Power delivery into difficult headphones

[Loquah]

I'm hoping you wonderful people can help me settle a question...

Let's say you have a difficult (headphone) load like a low impedance planar magnetic headphone with low sensitivity (60 ohm, 83dB at 1mW). You also have 3 amps:

• A Mainline which is maybe underpowered for the headphone given its 600mW power rating
• A tube-based transformer-coupled speaker amp with 8W per channel into 8 ohms
• A solid state speaker amp with 100W per channel into 8 ohms

When trying to drive these headphones with each amp at safe listening levels (e.g. 75-85dB), what would you expect to hear in terms of headphone performance? Would any of these amps provide significant issues or benefits at these listening levels?

For reference, my rough calculations say that this headphone requires:

• 2mW (0.35V and 5.8mA) to reach 85dB
• 51mW (1.75V and 29.2mA) to reach 100dB

[Paul Birkeland]

I would wonder why a 60 ohm headphone that needed 100mW of power would be considered hard to drive? 

[Loquah]

I would wonder why a 60 ohm headphone that needed 100mW of power would be considered hard to drive?

I totally agree and it's the reason for my question. I was trying to avoid naming the headphone to avoid any bias in responses based on its reputation, but that headphone is the HiFiMan Susvara. I recently posted a review describing how I was able to easily drive it to great sound quality from basically every decent headphone amp in my collection.

My challenge is that there are factions in the hobby who are adamant that you can't hear the best of this headphone until it's connected to a powerful speaker amp (such as option 3 in my original post). The argument proposed by this group is that whilst some amps might be quoted to deliver say 2W at 60 ohms, they are actually right at the edge of their performance capabilities and therefore distorting heavily. My response to this is that normal listening levels should require nowhere near those peak levels.

Is there any truth to the idea that a planar magnetic headphone such as this will require significant current capabilities (beyond the sorts of numbers I calculated) to deliver sound effectively across all frequencies at normal listening levels? I can't see the maths suggesting that to be the case and my listening tests align, but it's a hard-held belief and I want to make sure I'm not missing something.

My thinking is this:

• Safe listening levels are around 85dB
• Headphone needs 2mW to achieve 85dB
• A decently powerful amp still has >15dB of headroom even if there are peaks of 90-100dB in the music
• Therefore, power requirements at normal levels really shouldn't be a concern so long as you have greater than a few hundred mW available at 60 ohms and particularly if you have >1W available

The argument goes that these headphones sound fuller and better from more powerful amps. My belief is that they are just hearing the qualities of the amp and attributing it to power rather than amp quality.

Here's a link to my video for reference: Jump to 11:28 for the discussion of power

[Paul Birkeland]

My challenge is that there are factions

I think you understand the issue better than you are letting on

The Susvara according to Hifi News, "its impedance vs. frequency is practically constant – a minimum of 59.3ohm and a maximum of 59.4ohm – so the Susvara is essentially immune to frequency response modification caused by the finite source impedance of the partnering headphone amp"

But at the same time: "the Susvara is the least sensitive headphone we have ever measured. The Sundara and Arya were already on the low side at 105.7dB SPL and 103.5dB SPL, respectively, for 1V input at 1kHz, but the Susvara is way lower at 91.5dB, averaged for the two capsules. No wonder it includes no provision for connecting to portable music players."

Maybe there is something to that last statement.  1V into 60 ohms is 16mW, so 16mW to get 91.5dB, so maybe you'd want 250mW at 60 ohms, so 4V or so of output.   These might even not sound all that bad on a Crack.

[Loquah]

I think you understand the issue better than you are letting on

The Susvara according to Hifi News, "its impedance vs. frequency is practically constant – a minimum of 59.3ohm and a maximum of 59.4ohm – so the Susvara is essentially immune to frequency response modification caused by the finite source impedance of the partnering headphone amp"

But at the same time: "the Susvara is the least sensitive headphone we have ever measured. The Sundara and Arya were already on the low side at 105.7dB SPL and 103.5dB SPL, respectively, for 1V input at 1kHz, but the Susvara is way lower at 91.5dB, averaged for the two capsules. No wonder it includes no provision for connecting to portable music players."

Maybe there is something to that last statement.  1V into 60 ohms is 16mW, so 16mW to get 91.5dB, so maybe you'd want 250mW at 60 ohms, so 4V or so of output.   These might even not sound all that bad on a Crack.

 

Thanks Paul. Sounds like I had the right idea from a technical point of view, but also perhaps that they are marginally harder to drive than the published specs. Still, nothing here that seems to scary for an amplifier's power delivery capabilities I gather...

[Paul Joppa]

I suspect that there is some confusion in the marketplace among volts, watts, and impedance. Four volts at 60 ohms is about a quarter of a watt, but into 4 ohms it's 4 watts. So a rating of 4 watts or more is a minimum specification if you are using a speaker power amp, even though you are actually using only 1/4 of a watt.

[Loquah]

I suspect that there is some confusion in the marketplace among volts, watts, and impedance. Four volts at 60 ohms is about a quarter of a watt, but into 4 ohms it's 4 watts. So a rating of 4 watts or more is a minimum specification if you are using a speaker power amp, even though you are actually using only 1/4 of a watt.

I wondered about this too, Paul. Thanks