CanJam at RMAF 2015: Apogee Groove with Constant Current Drive

Current output headphone amplifiers are popping up all over the place these days. It's a technology with potentially significant advantages, but also significant drawbacks. The Apogee Groove ($295) sounds like it may have rid itself of some of the problems.

The magnetic field in a headphone voice coil is directly proportional to the current flowing through it...and not the voltage applied. With a voltage step going into the amplifier, a voltage amplifier will put out a similar step, but because of the reactive impedance the current waveform will lag as it puts energy into expanding the magnetic field around the voice coil. In a current amplifier, the output current will (almost) immediately rise to the proper level, but the voltage will spike significantly to respond to the initially high reactive impedance.


A pure current drive amplifier (transconductance amplifier), however, has a significant drawback when driving headphones and speakers. Because the current is always proportional to the input voltage, its voltage output will change proportionally with the impedance of the load. Many headphones have impedance plots that significantly change with frequency. For example, the Sennheiser HD 800 has a nominal 300 Ohm impedance, but at 90Hz the primary driver resonance increases the impedance to 600 Ohms. With a current drive amp, the voltage will double at those frequencies creating a significant increase in the bass heard at those frequencies. This is why it was so easy for Big Sound participants to identify the Bakoon current amplifier in blind tests—the bass was significantly boosted.

Apogee claims to have overcome this problem with the Groove, which includes some sort of impedance sensing circuit in the design. Unfortunately, this technology is confidential at the moment until patents are in process. I did, however, have the opportunity to hear the HD 800 at the show, and indeed it did not seem to exhibit this problem.

I've long thought that current drive topologies might be a very good thing in audio (Nelson Pass has done some, which include compensating networks to solve some of the problems) as it may dramatically improve articulation in the high frequencies. While the Bakoon current drive amp had problems with tonal balance due to varying headphone impedance, it also seemed to make the treble smoother without loosing articulation.

This is a subject you can expect me to investigate and explain further over the coming months. I do have a Groove in-house, as well as the Bakoon HPA-21 and Erzetich Perfidus, which, as I understand it, is also a current source amp.

gixxerwimp's picture

But someone should tell them that the hip guy with tattoo and skull cap on their website "working on music" is wearing the HD 800 backwards.

Tyll Hertsens's picture
I think I actually commented on that when they posted the pic on Facebook. Also, HD 800s are not good on an airplane. :)
cspirou's picture

I suspect that headphones these days are designed with the assumption that they will be used with a low output impedance amp instead of a transconductance amp. If current output amps have been used the entire time then headphones would have followed suit.

For single driver full range speakers I don't think this is an issue because the bass is inherently weak for a full range driver and speaker drivers are usually designed with the assumption of some sort of passive crossover compensation.

TheAudioGuild's picture

That's correct. I'm not aware of anyone designing dynamic headphone drivers intended to be driven by anything but a voltage source. So I can't imagine why anyone would want to design a headphone amp with a current source output except as a marketing gimmick (read "it's different").


bogdanb's picture

My knowledge about physics and electricity is almost 0. This story made me think about power sources in electronics now capable to adapt to 110 or 220. Wouldn't that be case with transconductance amplifiers? can they get a automatic reading of the impedance and adapt accordingly?

Tyll Hertsens's picture
Usually those are switching supplies that sense the incoming voltage. No relation to transconductance amps.
bogdanb's picture

What I had in my mind was a small impedance meter inside that would switch between "stuff" inside...
I had in mind the Multimeters that can measure amperage or voltage and display a value.
Would it be very complicated to have different voltage output behaviour based on measured impedance.

That if the headphone impedance is possible to be measured at the end of the cable (at the 3,5 jack).

Again I might be very far off the reality since I really don't understand well physics. (I had bad teachers)

bogdanb's picture

They are after all a company working with the professional side of music, for the audio industry(I would call that). Have you got a chance to ask their opinion about Harman's Target Response Curve? how would that work for a studio monitor headphone?
Does it make sense to adapt a studio monitor headphone to what, by my understanding, is the eq for flat sound when listen from a headphone?

Tyll Hertsens's picture
Really can't do that from the amp side as there's no way to sense the acoustic frequency response from the amp side.
bogdanb's picture

I was just curious in the audio professional world how that is perceived... what is the echo of Harman's idea in the professional world. Especially because they don't make headphones. Let's say like a referee p.o.v.

Tyll Hertsens's picture
Not tuned into the pro world well enough to know, but I doubt they're much aware of it yet.
steaxauce's picture

Tyll, if you get a chance, it'd be great to see measurements of the HD800 when plugged into the Groove. I'd like to to find out how their frequency response, square wave response, etc. when powered by the Groove compares to what you get driving them with a standard voltage amp.

On Apogee's FAQ page for the Groove, they actually advise against using it with multi-BA IEMS. From Apogee's FAQ: "Apogee does not recommend the Apogee Groove for use with multi-driver balanced armature in-ear monitors. Due to the design of the balanced armature drivers and crossover networks used in this type of headphone, the Groove’s Constant Current Drive amplifier technology may result in uneven frequency response when used with certain models."

I assumed their warning was because of the variable impedance of multi-BA IEMS, in which case it would apply to the HD800 as well. But maybe it's more complicated than that?

Tyll Hertsens's picture
I can't do it quite as comprehensively as my normal headphone measurements because my AP tester does not have a USB output. But I will do some measurements of FR using pink noise as source and square wave response.

thune's picture

I'm unconvinced 'impedance sensing' is more than marketing-speak that resulted from attempts to promote the benefits of current drive.