Is the 22uf C2 (feedback cap) in LM1875 optional

[AndrewT]

....... you not only “can” mix AC and DC coupling, but you “should” do it that way, provided it’s done correctly. ............

how do you do it correctly?
You rightly point out that AC coupling the NFB and DC coupling the input is bad.

You rightly point out that DC coupling the NFB and AC coupling the input can be made to work if the output offset is acceptable.

Pacific has shown that output offset can become excessive/damaging if the odds stack up against one.

WHY?

Because, when you mix AC & DC coupling, the resistances seen by the two input pins do not match. When these -IN & +IN resistances match the input offset currents create the SAME input offset voltage on both inputs and these matching offset voltages cancel at the output because these opamps are differential amplifiers.

Subject to production tolerances the input offset current for both inputs are similar. Slight adjustment of the IN resistances can be used to ZERO the output offset.

 

[BionicSniper]

Slight adjustment of the IN resistances can be used to ZERO the output offset.

Wouldn't the best solution then be to use a pot to find a good value then through a 1% R in there

 

[AndrewT]

Wouldn't the best solution then be to use a pot to find a good value then through a 1% R in there

when you care to look at the various implementations reported in this Forum, you will find that many do indeed use a pot to balance the output offsets.