Mesa Tigris integrated amplifier Measurements
A complete set of measurements of the Mesa Tigris was made in the two-thirds pentode, Stage I feedback configuration. Additional selected measurements were made with other settings, but, unless noted otherwise, these results refer to operation in two-thirds pentode/Stage I.
The Mesa's voltage gain at maximum level measured 40.3dB. For the measurements I set the level control somewhat lower, at a position that produced 34dB gain (approximately 2:00 o'clock with Stage I feedback/two-thirds pentode). The amplifier's gain depends heavily, however, on its operational mode. For example, with a fixed level control setting, two-thirds pentode, and Stage 0 feedback taken as a reference, switching to Stage I, II, and III results in gain losses of –3dB, –5.7dB, and –8.5dB, respectively. Switching from pentode to two-thirds pentode/Stage III results in a gain drop of 1dB. When evaluating the sound of the Tigris with various settings in a dealer's sound room, you should be careful to take these level changes into account before drawing any conclusions.
The Tigris' input impedance measures a high 62k ohms, the DC offset a negligible 0.1mV left, 0.3mV right. The amp is noninverting if you connect your speaker leads with the positive lead at the common, red-coded output terminal (using the more conventional connection, where common is negative, the amp is inverting). The unweighted S/N ratio ref. 1W into 8 ohms, left channel only measured, was 62dB over a 22Hz–22kHz bandwidth, 62dB from 10Hz to 500kHz, and 77dB A-weighted.
The Tigris' output impedance varied radically depending on its mode of operation. I measured 2.7 ohms at 1kHz, pentode/Stage III feedback, and 7.5 ohms at 20kHz, two-thirds pentode/Stage I. I didn't measure all combinations, but you can reasonably expect pentode/Stage III to offer the lowest output impedance. At all settings, however, the output impedance of the Tigris is high enough to have a substantial effect on the amplifier's frequency response with any real-world loudspeaker,
I also measured the output impedance from the Tigris' preamp outputs. At its maximum setting, the reading was 15.9k ohms. This will not be a good match for most consumer power amplifiers, even tube designs. At the control's middle setting, the value dropped to 3.2k ohms—still high, but usable. (I recommend an amplifier input impedance at least 10 times the value of the output impedance of the preamp feeding it.) At 9:00 o'clock the output impedance dropped below 1k ohms, but the output level is very low at that point: 25dB below maximum.
The output impedance at the tape outputs measured 25.5 ohms with a 25 ohm source impedance and 588.5 ohms with a 600 ohm source impedance, indicating buffered tape outputs. The tracking of the main level control is fair; I measured a maximum deviation of 0.38dB (right higher than left) at a 10:00 setting of the control.
Fig.1 shows the Mesa Tigris' frequency response. The high-frequency response into a resistive load falls off rapidly in two-thirds pentode/Stage I feedback, by about –4.7dB at 20kHz; the bass rolls off too, but less seriously. The response holds up better in full pentode/Stage III (not shown, but approximately –2dB at 20kHz, +0.5dB at 20Hz). More significant is the Tigris' performance with a simulated two-way loudspeaker load. The bass peaks at 70Hz, and the dip at about 5kHz will certainly be audible as a noticeable softening of hard transients. To some degree most tube amplifiers exhibit this sort of behavior, which is the result of high output impedance, but the Tigris is about the most extreme case I have seen. Fig.2 shows that the frequency response aberrations can be reduced significantly by going to Stage III feedback. If you hear a sweeter sound with Stage I, this is most certainly due to the sort of response variations shown here, not some exotic effect from the lower feedback setting, in my opinion.
Fig.1 Mesa Tigris, frequency response at (from top to bottom): 1W into 8 ohms, 2W into 4 ohms, and 2.828V into simulated loudspeaker load (0.5dB/vertical div., right channel dashed).
Fig.2 Mesa Tigris, two-thirds pentode, frequency response at (from top to bottom): 2.828V into simulated loudspeaker load with Stage III feedback and Stage I feedback (0.5dB/vertical div.).
Fig.3 shows the Tigris' reproduction of a 10kHz squarewave. The waveform is no longer recognizable as a squarewave, most of the change due to the severe drop in high-frequency performance. The 1kHz squarewave (not shown) is better, with just a small tilt at the top of the waveform.
Fig.3 Mesa Tigris, small-signal 10kHz squarewave into 8 ohms.
The amplifier's channel separation (not shown) is adequate at 50dB/55dB (L/R) at 1kHz, decreasing to 32dB/35dB at 20kHz, but this is still mediocre performance for a modern amplifier.
Fig.4. shows the THD+noise percentage vs frequency—by today's standards, a poor result. The right-channel response is significantly better than the left-, however, suggesting either that the tubes are mismatched or the bias is nonoptimum for the left channel. (There are no user-accessible bias adjustments on the amplifier.) I also measured the distortion at the preamp outputs. It shows a similar, significant increase at low frequencies (8.5% at 20Hz left, 1.8% at 20Hz right), suggesting that using this output to drive a powered subwoofer—an option suggested in the review—could have dubious results.
Fig.4 Mesa Tigris, two-thirds pentode and Stage I feedback, THD+noise (%) vs frequency at (from top to bottom at 1kHz): 4W into 2 ohms (4 ohm tap), 1W into 8 ohms (8 ohm tap), and 2.83V into simulated loudspeaker load (4 ohm tap) (right channel dashed).
Fig.5 shows how the distortion varies with changes in the feedback setting. Not surprisingly, higher feedback lowers the distortion, though the change is not likely to be audible. The 1kHz THD+noise waveform at an output of 2W into 4 ohms is shown in fig.6. The second harmonic dominates, though there is also clear evidence of higher harmonics being present.
Fig.5 Mesa Tigris, two-thirds pentode, THD+noise (%) vs frequency at 2W into 4 ohms (4 ohm tap, from top to bottom at 1kHz): with Stage 0, Stage I, and Stage III negative feedback (right channel dashed).
Fig.6 Mesa Tigris, 1kHz waveform at 2W into 4 ohms (top), distortion and noise waveform with fundamental notched out (bottom, not to scale).
The distortion spectrum resulting from a 50Hz input at 16W into 4 ohms is shown in fig.7. The second harmonic is at –28.4dB, or about 4%. The distortion shown should result in a rather warm sound—possibly pleasing, but inaccurate. The 19+20kHz IM spectrum at 11W into 4 ohms—the highest output with this demanding signal before visible signs of clipping start to appear on an oscilloscope trace—is plotted in fig.8. The 1kHz intermodulation artifact is at –29.5dB (about 3.2%), the 18kHz artifact is –28.9dB (about 4%)—neither figure is particularly noteworthy. The 19+20kHz spectrum at 11.8W into 8 ohms (not shown) is very similar to the result shown in fig.8.
Fig.7 Mesa Tigris, 4 ohm tap, spectrum of 50Hz sinewave, DC–1kHz, at 16W into 4 ohms (linear frequency scale).
Fig.8 Mesa Tigris, 4 ohm tap, HF intermodulation spectrum, DC–22kHz, 19+20kHz at 11W into 4 ohms (linear frequency scale).
The Tigris' THD+noise percentage vs level curves are shown in fig.9. The output powers at clipping (relaxed from our usual 1% THD+noise to 3%) are shown in Table 1. Changing the operating mode to full pentode/Stage III increases the 3% THD+noise clipping point at 1kHz into 8 ohms (8 ohm tap) to approximately 37W (27W at 1% THD+noise). John Atkinson measured the Tigris' maximum power output using the Miller Audio Research Analyzer, which uses a test signal more representative of a music signal: a 1kHz toneburst with a duty cycle of 10 cycles on, 40 cycles off. The maximum burst power was not that different form the continuous power figures: 25.6W into 8 ohms from the 8 ohm tap, 25.1W into 4 ohms from the 4 ohm tap; 20.1W in to 2 ohms from the 2 ohm tap (all figures for 3% THD+N). However, 28.9W was available into 8 ohms from the 4 ohm tap, suggesting the latter is optimized for a somewhat higher load than 4 ohms.
Fig.9 Mesa Tigris, distortion (%) vs continuous output power into (from bottom to top): 4 ohms (4 ohm tap), 8 ohms (8 ohm tap), 2 ohms (4 ohm tap).
Fig.10 shows how the THD+noise percentage with this toneburst signal changes with output level: the blue trace is with pentode operation and Stage I feedback and shows that 40W is available into 4 ohms in this configuration. By contrast, two-thirds triode operation (black trace) gives just 15W into 4 ohms for the same 3% THD+N limit. Fig.11 shows how the differing degrees of negative feedback affect the THD+N vs output power behavior. Only with Stage III feedback (black trace) does the distortion remain at a level generally regarded is inaudible over most of the power band.
Fig.10 Mesa Tigris, 4 ohm tap and Stage I feedback, distortion (%) vs burst output power into 4 ohms with two-thirds triode operation (black trace), two-thirds pentode operation (red), and pentode operation (blue).
Fig.11 Mesa Tigris, 4 ohm tap and two-thirds pentode operation, distortion (%) vs burst output power with Stage III feedback (black trace), Stage II feedback (red), Stage I feedback (blue) and no feedback (green).
The Mesa Tigris' technical performance is poor. If you fall in love with its sound, you must do so despite its measurements, because they will provide you with no reassurance. Given the right loudspeakers, the Tigris' low power is not necessarily a negative, but its frequency-response deviations into a real-world load and its high distortion do not, in my opinion, qualify it as a true high-fidelity device. But if you like this sort of sound, you'll likely consider the measurements irrelevant anyway.—Thomas J. Norton
Table 1 Mesa Tigris: Clipping
(3% THD+noise at 1kHz, all with a 119V line)
(The 8 ohm figures were from the 8 ohm tap; the 4 and 2 ohm figures from the 4 ohm tap)
|Both Channels Driven||One Channel Driven|
|Load||W (dBW)||W (dBW)|
|8||23.9 (13.8)||23.4 (13.7)||23.9 (13.8)|
|4||24.6 (10.9)||23.3 (10.7)||24.5 (10.9)|