Laboratory test report

The Naim NAP 250DR power amplifier only just exceeded its manufacturer’s claim for power output into 8Ω, as you can see both from the tabulated results and from the bar graphs accompanying this review. Newport Test Labs measured output at 1kHz as 89-watts per channel, both channels driven into 8Ω. As you’d expect, given the regulated power supply, power output with only a single channel driven was almost identical. Into 8Ω loads, power output was maintained at 89-watts at 20Hz, but dropped almost infinitesimally at 20kHz to 85-watts per channel… still higher than specification.

When driving 4Ω loads, the Naim NAP 250DR came within a whisker of meeting the theoretically ‘perfect’ ideal of doubling its power output… but not quite, with the amplifier delivering 156-watts per channel at 1kHz, 155-watts per channel at 20kHz and 137-watts per channel at 20kHz. Since Naim quotes power output at 150-watts per channel into 4Ω for the NAP 250DR, this puts the result at 20kHz just slightly under specification. The difference is so slight, however, that if the 240 mains voltage had been held stable (it varied between 238 volts and 252 volts during the testing), an output of 150-watts per channel at 20kHz may have been achieved on test.

Naim puts the recommended ‘minimum load impedance’ for the NAP 250DR at 2Ω, but not surprisingly, the amplifier’s output did not redouble when the amplifier was driving 2Ω loads, though it still delivered a very creditable 239-watts per channel both channels driven at 1kHz and 180-watts both channels driven at 20kHz. When tested into 2Ω loads using a high-frequency (20kHz) test signal however, the amplifier’s protection circuitry cut in to limit power output to 80-watts per channel. In practise this really won’t matter, since music will never contain continuous signals at 20kHz and, if the amplifier was producing its maximum output into 2Ω at low and mid-frequencies, the spectrum of music is such that an output of 80-watts at 20kHz would be more than sufficient to deliver all the high-frequency musical information at the correct playback levels. Frequency response was very good, with the amplifier reaching down to 4Hz (–1dB) and 2Hz (–3dB) in the low frequencies, and out to 39kHz (–1dB) and 70kHz (–3dB) at high frequencies. Within the audio band, the response was 20Hz to 20kHz ±0.1dB when driving a non-inductive load, and 20Hz to 20kHz ±0.25dB when driving a highly inductive (and capacitive) load that simulates that of a two-way bass reflex loudspeaker. Both these traces are shown on Graph 6. Channel separation was excellent at low frequencies (91dB at 20Hz), still excellent at 1kHz (90dB) but dropped down to 68dB at 20kHz. This is still far more separation than will ever be required for perfect stereo imaging and channel separation but it is a little lower than I might have expect from power amplifier. Inter-channel phase was appreciably better, coming in at 0.02° at 20kHz and swinging to just 5.95° at 20kHz. Channel balance was outstandingly good, one of the best figures I have seen, coming in at 0.007dB.

When driving 4Ω loads, the Naim NAP 250DR came within a whisker of meeting the theoretically ‘perfect’ ideal

Distortion at a power output of 1-watt was very low, as you can see from Graph 1 (8Ω load) and Graph 2 (4Ω load). At one watt into 8Ω, the NAP 250DR’s output spectrum contained a second harmonic distortion component at –98dB (0.0012%), a third at –95dB (0.0017%), a fifth at –107dB (0.0004%) and a seventh at –113dB (0.0002%). The peaks to the left of the graph are low-frequency noise components at 50Hz, 100Hz and 150Hz. The level of the second and third harmonic distortion components increased when the amplifier was driving 4Ω loads, but only to –90dB (0.0031%).

Driving 4Ω loads at 1kHz a fourth harmonic became obvious (at least visually!) at –114dB (0.0001%) while the fifth and seventh harmonics remained at approximately the same levels as when the amplifier was driving 8Ω loads. The levels of low-frequency components also rose a little, the result of the power supply having to work harder when the amplifier is driving 4Ω loads. The very low levels of distortion, combined with the complete absence of higher-frequency distortion components suggests to me that the amplifier would have a completely neutral tonal character in this regard.

Distortion levels increased when the NAP 250DR was operating at its maximum output power. At 80-watts into 8Ω, a 1kHz test signal resulted in a second harmonic distortion component at –85dB (0.0056%), a third at –90dB (0.0031%), a fourth at –95dB (0.0017%), a fifth at –96dB (0.0015%), a seventh at –104dB (0.0006%) and an eighth at 106dB (0.0005%). As you can see from Graph 3, higher-order components are also present, but they’re all around –120dB (0.0001%) or lower. These levels of distortion would be completely inaudible.

The same observation can be made for the NAP 250DR’s distortion into 4Ω loads at an output of 150-watts per channel. You can see that the amplifier is having a little more difficulty with the lower impedance, so that distortion is slightly higher, but apart from the second harmonic at –82dB (0.0079%) and the fifth at –86dB (0.005%), the first nine harmonics are all more than 90dB down (0.0031%) and most are down at –100dB (0.001%).

IMD was low, particularly the higher-order sidebands, which averaged around –90dB (0.0031%) for the first pair (18 and 21kHz), around –100dB (0.001%) for the second pair (17 and 22kHz) and –105dB (0.0005%) for the third pair of sidebands at 16 and 23kHz. The unwanted 1kHz regeneration was at –98dB (0.0012%). Ignore the noise at the left of this graph: it’s the result of a transient measurement error.

Newport Test Labs measured the signal-to-noise ratio of the NAP 250DR at 76dB unweighted and at 83dB A-weighted, referred to a one watt output. Referenced to rated output, these results improved to 94dB unweighted and 100dB A-weighted. These figures came on the back of a fairly high input sensitivity,

Newport Test Labs measured the signal-tonoise ratio of the NAP 250DR at 100dB A-weighted

with the Naim requiring only 105mV at its input for a 1-watt output, and only 934mV for rated power output.

Square wave performance was exemplary, with the 100Hz square wave showing barely any tilt (as would be expected from the nearly d.c. low-frequency extension). The 1kHz square wave, although not perfect, showed only modest rounding on the leading edge, with the rounding on the 10kHz wave being more evident but typical. Undoubted star of the show, however, was the square wave reproduction when the Naim was driving a highly capacitive load, which was completely atypical. There is a tiny amount of ringing, but it’s so negligible that it’s barely visible at all, so not only will this amplifier be stable into all loudspeaker loads, it should also ‘sound’ good in listening sessions, as amplifiers exhibiting this type of performance have in the past typically been rated very highly in subjective listening tests with all types of loudspeakers.

Damping factor was relatively low (40 at 1kHz) but even at this level is sufficient to handle the back-emf from large bass drivers. Mains power consumption at idle was 15.98-watts, increasing to 58-watts at a continuous output of 1-watt and topping out at 374-watts when the amplifier is operating at full power into 8 loads.

My opinion is that the Naim NAP 250DR performed very well on Newport Test Labs’ test bench. Steve Holding