EAR 324 phono preamplifier Measurements

Sidebar 3: Measurements

The EAR 324's gain in MM mode was a little lower than specified, at 50.6dB at 1kHz from both inputs, and was the same in both balanced and unbalanced output modes. The "-6dB" and "-12dB" settings reduced the gain by exactly those amounts. Switching in the MC transformer set to its "40 ohm" winding added another 19.5dB of gain, which will make the 324 suitable for use with low-output MC cartridges, particularly as its signal/noise ratios are respectably high: 82dB for the MM input (ref. 5mV input at 1kHz), 79dB for the MC input (ref. 0.5mV input at 1kHz). Both these figures are A-weighted; they worsen slightly when unweighted over a wide measurement bandwidth (to 75.4dB for the MM input and 66.3dB for the MC input), due to the presence of some low-level 120Hz hum that I could not eliminate by experimenting with the grounding between the 324 and my Audio Precision System One test set.

The preamp's output impedance was a usefully low 33 ohms from both the balanced and unbalanced output jacks, and it preserved absolute polarity with the front-panel button set to Normal (the XLRs are wired with pin 2 hot, the IEC standard). The MM stage's input impedance could be varied with the front-panel control. Set to "15k," it measured 16.7k ohms at 1kHz; to "22k," 23k ohms; to "33k," 42k ohms; to "47k," 64k ohms; and to "100k," 107k ohms. The MC stage's input impedance varied from 100 to 780 ohms at 1kHz, depending on the transformer winding used. I did not find any interaction between the phono inputs. However, the EAR was very fussy about the grounding between its inputs and the Audio Precision's outputs, with some low-frequency instability evident if the ground connection was not to its liking.

The EAR's RIAA error, measured through its MM input set to 100k ohms to minimize any interaction between the preamp and the signal generator, is shown in fig.1. Other than a small, 0.5dB rolloff at the frequency extremes, the response is very flat, with just a slight (0.2dB) emphasis in the lower treble. Channel separation, again measured through the MM input, was good in the treble, with the R-L leakage lower than in the opposite direction (fig.2), but degraded at lower frequencies, this presumably due to the action of the RIAA de-emphasis. (The input signal decreases with the frequency, meaning that the constant noise floor increasingly affects the measured separation.)

Fig.1 EAR 324, MM input, RIAA error at 1mV input at 1kHz (0.5dB/vertical div., right channel dashed).

Fig.2 EAR 324, MM input, channel separation (10dB/vertical div., R-L dashed).

Above 100Hz, with the level of the signal fed to the EAR 324's MM input set to 5mV at 1kHz, and pre-equalized with the appropriate RIAA curve, the THD+noise percentage remained close to 0.01% at all settings of the gain control (fig.3), which is superb performance. However, again there was some incompatibility between the EAR and the Audio Precision: I only obtained this low distortion if I measured each channel separately; with both left and right inputs connected to the AP, the THD rose to 0.06% in both channels.

Fig.3 EAR 324 line stage, MM input, THD+N (%) vs frequency at 5mV at 1kHz, right channel top, left bottom.

And whether I connected one or both channels, the measured distortion rose steeply with decreasing frequency below 100Hz, reaching our definition of "overload"—1% THD—just above 20Hz. I checked and rechecked this figure, using an unequalized signal and a different generator—a battery-powered Neutrik—with the same result. This means that the EAR's circuit has virtually no headroom in the low bass. While this may not have any subjective consequences with music recordings, which tend to have little energy below 40Hz (unless they are of pipe organs or synthesizers), it means that record warps will drive the preamp into nonlinear behavior. It should be noted that AD noted no audible problems with any of the cartridge/arm combinations he tried, but I was alarmed by this behavior regardless.

Moving higher in frequency, the 324's overload margins were only moderate, this correlating with the higher-than-normal MM-stage gain. I measured around 13dB through both the MM and MC inputs at 1kHz and 20kHz, this figure not affected by the front-panel gain setting. (The 13dB figure is referenced to standard 1kHz input levels of 5mV, MM, and 0.5mV MC.) This performance suggests that the EAR is best used with low-output cartridges.

Fig.4 shows the spectrum of the EAR's output while its MM input was driven with a 50Hz tone at 440µV; ie, just below reference level. The second harmonic is the highest in level, at -60dB (0.1%), followed by the third, at -64dB (0.06%). You can also see the 120Hz hum components I referred to earlier. Fig.5 shows a similar spectrum for a 1kHz tone at a high 15mV input level, again taken through the MM input. Even though this signal approaches the overload point, the harmonics are only just visible above the noise floor, the latter shaped by the RIAA de-emphasis. The EAR also turned in low levels of intermodulation distortion when reproducing high levels of 19 and 20kHz tones (fig.6).

Fig.4 EAR 324 MM input, spectrum of 50Hz sinewave, DC-1kHz, at 300mV into 100k ohms (linear frequency scale).

Fig.5 EAR 324 MM input, spectrum of 1kHz sinewave, DC-10kHz, at 1V into 8k ohms (linear frequency scale).

Fig.6 EAR 324, MM input, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1V into 8k ohms (linear frequency scale).

With the exception of its very limited low-frequency headroom, the EAR 324 turns in good measured performance provided it is used with low-output MM or MC cartridges through the appropriate inputs. But that LF performance is still a matter for concern.—John Atkinson

1087 East Ridgewood Street
Long Beach, CA 90807
(562) 422-4747