• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 63 subscribers
  • Views 474 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS62136: Output Capacitor

maceti
Prodigy 230 points
Part Number: TPS62136


Hi,
I will use TPS62136 for 5G Module.
In the example power diagram in the 5G Module HW Design Guide, it is recommended to use 2 x 220uF capacitors at the module input.
Reference circuit:
https://drive.google.com/file/d/1CRF08FVHV61a5waHIUr0RnXt7wvWT8UJ/view?usp=sharing

Question 1: If I connect two 200uF Tantalum and ESR 0.4 OHM capacitors to the output of TPS62136 as suggested here, will I have a problem?
Question 2: I have stock of the 470uF 6.3V 0.4 OHM ESR tantalum capacitor in the link below. If I use 2 x 470uF instead of 2 x 220uF, will I have problems and what should I pay attention to?

I have used 2 x 470 uF in a different regulator before and the possibility of instability in this case was written to me in the last post here. By the way, I solved my problem here, the problem was caused by the production of the capacitors I used.
e2e.ti.com/.../tps62086-explosion-of-the-tantalum-capacitor-connected-to-the- tps62086-output

Regulator input will be 5V 3A adapter. I am looking for the best and smoothest combination I can create with the TPS62136 for a 5G module powered by 3.7V, which may have high current consumption and peak currents.

Thanks.

 

  • 0
    TI__Expert 7860 points

    Hi Maceti,

    Do you have a full schematic of your TPS62136 design showing the inductor used as well?

    What are your design goals/constraints? Since this is a 5G module application I would assume you want low ripple and noise? 

    The datasheet provides information in section 10.3.2, saying that tantalum capacitors should provide a large enough AC signal to the VOS pin due to the larger ESR.

    Thanks,

    Joseph

  • 0 in reply to Joseph Conrad
    Prodigy 230 points

    Hi Joseph,

    Thanks for your reply.

    More things are starting to become clear on my system.

    First of all, I want to answer your questions.

    The coil I use: 74438357015 , Wurth, 1.5UH 6.2A 20 MOHM SMD

    My schematic is simply as follows. Additionally I will add 1uF, 100nF, 33pF and 10pF in parallel to each VCC input in addition to a 470uF capacitor. For noise suppression.
    https://drive.google.com/file/d/1ZxybmDAZTEwwT80U933iaRG1qqfvwPUg/view?usp=sharing

    The need for the 5G module is to be able to provide 3A continuously and have a ripple of less than 100mV. According to the graphics of TPS62136 on Ripple, it already provides this. It can also provide 3A continuously. Is it correct?

    I rechecked section 10.3.2 of the datasheet as you said. Here it says that the "minimum total trace resistance" value should be 10mohm to provide sufficient AC signal to the VOS pin.
    So what I understand from here is that there should be a maximum length path between the output of the regulator and the 470uF tantalum capacitors, the tantalums should not be placed directly near the output of the regulator. I can leave a distance of 30mm between the Regulator and the 1st tantalum and 60mm between the 2nd tantalum. I am calculating a trace resistance of 14mohm on a PCB with 1 oz copper thickness. Do you think everything is ok this way?

    In addition, I need to use a 47uF capacitor at the regulator output just close to the regulator and place the 470uF capacitors close to the supply pins of the module, right?

  • +1 in reply to maceti
    TI__Expert 7860 points

    Hi Maceti,

    The 1.5uH inductor is per design recommendation and TPS62136 can support up to 4A output current.

    I agree with your reply on capacitor placement. Placing the 47uF close to the buck converter is recommended, and then placing the tantalum capacitors at the load will help reduce ripple and allow for stability from the trace resistance.

    The main concern with having high capacitances close to the device is when using ceramic capacitors that have a low ESR. For example, if you achieved 470uF using only ceramic low-ESR capacitors, then there would be very little ripple sent to the VOS pin that is needed for stable operation. Since in this design you are using tantalum capacitors with a relatively large ESR, there shouldn't be as big an issue using a high capacitance.

    Thanks,

    Joseph

  • 0 in reply to Joseph Conrad
    Prodigy 230 points

    Hi Joseph,
    Thanks for your reply. The information you provide is very valuable.
    This is how I design.
    best regards

  • 0 in reply to Joseph Conrad
    Prodigy 230 points

    Dear Joseph,

    I have one last question about this.

    I'm trying to calculate the capacitor I need to connect to the SS pin of the regulator to avoid inrush current in the case I'm using 2 X 470 uF.


    According to the Soft-Start Capacitor calculation on page 15, when I connect 33nF to the SS pin, it charges the capacitors in 9.24 milliseconds. According to the capacitor charging calculation, it charges the capacitors with approximately 300mA.


    How does the regulator know that I'm using the 2 x 470uF here? What logic does it have?
    Does it limit the output to 300mA until the capacitors on the output are full and the voltage rises to 4.2?

    Thanks.

  • 0 in reply to maceti
    TI__Expert 7860 points

    Hi Maceti,

    The Soft-Start capacitor size doesn't directly determine the current that the output capacitors are charged at. Instead, the voltage at the SS pin will limit the FB pin regulation setpoint so that it ramps up linearly. Since the device is regulating the output voltage rise time during soft-start, the startup inrush current will be limited, but not directly controlled by the device (unless the device current limits are hit). You should be able to get an estimate of the output capacitor charging current during soft-start, however it will also depend on the load that is connected.

    Let me know if this helps clarify the question.

    Thanks,

    Joseph

  • 0 in reply to Joseph Conrad
    Prodigy 230 points

    Hi Joseph,

    Thanks for your detailed answer. I understood how Soft Start works.

    You give clear and explanatory answers to all my questions. Also many thanks for your great support!

    Tüm sorularıma net ve açıklayıcı cevaplar veriyorsun. Harika support un için ayrıca çok teşekkürler!

    Finally, as 2 x 470uF 10V, I can say I've clarified using the capacitor below. This capacitor has an ESR of 100mohms.
    www.digikey.com/.../563910

    As I said, I'm thinking of setting a 10ms soft start time with a 33nF capacitor on the SS pin. In this way, I aim to prevent the problems that may arise from this by taking the inrush current under control. Is there any error in my calculations here or something else I should pay attention to?

    Thanks.

  • 0 in reply to maceti
    TI__Expert 7860 points

    Hi Maceti,

    You are on the right track since you have accounted for the large capacitance with some additional resistance to help stabilize the output voltage. 

    Your calculation for output current is also within range. I calculated around 350mA for no load during startup.

    Thanks,

    Joseph