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.

WEBENCH® Tools/TPS62177: TPS62177 Input capacitor value

Part Number: TPS62177
Other Parts Discussed in Thread: TPS62175, TPS62745, TPS62122

Tool/software: WEBENCH® Design Tools

Hi,

We have simulated the the TPS62177 the results for our requirements.

Please find the attachments.

Our circuit uses the Current source as input (4-20mA loop device)

When we implemented the Simulated circuit it is not working for the 4mA it is working only after the 5mA

Then by increasing the input capacitance to 22uF the board starts working.

What could be reason and what is the minimum capacitance for the TPS6217x. for 7V input and 5mA output @3.3V outputWBDesign251.pdf

  • Hello,

    If you did not have a current limited supply, the 2.2uF Cin value would be able to work.  However, since the input supply is current limited, a larger Cin is going to be required to supplement the dynamic loading during start up.  If you monitor the input voltage (Vin pin) of the device during start up, I would expect that it is dropping  significantly (possibly below the UVLO threshold) with the 4mA limited supply and 2.2uF Cin, whereas the drop will not be as large with the 5mA supply or 22uF Cin.

    I have also forward this request to our WEBBENCH team as it does not appear that the WEBBENCH model is able to account for your current limited input supply.

    Thank you,

    -Matt

  • Hi Matt,

    Thanks for the explanation and you are correct the input voltage is below UVLO 3.7V @ 4mA.

    Other than using 22uF, Is there any alternate approach ? (As our application demands less Capacitance at the input side we are forced to use less than 4.2uF.)

    Thanks,

    - Koti

  • Hi Koti,

    I notice that the WEBENCH design report you shared has  output of 3.3V and 10mA, for this condition your input average current is above 5.2 mA for 7V input. This will cause issues if you are using supply current limited to 4mA. You must have current limit well above the average input current requirement. Also note that this value is shown in OpVals section - it is dependent on efficiency. The average value in SPICE simulations / on actual board may be slightly higher or lower - so it is recommended to keep sufficient margin above the average current value shown in OpVals.

    Please confirm if you are testing for 5mA output current  as mentioned in the post or 10mA as shared in the WEBENCH design report. If you are using 5mA output current  and still seeing failure - please share your simulation circuit so we can debug and identify the issue. As Matt mentioned this could also be an issue with Vin going below UVLO during startup - if this issue is the cause then again as Matt suggested increasing Cin will help with that.

    I am not sure of any alternate approaches as the startup current will go higher than current limit causing reduction in input voltage. 

    Regards,

    Srikanth Pam

     

  • Koti,

    As Srikanth highlighted above, this design will not work if the output current is 10mA.  So, please confirm that the output current is limited to 5mA before making any further changes.

    Although a larger Cin value is your best solution, you can try using a 5:1 or similar resistor divider between Vin and EN to delay the device turn on until Vin exceeds ~6V.  The additional voltage accumulation and larger EN pin hysteresis will effectively act as a larger capacitor during start up.

    Alternatively, as the charging of the output capacitance is the heaviest load during start up, you can also look into reducing the output cap to 10uF, but to ensure you maintain stability, this will require you to also increase the inductance to 22uH.

    Please see the Tech document below for more information on optimizing the output filter design for the tps62175:

    Sincerely,

    Matt

  • Hi Srikanth,

    Thanks for the response and your observation.

    The simulation we have done for 10mA because it is maximum current the system needs in one of the condition.

    in normal/idle condition the system is taking 5mA.

    In both the cases input current is limited to less than 4mA.

    Thanks Koti

  • Hi Matt,

    Thanks for the approaches i tried first one but no significant results. still same result.

    for the second approach i haven't tried yet. again we have less inductance restriction at the output side.

      

  • Hi,

    As I can see in the datasheet of TPS62745 the input capacitance can be 4.7uF to 10uF.

    As you Know my requirements can i use this part for my application with 4.7uF as the input capacitor.

    Thanks 

    - Koti

  • Koti,

    Given your application conditions, changing the device is not going to make any significant difference in the result as your problem is inherent to your system architecture.

    For instance, a 4mA limited 7V input supply can only provide 28mW (Pin_max) of power, and you have confirmed that under some conditions your output requires 3.3V at 10mA, which is 33mW.  Taking into account the converter efficiency, the max output power will be limited to ~25.2mW (0.9*Pin_max), if you are operating at 90% efficiency.  

    Although using a larger input capacitor will allow for some resilience to load transients, the overall average output power has to remain below the above calculated maximum.  Additionally, the minimum input capacitor will need to be sized based on the magnitude, duration, and frequency of the load transients to ensure that sufficient input voltage is maintained throughout operation.

    Would it be possible for you to provide a timing diagram of the output load? 

    Sincerely,

    -Matt

  • Hi Matt,

    Regarding power, Your are Correct.

    Sorry its my mistake with the Numbers. Please find below Inputs

    Input Voltage is 6.8V Nominal

    Input Current is 3.6mA

    and Output Voltage is 3.3V

    Output Current is 5mA

    But in my condition it doesn't even turns on with 4mA Load

    Then I tried your suggestion of changing 22uH and 10uF as output. Then minimum input cap i can tune is 14.4uF nut my requirement is 4.7uF or less.

    Please suggest my point is if this is only problem with UVLO then if can go with the TPS62122 I feel it works. or suggest some other part.

    and the main reason why I am particular about these two parts is the Ripple frequency range is higher.

  • Koti,

    I believe the problem you are seeing is caused by the additional loading of the output capacitance during start up which is causing the input capacitor to discharge below UVLO due to your current limited voltage supply.  

    I would not expect that simply switching to the TPS62122 alone is not going to be enough to resolve your problem because the lower UVLO thresholds also means the converter will also be enabled at lower input voltage.  Thus, in order to maximize the energy storage at start up, you will also need to use the programmable EN threshold on either device to maximize the amount of voltage drop you can withstand before the converter is disabled due to the input capacitor discharging.

    Additionally, if you can tolerate some additional ripple on your output voltage, the TPS62122 will allow you to reduce the output capacitor to 4.7uF which will notably reduce the additional load on your input supply during start up.  

    Attached are the Spice models for the TPS62745 and TPS62122 that I created to emulate your system.  As current limited voltage supplies can be tricky to accurately model, I started with a rough model of my own using a 4mA current source and diode/voltage clamp on Vin, but you may need to make some adjustments to better match your system performance.  Nonetheless, you should find these as a good starting point for you to evaluate how the different components will impact performance. 

    Please note that the TPS62122 model does have a bug (being fixed) in it where the EN pin hysteresis is not being properly modeled which causes the regulator to oscillate on/off during start up.  In application, I would expect to see a 1+V drop in Vin (similar to the 62745 model) during start up, but based on how you setup the EN divider this should not cause the converter to be disabled.  And, once the output voltage become stabilized, the limited input power available will recharge the input cap.

    Lastly, I did find that the TPS62745 model I created appears to be operating with a 4:1 resistor divider on the EN pin and 22uH/10uF output filter.  So, you this may be worth double checking before switching to the 62122, but this may also be an artifact of the input supply model not being entirely accurate.

    TPS62122-2020-09-11T16-18.zip

    tps62745-2020-09-11T16-19.zip

    Sincerely,

    -Matt