Inverting LM4780 circuit

[CommanderLake]

I'm looking at designing my own 8 channel amplifier board with 4x LM4780's and the datasheet says an inverting config might reduce THD+N and eliminate the need for the 68μF Ci caps but it says RB on the positive op-amp input to ground needs to be the sum of Rf and Ri when in theory and on the Falstad circuit simulator the resistor does nothing at all, please explain this?

 

[(*steve*)]

Since you are referring to a datasheet and the circuits and advice given therein, it would be good if you provided a link to it.

There may be more than one datasheet, so it is best if you give the link to the one you're using.

That way we'll all be on the same page.

 

[(*steve*)]

My guess is that this is the simple rule of thumb used to negate the effects of input offset current. And these effects are unlikely to be modeled in a simple circuit simulator.

 

[CommanderLake]

Sorry I assumed you would just look on the ti.com website for it anyway here it is: www.ti.com/lit/ds/symlink/lm4780.pdf

 

[(*steve*)]

I gave the reasons. It's also simply polite to provide a link to the datasheet you're using.

And on a 30 page datasheet it would be nice to be referred to a particular page.

(It's on page 23 for anyone else that's interested)

Yes the formula for Rb gives a similar impedance from each input to ground. This results in the effects of input bias current being canceled out.

An Apex Application note (AN-17) describes the advantage of an inverting amplifier (with respect to distortion) as:

"The inverting amplifier configuration forces common mode potentials to zero. By doing so, non-linearities due to common-mode effects are also reduced to zero."

 

[CommanderLake]

Well I know now and thanks for the info.

 

[CommanderLake]

Looking at figure 39 on page 24 of the datasheet I cant find any mention of the CN capacitors on the datasheet, what do they do?

 

[Harald Kapp]

The CINs decouple the input from the amplifier. They will let AC signals pass, but block DC. Therefore any offset from the signal source is eliminated.

The CNs area a short circuit for high frequency noise. The have a rather high impedance al low (audio) freqeuncies, but a low impedance for high frequencies.

Note the small but important difference between CIN and CN.

 

[CommanderLake]

What would be more suitable ceramic or mica?