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.

LM706A0: Stability Errors with WEBENCH Design

Part Number: LM706A0
Other Parts Discussed in Thread: LM70660, LM5148, LM5085, LM5137

Tool/software:

Hey Buck Regulator Team,

I have a customer evaluating the LM706A0 for a new design and we are running into an error when trying to generate and simulate a design based on the following input/output conditions:

  • Input range: 42V - 57.4V
  • Output: 24V at 10A

The error mentions that the design is not stable due to the phase margin. Can you please help to provide guidance as to whether or not this device can support the given application or if there is an issue with the WEBENCH model?

If there is another device in your portfolio which would be better suited for these power conditions we are open to your suggestions as well.

Best regards,

Matt Calvo

  • Hi Matt,

    Can you close this thread by clicking on "resolved" as the conversation is now through internal email. Thanks!

    Ben

  • Hey Ben,

    I intentionally posted this technical use-case question on the this page so that my customer and I can collaboratively get support from your team on this channel. My customer does not have visibility into our internal email thread so using this thread for design discussion is a more efficient way of getting them direct support from your team.

    I'll go ahead and reply to your email and forward your reply to the customer but I'd like to also continue the technical follow-up here so that my customer can be involved too.

    -Matt

  • Just for clarity purposes, I'm pasting the answers to my original questions below:

    From Buck Regulator Team:

    10A at 24V is a lot of power even with a heatsink. It may be doable with a large heatsink but I suggest using two LM70660 (multi-phase) to keep the thermals down. WEBENCH is not broken. The intended goal for the 10A devices was to limit the output voltage to 5V. Anything over 5V, WEBENCH will output an error. 

    My customer and I will report back here with any technical follow-up questions regarding the advice given above!

    -Matt

  • Hi Matt,

    Thanks for the clarification. Keep us posted. Thanks,

    Ben

  • Thank you, Matt!

    It would be helpful if WEBENCH could be updated to report the correct error - something like "Output voltage > 5V is not supported" rather than a misleading error.  It would also be best to update the LM706x0 datasheet to reflect this for the LM706A0 part in the table within section 5.3 Recommended Operating Conditions.  And the LM70XX0 Quickstart Calculator spreadsheet, as well.

  • WEBENCH doesn't have a way to report that the suggested use case is not supported. I have asked that already to the WEBENCH team.

  • Hey Ben,

    Thanks for your help with this! A couple follow-up questions from our side:

    • Understood that WEBENCH cannot report those details. In terms of things that can be updated...do you think that the LM706A0 datasheet and calculator spreadsheet should be updated to more clearly convey the fact that the intended goal for this device was to limit the output voltage to 5V? This clarification in the ROC table and other collateral would likely help other customers who might run into this same situation
    • We are open to considering 2x LM70660s, but we are also wondering if it may also make sense to consider a more traditional controller with external MOSFETs so that we can sink the heat directly. Do you have any advice on the tradeoffs here and do you have any recommendations for controllers w/o integrated FETs that may be best to consider?

    -Matt

  • Hi Matt,

    Understood that WEBENCH cannot report those details. In terms of things that can be updated...do you think that the LM706A0 datasheet and calculator spreadsheet should be updated to more clearly convey the fact that the intended goal for this device was to limit the output voltage to 5V? This clarification in the ROC table and other collateral would likely help other customers who might run into this same situation

    We have a section in the datasheet (7.1.3.1 Derating Curves) that shows the output current vs. ambient temp. Depending on the conditions, the part will operate up to 10A but additional measures must be taken such as cooling, airflow, lower thermal resistance of the PCB or adding a heatsink on the IC to ensure the junction temperature does not reach 150C. What I mentioned above, where the part was intended to operate with an output of 3.3V or 5V, was assuming there were no additional cooling and operate at an ambient temperature above 85C. 

    We are open to considering 2x LM70660s, but we are also wondering if it may also make sense to consider a more traditional controller with external MOSFETs so that we can sink the heat directly. Do you have any advice on the tradeoffs here and do you have any recommendations for controllers w/o integrated FETs that may be best to consider?

    I would start a new thread if the discussion goes towards the use of a controller. For tradeoffs between controllers and converters, there is an application note that was released a few months back. See link below:

    https://www.ti.com/lit/an/snvaaa7/snvaaa7.pdf

    In my opinion, I would look into LM5148, LM5137, or LM5085. 

    Ben