The Audio Research VSi55 is a high-end integrated tube amplifier designed to deliver the warm and transparent sound that defines the American brand’s identity. Featuring a push-pull configuration with 6550C output tubes and a 6N1P driver stage, it combines the precision of modern engineering with the musicality of traditional tube designs. Its build quality is solid, the aesthetics are understated yet elegant, and its performance stands out for dynamics, detail, and control — making it a well-regarded amplifier among both experienced audiophiles and those seeking a refined introduction to tube amplification.
This tube amplifier arrived in my workshop after suffering an unusual fault: a screw had accidentally fallen inside the chassis, shorting the PCB traces. The short circuit caused the control chips managing input selection and volume to explode, also damaging the 6N1P input tube.
After replacing the damaged chips and checking the affected traces, the amplifier returned to full operation. I then performed several instrumental measurements to evaluate channel behavior and output transformer quality. Although overall performance was acceptable, I noticed an unexpected resonance at very low frequencies, likely caused by undersized magnetic cores in the output transformers.
Distortion spectrum at 15 watts
The 15-watt distortion test shows a behavior typical of a push-pull amplifier with moderate feedback, though higher-order harmonics appear slightly more pronounced than expected in a high-grade design. This points to a less-than-optimal transformer design or feedback network configuration.
1 kHz square wave at 15 watts: evident uncompensated ringing
The 1 kHz square-wave response reveals a clear ringing that should have been corrected during design through an appropriate snubber network or alternative feedback tuning. This behavior confirms the structural resonance previously observed in the frequency domain.
Frequency response on resistive load
The frequency response on a resistive load reveals a pronounced resonance around 70 kHz. Due to the low Q factor of the resonant element, the curve is broad and starts deviating as early as 10 kHz, affecting high-frequency linearity.
In the following plot, taken with a reactive load, peaks differing by nearly 5 dB indicate insufficient damping. For safety reasons, I avoided further testing at higher signal levels, not knowing the circuit’s stability limits and wanting to prevent destructive oscillations.
Overall, the amplifier performs well within the audible range: no visible waveform distortion is observed at high frequencies, and the sound remains balanced, though lacking the transparency expected from premium designs. It’s a good example of how even minor construction compromises or undersized transformer cores can influence an amplifier’s overall dynamic performance.