOHm coupling.

---

As you see, the distortion level is so low. Please let me know where the problem is coming from.

Thanks,

Farshad,

First, I am not sure what do you mean when you say: "input:5.5 Tone-burst signal at 100kHz frequency;" what is 5.5, anyway?  Could you show the actual input signal with clearly mark time scale?  Your first design has a gain of -10 but it would not work the way you show it since the 47k resistor at the non-inverting input seems to be floating-in order to work, it must to be connected to DC source (e.g. GND) to bias the input common-mode below the positive rail.

All in all, I believe your main problem is that with the high-pass filter at the output, with a corner frequency of about fc=3.2kHz, you do not have enough effective bandwidth to pass the signal due to the internal output impedance of the TLV2781 interacting with your load and modifying the open-loop gain to a point where there is no loop-gain left to correct for the output signal non-linearity.  You may verify this by either removing your 50ohm output resistor or lowering the close-loop gain (shorting across 47k feedback resistor) - the output should then look as you expect.  In the second design the signal gets distorted at the higher level because the close-gain is lower = G=~5.3 (not G~3).

If you cannot lower the close-loop gain or significantly increase the size of 50ohm load resistor, you need to use faster op amp (OPA350, OPA365, OPA355).

For more details on the issues involved here, please read the series of articles addressing the underlying causes of the problem you see:

http://www.en-genius.net/site/zones/acquisitionZONE/technical_notes/acqt_050712

Hi again Marek:

Thanks a million for your time and help,

first of all, I think you noted the right pointed, and the problem is coming from the gain/load...By the way, in order to verify the results, I have included some pictures...

1) A 5.5 tone-burst signal is shown here. It has 5.5 cycles, and the frequency of the whole signal is 100 KHz, so the frequency of each cycle is 18.18 KHz. (We usually use this signal to verify our design)

2) About the load: I measured the output with 1 MOhm coupling, and the signal is perfect. But, when I use the 50 Ohm coupling, the signal gets distorted. I have input a Sin signal with 100 KHz, and showed the input/outputs here:

the input is as below: (Sine signal with 100 KHz generated by a signal generator:

Output measured with 1 Mohm coupling:

and finally, the output measured with 50 Ohm coupling:

As you see the output gets distorted when it is measured with 50 Ohm coupling. However, it is easier to observe the distortion in Tone-burst signal.

Now, do you confirm that the problem is coming from what you mentioned in your previous comment?!

P.S: The first design is working, This design is used base on a design in a TI application note...

By the way, the restriction I have is that the input voltage is less than 3 V. So, I have to use an op-amp with input voltage less than 3 V.

Thanks a lot.

Regards,