Threshold FET Nine/e preamplifier Measurements
Sidebar 2: Measurements
The frequency response of the FET nine/e is shown in fig.1. The lower curves show the response through the line-level stages taken at the main outputsand require no comment. The upper curves show the phono RIAA response taken at the tape monitor outputs. The deviations within the audio band are minor; the rising response seen at the extreme top end peaks above 100kHz. In an effort to track down any possible adverse effects from this, we fed a (pre-equalized) 10kHz squarewave into the Threshold's phono input (set for low gain). The resultant squarewave is shown in fig.2. The only visible artifact is a small, well-damped overshoot with no ringing evident.
The line-level crosstalk is shown in fig.3, which is effectively identical from left to right and from right to left. The phono input crosstalk (fig.4) is shown from left to right (top curve) and right to left (bottom curve). Somewhat asymmetrical, in that there appears to be less crosstalk from right to left than vice versa, except at high frequencies, this measurement is actually showing the phono input's noise floor. Note that fig.3 shows the common increase in crosstalk at higher frequencies due to capacitive coupling between channels.
THD+noise for the line stages is shown in fig.5 (100mV input), and remains very low across the audio band. Fig.6 gives the THD+noise for the phono stage, with the bottom curve indicating that for the low gain setting (5mV, moving-magnet and high-output moving-coil), and the top the high-gain setting (0.5mV input, low-level moving-coil). The increase in the latter is presumably due to the rise in the noise floor rather than to any lessened linearityremember that RD did note the preamp's phono stage noise to be higher than he expected.
The unity gain position of the Fet nine/e fell just below the 12:00 position of the level control. Output impedance was 100 ohms, give or take a few tenths of an ohm between channels and at various level-control settings. The input impedance of the line stages at 1kHz was, however, approximately 7800 ohms at full volume, increasing marginally to just above 8500 ohms at lower settingslower than spec. The output impedance of the apparently non-actively buffered tape-monitor jacks was 565 ohms, higher than specified. (Perhaps Threshold has changed the series buffer resistor value in production.) Gain measured 20.1dB from line in to main out (level control full-on), 37.5dB from phono input to tape output (low-gain setting on phono), and 54.3dB from phono input to tape output (high-gain setting on phono). The tape outputs showed a minimal 0.8dB drop for a line-level source into a 100k ohm load.
Phono overload levels (for 1% THD+noise) were moderate, at 38mV (low gain) and 5.48mV (high gain) at 1kHz, 4mV (low gain) and 0.58mV (high gain) at 20Hz, and 378mV (low gain) and 55.2mV (high gain) at 20kHz. All overload readings were taken using an unequalized signal into the phono stage. Both the line and phono stages of the Threshold were non-inverting. Main output DC offsets were 4mV (R) and 3.4mV (L).
Finally, to produce an output of 2Venough to drive all but a handful of amplifiers to full output and moreat 1kHz with the level control on full required an input of 200mV at the line inputs (all of which had identical gain), 2.5mV to the phono section set to low gain, and 0.357mV to the phono section set for high gain.Thomas J. Norton