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.

LM5170-Q1: Two-Phase Interleaved Buck Converter

Part Number: LM5170-Q1
Other Parts Discussed in Thread: LM5170
Hello,
I am working on a high-power Buck Converter. The specs are as follow:
Vin: 40V ~ 60V
Vo: 24V
Io = 50 A
# of phases = 2
fsw = 80 kHz
Control Mode: Output Voltage Control Mode
Mode of Operation: Buck only
Controller IC: LM5170-Q1
I have downloaded the PSpice simulation model for this IC. The original downloaded model works fine, however, the system becomes unstable when I use the values of resistors and capacitors I get from the Excel Design Sheet. Can you please have a look at it ?
I am attaching both the  Excel Design Sheet and PSpice model with the email. 
Is there another TI Controller IC that has 4 phases ? In the second version of the above power-supply, I will be building a 100A converter. 
Thank you.
  • Hi Zuzahid,

    Thanks for using the e2e forum.

    Today is a public holiday in Germany, therefore responses may get delayed.
    I will look into your question and get back to you by the middle of this week.

    Thanks for your patience.

    Best regards,
    Niklas

  • Hi Zuzahid,

    Sorry for the long delay.

    If the system becomes unstable when adjusting the the resistors and caps according the the calculator, it might be related to the loop compensation.
    Looking at the bode plots of the calculator attached, the system should run stable at a crossover fsw of 200Hz and approx 90 degree phase margin.
    To check for a loop compensation, I would recommend to try a higher crossover fsw design and check if the behavior improves.

    To check for alternative pure Buck converters, I would recommend to use our part search feature, or open a new thread to get a part recommendation from a Buck expert.
    https://www.ti.com/power-management/non-isolated-dc-dc-switching-regulators/step-down-buck/products.html

    Best regards,
    Niklas

  • Hi Niklas,

    I have tried many different cross-over frequencies for both Current loop and Voltage. And each one them were stable loops. I followed the general rule of thumb (Current loop cross-over frequency is 1/10th of switching frequency and the Voltage loop cross-over frequency is 1/10th of Current loop cross-over frequency) The problem is, each iteration takes about 90 min. 

    Is there an easy way of solving this issue ? Can you please help me in stabilizing the converter ?

    Thanks.

  • Hi Zuzahid,

    Thanks for the update.
    So if I understand this correctly, the calculations based on the calculator all show stable loops, but once the circuits is simulated in PSpice, the waveform simulations always show instability?
    Or did the simulations look okay and it is the physical circuit that is not stable?

    If it is the simulation, can you show a waveform shot of the simulation, so we can see in more detail how the device starts to oscillate?

    Thank you and best regards,
    Niklas

  • Hi Niklas,

    I am sorry I did not mention it clearly. I am running the simulations in the PSpice for TI. I have attached all the PSpice simulation files and the design Excel sheet with the first question. 

    I have used the LM5170 Design Excel Sheet to design a 2-phase Buck Converter. The Excel sheet shows a stable Closed-Loop system, however, when I implement the same system in PSpice for TI, it gives me very strange results. I have attached the waveforms of two Inductor Currents and Output Voltage. Please note that the output voltage is set to 24V. If you notice the waveforms, the output voltage never reaches the set voltage. Also, the current in the inductor 2 shows that the system is unstable.

     LM5170_Waveforms.pdf

  • Hello, 

    Thanks for the feedback. The sim time seems to be 8ms, how long is the soft start time in this design? What happens if you try to run the sim for a longer time (e.g. 15 ms)? Does it look different if the sim runs longer? We're looking forward hearing from you.

    Kind regards,
    EM 

  • Dear EM,

    The Soft-Start time is only 2mSec. I have tried the soft-start time of 6mSec and the result was similar. The current in Inductor 2 goes to Zero and remains zero. The simulation takes ~ 1hour per 8mSec of simulation time, so I had reduced the soft-start time back to 2mSec.

    I have attached the Excel Design Sheet with the original question. If you see the above the simulation results, the current in the second Inductor does not follow the reference and goes to Zero. I have tried many different cross-over frequencies but none of them seem to work. Please compare the circuit parameters in the simulation file with the Excel sheet, and it can be seen that I am using the exact same values. 

    Thanks.

  • Hi Zuzahid,

    I am very sorry for the long delay.

    I took a review of your PSpice model and I could not directly find any issues within the model.
    Like you already mentioned, the model uses the same structure as the original default model with only the values changed according to the calculator.

    As the model runs fine in the default state, there are two suggestion I have:
    - As the default model has a load step on the output side at 4ms and 6ms, I would recommend to run a simulation with one continuous resistive load, to check if the model starts oscillating in steady state operation
    - If this is still the case, I would recommend to go back to the default model and only modify the model step by step to find the source component for the instability

    I know that this might consume a lot of time and resources due to long simulation times, but I do not have a quicker solution for this case.

    Best regards,
    Niklas