Columns Retired Columns & Blogs |
Rotel RC-850 preamplifier Measurements
Sidebar 2: Measurements
Fig.1 shows the frequency response of the complete preamplifier, measured via the phono inputs with the tone controls out of circuit. This is commendably flat between 40Hz and 30kHz, with any error within the tolerance of the inverse network. There is a faint hint of lift between 3kHz and 16kHz, however, which may tie in with the subjective nature of the treble. Integral infrasonic filtering curtails the response below 25Hz, this sensible in view of the less-than-ideal turntables likely to be used with this inexpensive component. The measured response rolls off slightly earlier than the specification, 3dB at 105kHz rather than 0.2dB at 100kHz, but I must say that this wide-open HF response on the phono input might be too susceptible to RF pickup in a region less of a radio-frequency desert than Santa Fe if there is any non-linearity present to demodulate the RF. Yes, the input NE5534 op-amps appear to be uncompensated, in order to maximize their gain-bandwidth product, but as this IC's open-loop gain drops from 55dB at 100kHz to less than 30dB above 1MHz, it is asking a lot of its intrinsic linearity at 100kHz and above to ask for extended bandwidth and high closed-loop gain.
Response via the line inputs was also extended in the highs, the 3dB point lying well above than 100kHz. (I measured 0.5dB at 102kHz.) The tone controls were gentle in their action, the maximum boost at 20Hz and 20kHz being 4dB or so, while the maximum amount of cut at those frequencies was between 6.5dB and 7dB. This gentle action is preferable to controls with a more aggressive action, in my view, for modifying system or program imbalances.
I found the input overload levels to be somewhat lower than specified, 1.9V RMS into the line inputs being sufficient to drive the preamp's output to its maximum 10.75V RMS level with volume control full up compared with the 5V specified. (If this seems too low to handle the maximum 2V RMS typical of a CD player, remember that the volume control precedes the active circuitry and lowering its setting will avoid the circuitry being driven into clipping.) As the specified overload level is given as just 5V, however, with no indication whether this is a peak, peak-peak or RMS level, I assume that it is a peak-peak 5V, which is equivalent to around 1.8V RMS, tying in with the 1.9V measured. Similarly, the phono overload at 1kHz is given as 180mV, whereas I measured around 65mV RMS, equivalent to 183mV peak-peak.
Phono input overload at 20Hz and 20kHz occurred at 30mV RMS and 430mV respectively, the latter actually due to the output stage clipping rather than due to the phono circuit overloading. There appears to be ample headroom, therefore, and it is extremely unlikely for the Rotel to have problems even with high-output MM cartridges.
Looking at interchannel crosstalk to see if there was any correlation with my impression of a less wide soundstage than ideally should be the case, the line-level inputs featured separation of more than 80dB at 20Hz and 1kHz, and 63dB at 20kHz. This is excellent performance, crosstalk at the first two frequencies being buried beneath the noise floor, and is better than many high-end products. Separation via either the MM or MC stages was only a little worse, however, at 70dB at 1kHz and below, this worsening to 56dB at 20kHz. All these figures are better than that offered by even an excellent pickup cartridge like the Linn Troika, so must be ruled out as being connected with the subjective impression of a rather narrow stereo stage. Noise on the phono input was quite low, measuring 72dB set to MM and 54dB, MC. Both figures are unweighted and improved by 45dB when A-weighted.
The output impedance measured close to spec at 1150 ohms, low enough for the RC-850 not to be too bothered by high-capacitance cables, though cables between it and the power amplifiers would best be kept reasonably short, 2m or so.John Atkinson
- Log in or register to post comments