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.

LM60440: LM60440 PSRR

Part Number: LM60440
Other Parts Discussed in Thread: TPS7A85A

Hi 

I have been given a design requirement of a PSRR of 60dB@1KHz. However, the datasheet indicates up to 7% line regulations in the low load range which is probably even worse at 1KHz do to the LC resonance at 6KHz (15uH, 47uF).

What do you think about the 60dB@1KHz requirement using the LM60440 (A or D) webbench design? and please elaborate on how you translate the PSRR to LM60440 spec in case you do not agree with my rough estimation from above. 

  • Hi,

    If you are interested in creating a low output noise application, it is always recommended to have a DC-DC step down converter cascaded with a LDO.

    Please refer to the following training material (https://training.ti.com/ldo-basics-power-supply-rejection-ratio-psrr).

    For example you can have the LM60440 provide a stepdown from 12V to around 5.4V output. Then the 5.4V output will be the input to the LDO and the output can be a true and clean 5V rail. As an example, you can look at the TPS7A85A Figure 9 for a rough idea of the PSRR vs frequency as a starting part.

    I recommend posting an E2E post on the LDO forum to get an appropriate engineer to suggest an LDO appropriate for your design application.  

    Regards,

    Jimmy

  • Yes I know that I can use a ldo after the buck but I am asking for the relationship between PSRR and line regulations? And if 60db@1khz is possible with this converter not how to make a low noise ldo. 

  • Hi,

    Most of the DC/DC step down converter datasheets do not have a PSRR information available. 

    Also the line regulation curve is generated assuming the bench top input supply is providing the typical input value. However this does not mean there is no input ripple present. The line regulation graph provides some reference/guidance on the expected output voltage ripple at different load current conditions for a set input voltage. 

    To properly calculate the PSRR of the device, both input ripple voltage and output ripple voltage (at a known power condition) should be properly measurement and evaluated.

    The best way to evaluate this is to take the LM60440EVM,  input voltage with some input ripple and measuring both Vripple_in and Vripple_out. The PSRR can then be calculated (from the equation presented in the TI training material) for a given input voltage and power load condition with this method. PSRR (db) = 20log (Vripple_in / Vripple_out).

    Regards,

    Jimmy 

  • Sorry for not being clear so I will try to elaborate.

    What I am asking for is a way to get a rough estimate of the PSRR performance just having the data from the data sheet.

    Looking only at the datasheet my gut feeling is that the LM60440 will not be able to do 60db@1khz over the full load. This I base on the loop gain is ~40db@1khz, the line regulations is betwee 3%-7% over load and the fact that the LC output filter has resonans at 6khz close to the 1khz. 

    So is anyone able to agree Og disagree with my gut feeling either on past measured experience or knowledge on the PSRR performance of LM60440? 

    I know how to improve the PSRR with additional components and how to measure PSRR on a real device. 

  • Hi Tim,

    Let me talk to the previous apps engineer to see if there is anything here that can be commented.

    From my experience of general DC/DC switching regulators but limited experience on the topic of PSRR, I don't expect it meet a PSRR spec without any post filtering since the switching component inherently generates noise at the output. The switching behavior of the internal MOSFETs create fast edge discontinuous voltage spikes that couple as high frequency noise onto the output ripple. 

    The noise also gets worse at higher load currents and higher input voltage which I expect you'd use this device for. There may be some devices that have features to help minimize the output noise, which is a combination of high frequency spikes and low frequency ripple but my gut feeling here is that most step down converters that are loading 1A+ applications will not have a PSRR performance that can meet your requirement without additional post filtering. 

    When I hear back from the apps engineer, I'll let you know of the resulting conversation.

    Thank you for your patience.

    Regards,

    Jimmy

  • Hi Tim,

    I've previously reached out to the apps engineer and came to the same conclusion.

    There has not been any test done on this and newer step down converters regarding PSRR functionality.

    It is generally recommended to use a post filter or LDO since this device is a switching regulator which inherently introduces noise into the system.

    As such it is believed that the device would not meet your 60dB @1kHz without additional filtering.

    Regards,

    Jimmy