The subject of mismatched speaker wires came up recently and I said it could matter due to varying speaker impedance with the wire as a divider from the amplifier. That was based on tests done long, long ago (but in this galaxy). At that time zip cord from maybe 22 down to 18 AWG was pretty common so I got to thinking my experience was no longer relevant given 14 - 10 AWG is pretty common now.
I had a old schematic with three different speaker models. I am no longer sure exactly the speakers, but one is a fairly benign load similar to my Magnepans. The second is (was) a more common but not terribly demanding load. The last is a speaker with fairly wide excursions and a low-impedance (ribbon) tweeter. I modeled a solid-state amplifier with damping factor of 200 at 8 ohms (0.04 ohm output impedance) rising at high frequency (typical due to reduced feedback). In the simulations SSn shows the result of the amplifier driving speaker n with no wire, and SSn_wire is with the extra 20 feet of speaker wire (modeling the mismatch). I only included wire resistance, no other parasitics (no extra inductance, capacitance, or conductance terms -- usually not terribly relevant at audio frequencies, and/or I was too lazy to include them all, take your pick). The top plot shows impedance in ohms, and the other the voltage at the speaker terminals in dB.
First is a simulation with and without an extra 20 feet of 18 AWG wire. Speaker one shows negligible difference except a slight change in effective gain. Speaker 2 shows more, with a dip in the bass, and the extra resistance actually reduces the HF peak. Speaker 3 shows a bit more deviation. These deviations are not large, making me wonder about my original tests (done ca. 1982 and no I don't have the records) when DBT's showed people could hear a difference. I am not sure what speakers were used then; I think it was a pair of Magnepans (my MG-IIIa), a pair of B&W 801, and an ESL pair (not sure which one, there were several we used).
Now below are the results if instead I use 20 feet of 12 AWG cable instead. The differences are much less as we'd expect, probably inaudible.
So chances are mismatching your speaker cables assuming a reasonable amplifier, speaker, and cable size and length is not going to be audible. Of course this is one example and certainly does not cover all systems but I tried to use somewhat real-world examples.
FWIWFM - Don
I had a old schematic with three different speaker models. I am no longer sure exactly the speakers, but one is a fairly benign load similar to my Magnepans. The second is (was) a more common but not terribly demanding load. The last is a speaker with fairly wide excursions and a low-impedance (ribbon) tweeter. I modeled a solid-state amplifier with damping factor of 200 at 8 ohms (0.04 ohm output impedance) rising at high frequency (typical due to reduced feedback). In the simulations SSn shows the result of the amplifier driving speaker n with no wire, and SSn_wire is with the extra 20 feet of speaker wire (modeling the mismatch). I only included wire resistance, no other parasitics (no extra inductance, capacitance, or conductance terms -- usually not terribly relevant at audio frequencies, and/or I was too lazy to include them all, take your pick). The top plot shows impedance in ohms, and the other the voltage at the speaker terminals in dB.
First is a simulation with and without an extra 20 feet of 18 AWG wire. Speaker one shows negligible difference except a slight change in effective gain. Speaker 2 shows more, with a dip in the bass, and the extra resistance actually reduces the HF peak. Speaker 3 shows a bit more deviation. These deviations are not large, making me wonder about my original tests (done ca. 1982 and no I don't have the records) when DBT's showed people could hear a difference. I am not sure what speakers were used then; I think it was a pair of Magnepans (my MG-IIIa), a pair of B&W 801, and an ESL pair (not sure which one, there were several we used).
Now below are the results if instead I use 20 feet of 12 AWG cable instead. The differences are much less as we'd expect, probably inaudible.
So chances are mismatching your speaker cables assuming a reasonable amplifier, speaker, and cable size and length is not going to be audible. Of course this is one example and certainly does not cover all systems but I tried to use somewhat real-world examples.
FWIWFM - Don