• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 64 subscribers
  • Views 450 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LM5576: the reason why Vout droop/non-regulation

Ella KIM
Expert 4102 points
Part Number: LM5576

Hi team,

Can I ask for your help to find the problem on TINA simulation results?

LM5576_calculator.xls

From the excel calculator, 75uH of Lout is recommended under the condition of 50Vin, 28Vout, 2Aload. But unavoidably customer has to go with 22uH of Lout. So I tried to simulate it with a TINA model as attached, where I ran into problems below : 

1. I applied 500mA of load. As you can see in capture1, it regulates into 28V well but I'm not sure COMP waveform is a normal waveform. Please check and advise on whether Vout is well regulated with the correct COMP waveform.

2. I applied 2A of load and it shows abnormal VOut regulation, only upto 16V, as attached in capture2. Why is it not regulated? 

LM5576_22uH_snvmbp5.TSC

Thank you!

Capture 1 & Capture 2

  • 0
    TI__Expert 8130 points

    Hi Ella,

    I don't think you attached the images. They are not showing on my end. Please provide capture 1 and capture 2 for review.

    Thanks,
    Alejandro

  • 0 in reply to iAlejandro
    TI__Expert 4102 points

    Alejandro,

    Sorry for missing it out. I attached the images again.

    Could you please review it asap? Customer has the issue on their board right now and tight schedule.

  • +1 in reply to Ella KIM
    TI__Expert 8130 points

    Hi Ella,

    1. Vout is regulated. Notice how the FB voltage is ~1.225V as it should be.

    2. Looks like it is hitting current limit.The low inductance value has lowered the max output current range. Current ripple increases with decreasing inductance. Notice how the COMP value is fixed (error amplifier is asking for more power) and FB is staying well below the ~1.225V regulated voltage.

    Regards,
    Alejandro

  • 0 in reply to iAlejandro
    TI__Expert 4102 points

    Alejandro,

    Please see the capture2, not 1. Vout is not regulated with increased load (2A). The target Vout is 28V. (I set R5 = 22.2k and R6 = 1k following the eq (15) in DS.) FB is only <0.8V, but it should be 1.225V as you mentioned.

    Could you advise why the LM5576 doesn't regulate the output when the load increases? I'm sure Rramp and Cramp are designed well following eq (3) and (4) in DS. Maybe COMP pin components are the issue?

  • 0 in reply to Ella KIM
    TI__Expert 4102 points

    Alejandro,

    I found the problem. The low frequency (approx. 286kHz with R_RT=21.6k) was a problem as it triggers current limit as you mentioned. After I change R_RT 10kOhm, thereby the switching frequency becomes 518kHz, I'm able to get 28Vout with 2A load.Thank you for your advice!

    But it still shows an unstable regulation. I guess it's because the compensation value is not correct. I don't have any tool to find its values with Lout 22uH. Could you help me to find the proper values? - For the sim above, Ccomp = 22nF, Rcomp=15kOhm

  • 0 in reply to Ella KIM
    TI__Expert 8130 points

    Ella,

    That's good to hear! Yes, increasing the switching frequency reduced the ripple current therefore not triggering current-limit.

    The components can be designed with the guidelines given in the following section of the datasheet. I would recommend to capture the error amplifier gain and phase results and see which component needs a modification. Increasing Rcomp, while proportionally decreasing Ccomp, increases the error amp gain and vise versa. 

    Regards,
    Alejandro

  • 0 in reply to iAlejandro
    TI__Expert 4102 points

    Hi Alejandro,

    When I read thorough the section of DS you mentioned, it describes type 2 compensation quite simple without detail design process. Only thing I need to do is to place fz = 1/(2pi*Rcomp*Ccomp) onto fp(mod)=1/(2pi*Rout*Cout). Is my understanding correct?

    iAlejandro said:
    I would recommend to capture the error amplifier gain and phase results

    How do I capture the gain and phase margin? I want to know the phase margin and BW but I don't think it's possible only by the calculation in the datasheet. Could you please proceed a simulation for customer to find proper Rcomp, Ccomp value, and phase margin? Or would you provide me any simulation model or calculation tool? I tried to get a bode plot in TINA model but I need your help to modify the TINA model properly.

  • 0 in reply to Ella KIM
    TI__Expert 8130 points

    Hi Ella,

    The TINA model should have the option to run Body Analysis. I will look into it and get back to you as soon as possible.

    Thanks,
    Alejandro

  • 0 in reply to iAlejandro
    TI__Expert 4102 points

    Thanks, Alejandro. Your quick feedback would be really appreciated! Customer has been waiting so long to find proper compensation values..

  • 0 in reply to Ella KIM
    TI__Expert 8130 points

    Hi Ella,

    Unfortunately, I am unable to get the Bode Analysis working on TINA. TINA is an outdated simulation tool with little support. I would recommend to run tests on bench. There are no simulation or calculation tool for this device found in the product folder.
    I've contacted the internal team for further inputs on TINA simulation tool.

    Regards,
    Alejandro

  • 0 in reply to iAlejandro
    TI__Expert 4102 points

    Thank you Alejandro for checking..

  • 0 in reply to Ella KIM
    TI__Expert 8130 points

    No problem. For the meantime, I would highly recommend to run tests on bench.

    Regards,
    Alejandro