ADP7182 Noise reduction question in datasheet

Hello all,

first time poster here, so I am not sure if this ends in the right forum:

I am working with some poer regulators: LDO type, positive and negative:

ADP7118 and ADP7182 in SOT23-5 Package, just for some information. Adjustable versions

Now I am having trouble to follow the datasheets in regard of the noise reduction techniques.

page 16 ( on Rev D datasheet)

noise = 11uV * (Rparallel + R2)/R2;   this one is kind of understandable ( its for the positive LDO)

now regarding the negative LDO, suddenly there appear to be 2 formulas, one is just describes in textform? what is that about, the other one has the same reading but suddenly is slightly different and in the text it says:

The noise of the LDO is approximately the noise of the fixed output
LDO (typically 18 µV rms) times RFB2, divided by the parallel
combination of RNR and RFB1

The noise of the LDO is approximately the noise of the fixed output
LDO (typically 18 µV rms) times the high frequency ac gain

now is that supposed to be the same, those two sentences ( simply someone forgot to dlete one after reviweing it

those two statement dont make any sense in combination, either the first one is adifferent noise, for example fixed output vs adjustable output, but two noises for one and the same thing, cannot exist.

So I would appreciate any further explanation, since i cannto explain this any more

Thanks a lot for reading,

cheers

  • 0
    •  Analog Employees 
    on Jun 28, 2018 8:51 PM over 2 years ago

    Hi Mex,

    I'll take a look into this and will get back to you soon.

    Thanks,
    Alex

  • 0
    •  Analog Employees 
    on Jun 29, 2018 3:20 PM over 2 years ago

    Hi Mex,

    The statement:

    "The noise of the LDO is approximately the noise of the fixed output LDO (typically 18 µV rms) times the high frequency ac gain" is correct. At high frequency operation, Xcnr is almost negligible on the equation compared to RFB1. The simplified equation should give you equation #2 on the data sheet. 

     

    AC gain = {[Rfb1 // (RNR +Xcnr)] + Rfb2} /Rfb2 ; RNR >> Xcnr; Xcnr is negligible.

     

    Simplified equation:

     

    AC gain = [(Rfb1 // RNR) + Rfb2] / Rfb2 which when further simplied is equal to equation #2.

     

    The statement:

    "The noise of the LDO is approximately the noise of the fixed output LDO (typically 18 µV rms) times RFB2, divided by the parallel combination of RNR and RFB1" is incorrect and should reflect same equation as the ADP7118 (positive LDO) equation #9. We'll update the data sheet to reflect the correct description of the equation.

    Basically equation#9 of the ADP7118 datasheet equates to the noise of fixed output multiplied by the AC gain since at high frequency Xcnr is negligible for the AC gain.

     

    Thanks,

    Alex

  • Ok,

    just to point out the crucial:

    ADP7182 ( negative LDO ) has errors in the datasheet.

    For some additional questions: How am I suppose to run a LDO in high frequency operation

    ( I assume this refers to simply some high frequency noise spectrum )

    I get that this whole noise reduction should not be critical to most of the operation the IC is used in, but for some it might help, so you are welcome that I have pointed out to an error, overseen so far

    For the sake of Jim Williams I donate this error finding to linear and thereby to analog, for ht enext error I will be very happy to receive some finders fee, since it is your own forum, you will find a way to contact me

    cheers.

  • 0
    •  Analog Employees 
    on Jul 4, 2018 3:34 PM over 2 years ago

    Hi Mex,

    Thank you for pointing out this paragraph error on the data sheet. We will make necessary changes to correct this.

    To "run the ldo in high frequency operation" simply refers to the high frequency input noise relative to what is seen on the output.

    Cheers,

    Alex