B&W 606 S3 on an Onkyo TX-NR6100: 100 watts do not cancel a 3.7-ohm dip

A 3.7-ohm dip on a receiver rated to 4 ohms: marginal, not dangerous.

The TX-NR6100 is rated 4–16 Ω and makes 100 W per channel into 8 Ω, 20 Hz–20 kHz, 0.08 % THD, two channels driven, FTC. The 606 S3 is 8 Ω nominal with a 3.7 Ω minimum. The calculation divides one by the other, gets 0.925, and returns impedance marginal: connect it and listen to it, knowing the reserve runs out at the top of the dial. The level is not the constraint — 88 dB at 2.83 V into 8 Ω is a genuine one-watt figure, and at 4 m with 100 W the result is 96.0 dB, 9.0 dB below the 105 dB reference peak.

A rating and a dip are not the same kind of number. The «4–16 Ω» on the back of the receiver describes the nominal impedance the amplifier is built to drive — the averaged number a manufacturer prints on a speaker’s label. The 3.7 Ω is the lowest point of a curve, held in a narrow band for fractions of a second as the music passes through it. Almost every passive speaker dips below its nominal rating somewhere; a rating that had to cover the worst point of every curve would read «2 Ω» on nearly every receiver sold. That is why this pairing, and hundreds like it, have worked for decades while looking wrong on paper. This page used to return a flat thermal-overload warning for anything under the rating. It is now graded by ratio, and 0.925 sits inside the rating’s own margin — the warning is reserved for dips below 0.85, where the load really is deeper than the amplifier was designed around.

Where the margin does run out. This is a seven-channel receiver on one shared heatsink, and the case that matters is the one people buy it for: several channels loaded, film peaks, a warm cabinet in a closed rack. Sustained loud bass is the signal that holds the load near its minimum long enough to matter, and what happens then is the protection circuit muting the receiver — at the loudest moment of the film. An interruption, not a repair bill. Give the cabinet air and leave a few decibels below the top of the dial and you will not meet it.

B&W qualifies its own power rating. The window is 30–120 W into 8 Ω «on unclipped programme». That last phrase is the specification: a 40 W amplifier driven into clipping delivers more destructive energy to a tweeter than a clean 120 W one. At 100 W FTC the Onkyo sits inside that window with the qualifier satisfied, which is the part worth reading — it matters more than comparing the headline numbers.

How much of this is measurement choice. Onkyo’s US sheet prints 210 W per channel at 6 Ω, 1 kHz, 10 % THD, one channel driven, right beside the 100 W FTC figure. The European sheet for the identical receiver prints only 165 W IEC (6 Ω, 1 kHz, 1 %, one channel). One amplifier, three numbers, 2.1× apart, and a buyer comparing an EU listing with a US one sees two receivers that do not exist. Onkyo also quotes dynamic power down to 3 Ω (240 W front), which is a real statement about current capability rather than a marketing line — and it is the number that agrees with the graded verdict rather than the old flat one. Note the cable while you are here: 12 AWG over 15 ft costs 0.6 % and gives a damping factor of 82; 16 AWG would be 1.5 % and 47, and 20 AWG 3.7 % and 23. For calibration, a Triangle Borea BR03 at the same 4 m reads optimal at 98.0 dB — a milder minimum and more sensitivity, not more watts.