• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 963 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.

TPS40061 above 10A (input 20V-50V, output 5V)

Sal Salami
Prodigy 40 points
Other Parts Discussed in Thread: TPS40061, TPS40060

Hi all,

 

We are using TPS40061 in our design. I looked at the datasheet but there is no mention of the maximum output current. However in the specification page for the part the maximum current is set to 10A. I am trying to have output current between 12A to 16A. In our design for the input of 24V, it seems the conversion to 5V is efficient (efficiency around 85% to 90%). However when I use 48V as input, the efficiency drops dramatically (70%) ( board is 10degrees hotter and 10W more consumed for the same load). I am using load between 12 to 16A.

Couple of question:

Why the web site put 10A for the part and no mention in the datasheet.

Is the above efficiency for voltages is normal for the part at 48V and load.

Is there any advice to make it more efficient for 48V to 5V (switching frequency, etc)

Any other part that can handle 20V-50V input for delivery of 5V and 15A

 

Thank you for your time.

Cheers,

 

 

Sal

  • 0
    TI__Guru 56965 points

    Sal,

    Why the web site put 10A for the part and no mention in the datasheet.

    The Website offers recommended applications where the IC can be used, but it has some limitations.  To help our customers, TI recommends a range where we believe the device will provide good performance with a reasonable design.  That requires us to determine what is "good performance" and "reasonable design" for a given part based on its features and capabilties.  Our experience tells us that the performance that a power supply based on the TPS40061 controller can provide is generally not what customers are looking for when load currents increase above 10A.

    Is the above efficiency for voltages is normal for the part at 48V and load.

    There are few things worse than being asked what is "Normal" I'd say if you didn't make any changes to the circuit, they are not unexpected.  From 24V to 5V, the converter is oppertating at about 20% duty cycle.  The high-side FET is carrying the current 20% of the time and the low-side FET about 80% of the time.

    When we increase the input voltage to 48V a few things happen:

    1) Switching Losses in the high-side MOSFET double due to the increase in switch voltage

    2) Conduction Losses in the high-side FET are reduced by half due to the shorter on-time and lower RMS currents that result.

    Typically 1 & 2 would drive a customer to select a higher Rdson high-side FET with lower gate charge and faster switching speeds.

    3) Losses due to gate drive & linear regulators (like BPN10 and BP10) double as they're drawing their current from a higher input voltage)

    4) Losses in switching capacitances, such as MOSFET output capacitance & snubbers increase by 4x as they store more charge from a higher voltage

    5) Conduction Losses in the low-side FET increase by about 12% due to the longer OFF time when the low-side FET is conducting.

    Is there any advice to make it more efficient for 48V to 5V (switching frequency, etc)

    Lowering the switching frequency would help, though this will drive a larger inductor and more capacitance to meet the same requirements.

    Selecting MOSFETs better for the new operating point would also help.  As stated above, as a first cut, you could try selecting a high-side (P-channel) FET with approximtely 1/2 the gate charge and 2x the Rdson to rebalance the switching and conduction losses for the increased input voltage

    You could look into selecting a lower Rdson low-side (N-Channel) MOSFET that will offer less conduction losses, though be careful not to make it too low where even the low-side switching / capacitive losses increase total losses.

    Any other part that can handle 20V-50V input for delivery of 5V and 15A

    Currently, the TPS40060 and TPS40061 are the only devices available that can deliver 15A from a voltage as high as 50V, but TI is constantly developing new products and may have an improved solution available in the near future.

  • 0 in reply to Peter James Miller
    Prodigy 40 points

    Peter,

    Thanks for your comments. Very helpfull. It open a few more questions for me. Mostly to help me choose how to proceed from here. At the moment my design is in production and my only option is change the parts (MOSFET, inductor, caps, resistors).

    1 - Do you recommend to have different circuit for 48V,15A with different MOSFETS or you believe the same circuit for 24V and 48V delivering 15A can be designed. I want to know by changing the parts and design for 48V weather I still get the same efficiency for 24V.

    2 - Do you recommend I put effort now to have a better circuit for 48V, 15A or I better wait for your improved solution. Do you believe I can get the efficiency up to 90% by modifying my circuit (consider limitation of MOSFET in designing such circuit and my options in changing the parts only now) for TPS40061.

    3 - Is the improved solution you referred close to release or in design stage. Is there any date mentioned in releasing the sample. is it in the same package as TPS40061.

     

    I understand to answer to the question 1 and 2 you need to know my level of electronic design. For now assume I have very good knowledge of electronic design (close to expert).

     

     

    Regards,

     

     

    Sal

  • 0 in reply to Sal Salami
    TI__Guru 56965 points

    1 - Do you recommend to have different circuit for 48V,15A with different MOSFETS or you believe the same circuit for 24V and 48V delivering 15A can be designed. I want to know by changing the parts and design for 48V weather I still get the same efficiency for 24V.

    While a common circuit can be designed, it must be optimized for a perticular operating point.  If the circuit is optimized for 24V input, it will not be optimum at 48V input.  If it is optimized for 48V, the reverse is true.  If it is balanced between these two points, it will actually be non-optimum for both.  The best design process would be 2 designs using different inductors and MOSFETs for the 2 different input voltages IF they don't require a common design.

    2 - Do you recommend I put effort now to have a better circuit for 48V, 15A or I better wait for your improved solution. Do you believe I can get the efficiency up to 90% by modifying my circuit (consider limitation of MOSFET in designing such circuit and my options in changing the parts only now) for TPS40061.

    I sent you a private message so we can discuss the details of your application and design cycle needs.

    3 - Is the improved solution you referred close to release or in design stage. Is there any date mentioned in releasing the sample. is it in the same package as TPS40061.

    I can not comment on unreleased solutions or devices.  My statement was more of a blanket statement that TI is continuously developing new devices for the non-isolated DC/DC controller market, so just because we don't have a solution today doesn't mean we will not have a new solution soon.