The M³ stereo headphone amplifier

Specifications & benchmarks

All measurements were done with an M³ amplifier with the following configuration (except where noted):

Test conditions
Power supply STEPS v1.2 regulated DC power supply, set to 27.0V DC
MOSFET quiescent current set to 85mA on all three channels
JFET cascode current source set to 5mA on all three channels
Output dummy load 33Ω 5W resistor between each stereo channel to output ground
Volume control50KΩ Alps RK27112A set to maximum
Bass boost control equipped but set to no boost
Opamps AD8610AR on all three channels
All resistors except R9 Vishay-Dale CMF RN55D
R9 Xicon MO
R4 / R3 4.02KΩ / 1.0KΩ (for voltage gain = 5)
C1 compensation capacitor 22pF Kemet Golden-Max C315 NP0/C0G
C4 capacitance multiplier capacitor 100µF Nichicon UHE
C5 opamp rail reservoir capacitors 100µF Nichicon UHE
C6 opamp rail bypass capacitors 0.1µF Wima MKP2
C7 rail reservoir capacitors Three 1000µF Nichicon UHE
C8 rail bypass capacitors 1µF Vishay-Roederstein MKT1817
All other parts as specified in schematic diagram

The tests were conducted using the following equipment:
  1. Wavetek 188 4MHz sweep function generator
  2. Protek 6510 100MHz dual trace oscilloscope
  3. Fluke 95 50MHz digital dual trace ScopeMeter
  4. Beckman HD100 digital multimeter
  5. Toshiba Satellite A65 laptop computer with Celeron 2.8GHz CPU and 768MB RAM
  6. M-Audio Firewire Audiophile mobile interface
  7. RightMark Audio Analyzer 5.5 software
  8. dummy loads 330Ω/33Ω/8Ω/none
  9. other custom test jigs, cables and software

Here are the measured results:

Measured resultsEquipment used
Input impedance 47.6KΩ 1, 2, 3, 4, 8, 9
Output impedance less than 0.05Ω 1, 2, 3, 4, 8, 9
Maximum output voltage
per channel, 1KHz, prior to onset of clipping
AD8610, 27V PSU: 13Vp-p (4.6Vrms)
OPA604, 40V PSU: 22Vp-p (7.8Vrms)
1, 2, 3, 4, 8, 9
Power output AD8610, 27V PSU: 0.6Wrms into 33Ω, 2Wrms into 8Ω
OPA604, 40V PSU: 1.8Wrms into 33Ω, 6Wrms into 8Ω
1, 2, 3, 4, 8, 9
Frequency response at 0dBV output:
0Hz - 1.5MHz, +0dB -3dB
1, 2, 3, 8, 9
Rise time 10KHz square wave, at maximum output, 10% to 90%:
1, 2, 3, 8, 9
Slew rate 10KHz square wave, at maximum output:
1, 2, 3, 8, 9
Noise level
Dynamic range
Total harmonic distortion
Intermodulation distortion
Stereo crosstalk
see RMAA test results below 5, 6, 7, 8, 9

RMAA test results

Several tests were performed, one with the M-Audio Firewire Audiophile mobile interface running in loopback mode to establish the baseline (lowest limit of measurement resolution), and anothers with the M³ amplifier added to the chain with 330&Omega, 33Ω or 8Ω dummy loads connected to the M³'s output.

The M³'s volume control knob is set to within 3dB of maximum for these tests. This is worthy of mention because many others test their amplifiers with the volume set lower, which would give artificially low noise and distortion results.

Click on the following links to see the results.

Comments on the RMAA test results

Even though the M³'s measured results are excellent, it can be seen from the loopback graphs that the M³ added almost negligible distortion to the M-Audio Firewire Audiophile's baseline. Hence, the distortion measured is predominently the sound interface's performance, not that of the M³ amplifier.

The slight rolloff at the low end in the frequency response graph is due to the M-Audio's coupling capacitor. The M³'s actual frequency response is flat down to 0Hz. Likewise, the high-end rolloff in the graph is due to the M-Audio's anti-aliasing filters. M³'s actual response extends to 1.5MHz (-3dB) in the tested configuration.

The small spike in the M³'s noise floor near 7KHz is a measurement aberration due to interference.

The M³'s RMAA results remain essentially unchanged whether its outputs are loaded with 330Ω, 33Ω or 8Ω. The stereo crosstalk is the only item slightly affected at 8Ω load, but even with this severe load, the results are still better than most headphone amplifiers driving 330Ω loads. Many amplifiers' distortion performance deteriorate significantly with lower impedance loads, there is no such problem here.

These results amply illustrate the benefits of M³'s high-current class A MOSFET output stage, three-channel active ground topology and other design elements. See the Technical highlights section for details.

Oscilloscope waveforms

The following shows the waveform response of the M³ amplifier. In all graphs except the Lissajous waveform, the top trace is the input and the bottom is the output.

The square wave graphs show that there is minimum slewing and ringing at the leading and falling edges. The 100KHz sine, triangle, and Lissajous graphs also show very small amount of phase shift between the input and output. Within the 20Hz to 20KHz audio band, there is no measurable phase shift.

The output waveforms are also unaffected by the load. There is no discernable change in shape or amplitude, whether the amplifier output is unloaded, loaded with 330Ω or loaded with 33Ω.

Main: M³ Main | Prev: Other options | Next: History