EHHA I Performance


Oscilloscope

Measurements were taken using the following test conditions:

  • +/- 23V DC rails (lab supply max)
  • 6.3V DC heaters
  • 101mA MOSFET output current
  • 50Ω load
  • 20 GS/S Tektronix digital oscilloscope
  • HP RF signal source

Input traces are on Channel 2 in blue while output traces are on Channel 1 in yellow. Data was taken at three gain settings: G=+1 / C2=C3=56pF, G=+4.3 / C2=C3=15pF, & G=+7.8 / C2=C3=open. Each was compensated for no (minimal) overshoot or ringing with 100kHz square waves.

The purpose of these measurements was to charcterize the large signal attributes of the EHHA. Plots are shown for sinusoidal inputs at 100kHz, 1MHz, and the 3 dB bandwidth. Additionally, square waves at 100kHz are shown, along with close-ups of both the rising and falling edge settling characteristics. What ringing there is on the input is due to the lead inductance of the hookup cable. Finally, the full scale output swing is shown to be within 4V (the Vgs of the output Mosfets) of each rail with a 10 kHz triangle wave chosen so that clipping/distortion would be visually obvious. With the +/- 23V supplies, 38Vp-p maximum output swing was measured.

Gain=+1 / C2=C3=56pF

100kHz 8.25Vp-p sine
1MHz 8.4Vp-p sine
1.5MHz 8.4Vp-p sine
100kHz 9.2Vp-p step
100kHz 9.2Vp-p step rising edge
100kHz 9.2Vp-p step falling edge

Gain=+4.3 / C2=C3=15pF

100kHz 2.0Vp-p sine
1MHz 2.0Vp-p sine
1.9MHz 2.0Vp-p sine
100kHz 2.0Vp-p step
100kHz 2.0Vp-p step rising edge
100kHz 2.0Vp-p step falling edge

Gain=+7.8 / C2=C3=open

100kHz 2.0Vp-p sine
1MHz 2.0Vp-p sine
1.96MHz 2.0Vp-p sine
100kHz 2.0Vp-p step
100kHz 2.0Vp-p step rising edge
100kHz 2.0Vp-p step falling edge
10kHz 4.0Vp-p triangle

Slew Rates

Slew rates for the three gain settings above can be obtained from the square wave traces.

Gain C2, C3 Slew Rate
+1 56pF 20V/μs
+4.3 15pF 50V/μs
+7.8 - 100V/μs

In summary, what these plots show is that the EHHA performs much better than is needed for audio. In each case the EHHA has more than 1.5 MHz 3 dB bandwidth large signal (>8Vp-p output swing) bandwidth. In fact, it's roughly 2 MHz at gains of +4.3 and +7.8. For a discrete amplifier, this is quite fast. For a tube amplifier, it is extremely fast. Moreover, the slew rate is well beyond what is needed for audio and will not be affected by headphone cable capacitance. What makes this speed more impressive is the nearly perfect 100 kHz square wave response that exhibits no peaking or ringing. Given these measurements with a 50 Ohm load, driving headphones in the audio band will be easy for the EHHA.

Audio Precision

Measurements were taken using the following test conditions:

  • +/- 30V σ22 power supply
  • 6.3V DC heaters
  • 100mA MOSFET output current
  • 50Ω load
  • Gain=+4
  • C2=C3=open

This is a prototype PCB with mosfet output devices fully cased with a rotary switch input selector switch, an Alps "Blue Velvet" potentiometer, RCA inputs, and a Neutrik output jack. Shown are the 1kHz sine wave spectrum, the 1kHz dyanamic range, the THD+N vs input amplitude, and the frequency response. In spite of the large 19" rackmount case and maximized spacing between the transformers and EHHA boards, the noise floor is limited by transformer electric fields. Better results would have been acheived with the power supply in a separate case.

1kHz Spectrum 1kHz Dynamic Range
Frequency Response THD+N vs. Amplitude

DC Offset

Over a variety of different conditions, including both open air as well as inside a ventilated 19" rackmount case, using Analog Devices OP07 servo opamps, DC offset settles to roughly 1mV or less. The opamp output voltage was measured to be in the range of 0.8V to 3V DC.