Measuring the MOH

I've used Sennheiser HD600 headphones for about 15 years. They're great headphones but they have some limitations for serious listening.

  • Extreme bass is attenuated; -10 dB at 20 Hz
  • Mid bass is boosted; +4 dB at 100 Hz
  • Extreme treble is attenuated; -12 dB at 20 kHz
  • But before I nitpick them, let me repeat these are great headphones. They have incredibly low distortion, smooth frequency response and crystal clear midrange. This makes them great for most acoustic music. Their limitations are minor; but after listening to my Magnapan 3.6 speakers (which I've also owned for 15 years) the limitations of the HD600, however small, seem accentuated. In comparison, the HD600 sound warm and dark with mid-bass emphasis coloration. I enjoyed the HD-600 but always knew what I was missing and preferred the Maggies for serious listening.

    However, I didn't know where to find higher quality in a headphone. There are more expensive headphones, but they don't necessarily sound better or have better measurements. Then I read about the HiFi Man HE-500. It's a planar magnetic headphone - like miniature Magnepans. Everything I could find about this headphone said it was excellent: from The Absolute Sound, to Inner Fidelity, to CNet. I had to give it a try!

    Efficiency (or Lack Thereof)

    Problem was, the HE-500 is less efficient than the HD-600. To play at the same volume level, the HE-500 needs:

  • 1.35x more voltage (2.6 dB quieter at the same volume/voltage level)
  • 11x more current
  • 16x more power
  • I have a great headphone amp, the Maxed Out Home from Headroom. No doubt it could drive the HiFi Man HE-500 with no problem. But it would need to drive 2 of them in parallel! Why? Because my wife and I use headphones to listen together in evenings. We need a pair with equal efficiency so we hear at the same volume. I could stick a resistor inline with her HD-600, but if I like the HE-500, I'm sure she will too. So now the question was, does the MOH have enough power to drive 2 inefficient headphones in parallel?

    I contacted Headroom, but the amp is so old they didn't have any schematics or specs. I didn't want to spend over a thousand dollars on a pair of headphones I couldn't use, so I had to test it myself. Here's how I did it.

    Testing the MOH

    All I had for testing is this computer with a sound card and software (Audacity). Problem is, the max input voltage to my sound card is 2Vpp (1.414Vrms). I couldn't test any higher voltage else it would overload the sound card. At that voltage, the HE-500 would play at:

    HE-500 efficiency: 0.31 Vrms @ 90 dB SPL (nominal 38 ohms, measured 47 ohm impedance)
    10 * log(1.414/0.31) = 6.6 dB
    90 dB + 6.6 dB = 96.6 dB.

    That's louder than a normal listening level, but transient dynamic peaks could be louder.

    I knew the MOH could easily produce more than enough voltage. The question was whether it could deliver enough current to swing that voltage over the low impedance load of 2 inefficient headphones in parallel. What it boiled down to was: how much current could the MOH deliver?

    I had some Dale metal film resistors lying around from another project, so I put them to use. I wired an unbalanced RCA cable with several resistors in parallel across the load. I was able to clip on resistor combinations to test the following impedances: 26, 12.7, and 8.5 ohms. I figured if the MOH could swing 1.414Vrms over an 8.5 ohm load, it must have the current capability to swing much bigger voltages over the 20 Ohm load that 2 HE-500s in parallel would represent.

    I played the 500 Hz test tone from Stereophile test disc 2 into the MOH, connected the MOH headphone output to my sound card line input using a normal cable (not my custom cable), recorded with Audacity, and adjusted the volume knob on the MOH until the level was just below clipping. I measured "just below clipping" by taking a spectrum analysis of the signal. When the sound board clips because the input voltage is too high, there are big obvious high frequencies in the spectrum. I used the highest volume setting that had a clean spectrum. By "clean" I meant all distortion artifacts below -80 dB. This is the distortion from the CD player, sound card and MOH combined. The MOH itself has much lower distortion than that.

    Next, I connected the MOH to the sound card with my custom cable and took the same measurement. I found there was virtually no difference in the distortion all the way down to 8.5 Ohms. Actually, the 2nd harmonic at 1500 Hz was 6 dB higher at 26 Ohm, same as 8.5 Ohm, yet still at -80 dB, which is negligable. What this means:

    i = V / R; i = 1.414Vrms / 8.5Ohm = 0.166 = 166 mAmp
    P = V*I = 1.414Vrms * 0.166 = 235 mWatt
    P = i^2*R = 0.166 * 0.166 * 8.5 = 235 mWatt

    So the MOH can output 235 mWatt and 166 mAmp of current with no distortion. It can probably output more, I just don't know how much more.

    But this should be plenty to drive dual HE-500s in parallel. Here's why: What would happen if the MOH delivered that same 166 mAmp into a 20 Ohm load (20 Ohms is dual HE-500s in parallel)?
    Vrms = 0.166 A * 20 Ohm = 3.32 Vrms
    Now the HE-500 play at 90 dB SPL with 0.31 Vrms. How loud would 3.32 Vrms be?
    20 * log(3.32/.31) = 20.6 dB louder
    90 dB + 20.6 dB = 110.6 dB SPL.
    That is ear-bleeding loud. And even at this level, the MOH is cruising along with negligable distortion and power to spare.