This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

OPA541: Distortion at the bottom of output waveform

Part Number: OPA541

Hi team,

Hi team,I got a question from customer.

OPA541.docx

Thank you very much for your help.

Best regards,

  • Hello Zhonghui,

    Please provide the schematic for the OPA541 circuit. The report says "the circuit mode is simple and forward input, the ratio is 1 to 4." Does that mean it has a gain of 4 V/V? 

    Additionally, I am having difficulty determining the output voltage swing from the O-scope images. What are the volts/div settings for the O-scope? If the output is only swinging a few hundred millivolts it may be crossover distortion. Is the output waveform initially undistorted, and then becomes distorted as the OPA541 heats? I see the heatsink in the images. Is there a theta, heat-sink-to ambient rating for it?

    Thank you, Thomas

    Precision Amplifiers Applications Engineering 

  • Hi Thomas,

    Thank you for your reply.

    ①Circuit diagram:

    ②The voltage/div setting of oscilloscope, each grid of y-axis is 100mV

    ③Whether it is a fever problem:

    A: the distortion waveform exists at the beginning. When the input voltage is 100mV, the amplitude attenuation is caused by the resistance voltage in front of opa541. After the four times gain amplification of opa541, the peak value of the output is 278 MV, and the current is not large. The temperature of the device is room temperature by hand. It is normal, and there is no high temperature phenomenon. When the frequency is adjusted to 10Hz, the distortion will not be obvious. The larger the frequency is, the more obvious the distortion is.

    Thank you!

    Best regards,

  • Hello Zhonghui,

    Thank you for the very complete information about the OPA541 application and the details about the distortion, and how it improved with lower frequency.

    All of this information supports my original thought that the distortion is what is commonly known as crossover distortion. The OPA541 uses a class AB output stage to conserve operating current when it is not driving a heavy current load. The transistors in the output stage nearly turn off when the output voltage is near 0 V. What is actually being observed is the output transistor base-emitter voltage characteristics as one output transistor hands off the load current to the other output transistor. One sinks current and the other sources current. The region where the transistors are nearly turned off is highly non-linear and that is where the distortion comes from. 

    Normally applied negative feedback reduces crossover distortion. Therefore, the lower the closed-loop gain (Acl) the circuit operates with, the lower the crossover distortion will be at a particular frequency. Because the op amp's open-loop gain (Aol) increases as frequency is lowered, the loop gain (Aol - Acl) increases and the distortion decreases with it. That is why you are observing a reduction in the distortion as the frequency becomes lower.

    A gain of + 4 V/V is not an unreasonable Acl for the OPA541 to be operated at with a 1 kHz input frequency. Unfortunately, crossover distortion is an inherent characteristic of a power op amp class-AB output stage such as that used by the OPA541. There isn't anything we can do to the op amp itself to eliminate this type of distortion. However, there may be some things that would decrease it somewhat:

    1. The amplifier circuit is operating with an Acl of +4 V/V, but it has the R31, R32 voltage divider before the non-inverting input. The overall gain when the voltage divider is included is +2.8 V/V. If the input voltage divider can be eliminated, the OPA541 could be operated with an overall Acl of +2.8 V/V. This lower Acl would increase the loop-gain under which the OPA541operates, and that should help lower the crossover distortion. The circuit changes would involve removing the 700 Ohm divider resistor R32, and changing the feedback resistor R29 from 3 k, to 1.8 k. I do not know how much this will reduce the distortion, but it should be noticeable on the O-scope.
    2.  Another technique to further reduce the distortion is to operate the OPA541 in a composite amplifier circuit. This requires adding an additional op amp into the circuit such that you get a multiplication of the Aol of each op amp. The overall Aol product is much higher than that of the OPA541 alone. Doing this would require a design effort and a complete re-layout of the PC board. Composite amplifiers are sometimes difficult to stabilize and compensation is often required. If this arrangement is applied the loop gain would be very high and that should have a significant impact on the crossover distortion.

    I hope this information helps.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas,

    Thanks a lot.

    It is useful.

    Best regards,

  • Hi Thomas,

    In your reply, it is mentioned that " The OPA541 uses a class AB output stage to conserve operating current when it is not driving a heavy current load.". I have two questions:

    1. what is the large load it can be called driving high current load (0.5Ω?).

    2.if it meets the high current load, that is, the load resistance is less than 2Ω in the experiment, will its output stage still be class AB, that is, will the distortion change?

    In the second scheme, I have two questions:

    1. what is the function and circuit form of the additional op amp combined with opa541?

    2.what is the impact on distortion, so that the distortion is very low?

    Thanks a lot.

    Best regards,

  • Hi Zhonghui,

    Regarding your OPA541 questions:

    1. what is the large load it can be called driving high current load (0.5Ω?).

    I suppose the term "high current" is relative. Certainly when the output current is in the hundreds of milliamperes or Amperes, we refer to them as high current op amps. This distinguishes them from the more precision op amps where the output current is typically 50 mA and less.

    Do note that the power op amps such as the OPA541 are most often applied with a wide output voltage swing. Many times the swing approaches their maximum in relation to the supply rails. In such cases the crossover region is a smaller portion of the output range than it is in your application where the output is only swinging about +/-400 mV.

    Regarding the OPA541 a 2 Ohm, or 0.5 Ohm load is a high current load if the output voltage is such that a high current flows. For the OPA541 that would be Amperes of current.

    2.if it meets the high current load, that is, the load resistance is less than 2Ω in the experiment, will its output stage still be class AB, that is, will the distortion change?

    Most often for class-AB output stages the crossover distortion increases with an increase in load current. That is because the change in transistor current that when operating within the crossover region to when it is driving an increasingly higher current is a more radical change compared to when the load current is much lower.

    In the second scheme, I have two questions:

    1. what is the function and circuit form of the additional op amp combined with opa541?

    I don't have a great example of a composite amplifier configuration using a power op amp, but there are some TI applications resources that show the concept. These should provide you with general information:

    https://www.ti.com/lit/an/sboa002/sboa002.pdf

    https://www.ti.com/lit/an/sboa357/sboa357.pdf

    I suggest doing a Google search on "Composite Amplifier Op Amp" and much more information is available online. 

    2.what is the impact on distortion, so that the distortion is very low?

    The distortion improvement will depend on how much additional loop gain is added by the added op amp to the composite amplifier. Distortion is improvement would have to observed because there a numerous secondary factors affecting the end result.

    There is one other technique that is sometimes applied to op amps to reduce crossover distortion, but it is at the expense of increased power dissipation in the circuit. The technique forces the op amp output stage to operate class-A and eliminates crossover distortion. The basic idea is to add a load to the op amp output that forces it to continually sink, or source current. I have seen this applied to low power op amps, and should be applicable to high power op amps but in the latter case the power dissipation might be a problem. 

    I found an online article that discusses moving an op amp's output to class-A operation. I think it discusses the concept well:

    https://tangentsoft.net/audio/opamp-bias.html#

    I expect this could be accomplished by adding a resistor from the OPA541 output pin to the VEE supply. A 150 Ohm (2 W min) resistor would add about 100 mA current load to the OPA541 when the output is +400 mV peak. You might want to give this a try and see if the crossover distortion improves for you circuit. 

    If the output voltage is made higher (more positive) the resistor current and power will increase in accordance with Ohm's law and the Power law so keep that in mind.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas,

    I have learned from your reply. Thank you very much.

    There is one last question: because I have used a lot of high current operational amplifiers, which cost hundreds to thousands. It seems that this kind of situation is very common. I would like to ask you, as a high current operational amplifier (max 3, 4A or above), to drive a strong load and output a continuous high current, due to the operational amplifier's own reasons, will cross distortion exist?

    Thank you again.

    Best regards,

  • Hello Zhonghui,

    It is highly likely that almost any integrated or discrete power op amp, and even linear audio amplifiers, are being operated with a class-AB output stage. Therefore, it is probable that you will find that most power amplifiers you evaluate will have crossover distortion to some degree; some better, or worse than others.

    One factor affecting how much crossover distortion a class-AB amplifier produces is the amount of current through the output transistors are in a quiescent (at rest) state as compared to when it is driving an output load. Typically, an op amp amplifier that has high quiescent current in relation to output drive current will produce less crossover distortion compared to one that has low quiescent current for the same output current drive level. That is because the high quiescent current op amp is "burning" more current when it is at rest and the output stage is running closer to true class-A.

    Even though it may not be apparent to the user nearly all op amps employ a class-AB output stage. It may not be as evident with a precision low-power op amp as they have very high open-loop gain (Aol), which reduces the distortion. I do recall years ago that Burr-Brown, acquired by TI in 2000, did produce a unique class-A output op amp that was produced for a specific customer who's graphics application was extremely sensitive to any distortion. The application demanded a class-A output op amp. As expected, that op amp burned a lot of current in the output stage and the devices ran very hot.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas,

    You've answered so well.

    Thank you.

    Best regards,