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.

DRV2700: How to reduce the output ripple of the flyback circuit

Part Number: DRV2700

Hi Team,

Original question post: https://e2e.ti.com/support/motor-drivers/f/38/t/909339

Customer did the test. After reducing the ripple, he found that after inputting a constant PWM, the voltage of the capacitive load could not be stabilized within the value set by himself.The specific circuit is in the original post. Is there a problem in calculating the relationship between PWM and output voltage? What other parameters should be considered? (The formula is derived from several formulas in the manual)

Thanks,

Annie

  • Hello Annie,

    I am looking into this problem in more detail, but may I ask what is the PWM sampling frequency the customer is using?

    Best,

    Carson Allen

    LPAA Applications Team

  • Hi Carson,

    Input frequency is 3.3KHZ, allow 1/3, simulation time 50ms

    Thanks,

    Annie

  • Hey Anne,

    Do you mind having the customer share the spice model/simulation with me and what they are looking for at the output(150V DC?). The exact output voltage can be hard to match to the formula in the datasheet as it also depends on the PWM frequency and the capacitor values across the circuit, especially since they are using such high capacitance at the output to fix the ripple. So I would need to play around with the circuit a little to determine an exact solution to the customers problem.

    Otherwise if the customer is just trying to output a DC I would try a higher frequency PWM (>20kHz) input if possible so that the voltage output of the filter would come out as DC instead of the way it is now.

    Best,

    Carson

  • 20200706start.TSC

    Hi Carson,

    I'm a student who asked Engineer Annie before.I want to use DRV2700 to make a PZT power supply driver, the power range is 0~150V, and the voltage ripple should be as small as possible.Attached is my current simulation file.At present, the problem is that under the condition of constant PWM, my simulation cannot obtain stable output voltage value.At present, I have been trying to find out the relationship between input PWM and output voltage.In this way, I can output PWM with MCU to get the voltage I want, and the ripple is also very small.

    Thanks,

    weijian

  • Thanks Weijian,

    I will take a look at this model, in the meanwhile I take a look at a couple things to help understand the relationship between PWM and output voltage.

    First read section 9.2.1.2.9 in the datasheet

    As well as this

    Taken from page 15 of the following EVM user guide:

    https://www.ti.com/lit/ug/slou407a/slou407a.pdf?ts=1594222809929

  • Hello Again Weijian,

    Using the model you provided I was able to adjust and simulate different combinations of the PWM frequency and output capacitance. For the 3.3kHz you have been using,I would lessen the capacitance at the output to the point where it still eliminates the unnecessary amount of ripple but takes less time for the voltage to stabilize in the simulation. 

    With the current output capacitance it takes over 500ms for the voltage to stabilize and based on the simulations I did, I got rid of both of those 10uF capacitors and instead used a single 2uF capacitor with no resistor and it seemed to stabilized at 100ms at around 145V with minimal ripple.

    Best,

    Carson

  • Hi Carson,

    I also adjusted my model according to the method you said, and I did get the result you said.The final result is a 140V output voltage with a small ripple, as shown in the figure below.However, the capacitance of the output load capacitance (C2) cannot be changed, because the actual piezoelectric ceramic used has an electrostatic capacity of 0.8uF and a power of less than 5W.

    Before changing the model, the stable value of 140V can be obtained by using two 10uF capacitors under the condition of constant PWM and 10 times (333HZ) lower frequency.The problem is that when I was looking for the relationship between PWM and output voltage,I could not find a formula to calculate the output voltage,as shown in the figure below,from <slou407a> specification.

    The formula marked in red cannot accurately calculate the output voltage.The calculation process and formula are very important for my future work.If there is no other way,Now I want to use MATLAB software to build the circuit model and calculate the relationship between PWM and output capacitance.There is no doubt that the work you did before has been a great help to me.Thank you very much! So do you have any better Suggestions or methods?Looking forward to your reply.

    Thanks,

    Weijian

  • Hey Weijian,

    Unfortunately there is no current equations that are evident in relating the output voltage to PWM, but I will ask my superiors for more information on the flyback configuration that has not been publicly released or is currently being tested. Otherwise you may need to proceed with that MATLAB method.

    One thing I do know is when I simulated for 30kHz the output voltage stabilized within 2 volts of the output predicted by that equation but also has a 2V ripple when I only had the 3uF at the output. The datasheet does indicate the PWM should be in that higher frequency range, so that equation may only be valid for that case. Ill go ahead and simulate at 30kHz with both of those 10uF capacitors to confirm if it will still stabilize around the predicted value.

    I will keep you updated on my results and any new information I discover.

    Best,

    Carson

  • Hey Weijian,

    Although the flyback configuration you are using is not the exact same as the one in the DRV2700 datasheet, this may still give you any additional insight to the workings of calculating the output voltage based on the configuration within the datasheet.

    VHV=VFB(1+RFB1/RFB2)-V1*RFB1/RFB2.

    Where:

    V1 will be: V1=V3(1+RINT2/RINT1) – V2*RINT2/RINT1 and V1 must be below VFB=1.3V.

    V2 will be determined by the PWM duty cycle and the PWM frequency's interaction with the lowpass filter's resistance and capacitance values

     

     

    Best,

    Carson

  •  Here is the image that was supposed to show in my last post

  • Hi Carson,

    After looking through the books here, I applied Kirchhoff's law and got the same result as you.As shown in the attachment.Thank you very much for the data

    Can you provide some videos about DRV2700 power flyback?For example, the output load is piezoelectric ceramics.

    Thanks,

    Weijian

  • Hey Weijian,

    The resources for the DRV2700 chip are limited and there are no videos regarding any of its applications, as it is becoming an older model that is being phased out. For any type of video I would recommend doing more research just on generic power flyback boost configurations and not just the one associated with this chip. 

    For anymore specific questions just ask them here or start a new thread and I will try use my resources to answer them to the best of my ability.

    Best,

    Carson