What does it mean by 'DC Transfer Characteristic'?

[wolfrain]

Hello guys.

I am a little bit confused with this concept - 'DC Transfer Characteristic'. In fact, I am doing a simple two-stage Op Amp, and up to now all the sizes has been figured out and put them into LTSpice and simulated it.

But I have a specification called Output Voltage Range, which requires at least +- 0.8V with respect to ground while maintaining the specified gain. In addition, I have been provided with a test circuit for this spec where I have to connect the negative terminal of the Op Amp to ground and a DC source (sweep from -10mV to 10mV) to positive terminal.

Here I attached the output waveform. This jpeg shows the systematic offset. But I don't get it. Why do I need to use this configuration to measure the output in this way? What is it for? Does it mean, like if the slope keeps linear between -0.8V to +0.8V, then I can have a good Op Amp? What does it mean? And why do I need to sweep the DC source in this way? What is the DC Transfer for?

Thank you guys, I will be really appreciated for your helps here.

 

[keith1200rs]

DC transfer is a static measurement - i.e. not "dynamic". The signals are assumed to be static so the device speed doesn't affect the results. If you did the same simulation in transient analysis with a voltage ramp the results would vary depending on the speed of the ramp. This eliminates time as a variable.

In your case, the simulation you are doing is trying to find out if the output range is +/-0.8V. To truly specify it, you would need to know what gain reduction may ba allowed at the extremes, but your circuit looks fine. The slope (=gain) is pretty constant over the output range -0.8V to 0.8V and it has a good bit extra before the gain starts to drop. You have a systematic offset, but that is a separate issue.

Keith.

 

[wolfrain]

DC transfer is a static measurement - i.e. not "dynamic". The signals are assumed to be static so the device speed doesn't affect the results. If you did the same simulation in transient analysis with a voltage ramp the results would vary depending on the speed of the ramp. This eliminates time as a variable.

In your case, the simulation you are doing is trying to find out if the output range is +/-0.8V. To truly specify it, you would need to know what gain reduction may ba allowed at the extremes, but your circuit looks fine. The slope (=gain) is pretty constant over the output range -0.8V to 0.8V and it has a good bit extra before the gain starts to drop. You have a systematic offset, but that is a separate issue.

Keith.

By saying the signals are dynamic, you are talking about the high-frequency (maybe RF) signal which would effect the circuits in ways like parasitic capasitors, cross talk, etc?

So, is it also the same thing if I define the input source as 0Hz, -10mv to 10mv? If I look at the transient waveform at this time, the waveform should be amplified by the DC Gain?

Thank you,
wolfrain

 

[keith1200rs]

By saying the signals are dynamic, you are talking about the high-frequency (maybe RF) signal which would effect the circuits in ways like parasitic capasitors, cross talk, etc?

yes, and slew rate, bandwidth etc.

So, is it also the same thing if I define the input source as 0Hz, -10mv to 10mv? If I look at the transient waveform at this time, the waveform should be amplified by the DC Gain?

Thank you,
wolfrain

Well, I don't think you can do 0Hz, but yes, a low frequency transient analysis should give the same results. An example where this is used might be a Schmitt trigger where you want the switching thresholds in each direction. An easy way to see that is a slow triangle wave input. Provided you do it slow enough the dynamic "transient" and "static" DC results will be the same. If you do it too quickly the response speed of the Schmitt trigger will affect the results.

An AC analysis won't help as that is a "small signal" analysis. What you are doing is a large signal analysis.

Keith.