Are you really asking how you can (effectively) claim A-B when the art teaches A-C-B, i.e, an intermediate connection?
Yes, but trying to not using "directly connected to" , wondering what is the best way to describe circuit connections for any new applications in the future.
I've thought about this one before, and was unable to think of claim language for A-connected-to-B that avoided prior art A-C-B while at the same time protected against infringer A-C-B. That is, I don't think you can have your cake and eat it too.
As a concrete example, consider:
An apparatus comprising:
a condenser having an outlet;
a compressor having an inlet, the inlet of the compressor coupled to the outlet of the condenser.
Assuming your spec and pros. history allows "coupled" to be interpreted as "directly or indirectly", this claim is good from an infringement point of view because an apparatus having a valve between the inlet and outlet still infringes.
On the flip side, if that apparatus with the intervening valve was prior art, it would anticipate.
But if you amend with "directly coupled" to avoid anticipation, you've lost on infringement. [Let's ignore the question of whether such an amendment is even worthwhile, because of obviousness]
If I missed something, I'd love to know because this concept seems really important.