The "0-Ohm" Headphone Amplifier

Transcription

1 The "0-Ohm" Heaphone Amplifier The Sonic Avantages of Low-Impeance Heaphone Amplifiers y John Siau December 2011 The circuits use to rive heaphones are often ae to a prouct without careful consieration of the ifficult loas presente by high-quality heaphones. The most common circuit is an opamp river followe by a 30-Ohm series resistor. The series resistor provies short-circuit an overloa protection while isolating the opamp from the inuctance an capacitance of the heaphones. The series resistor protects the opamp while keeping it stable. In contrast, toay's state-of-the-art heaphone amplifiers eliminate the series resistor, an use a high current river. This change reuces istortion an flattens the frequency response when a heaphone is riven. These new high-en esigns are often calle "0- Ohm" heaphone amplifiers, an are essentially miniature power amplifiers. This paper provies measurements which emonstrate the significant avantages of heaphone amplifiers with very low (near 0-Ohm) output impeances. A low output impeance increases the amping factor of the amplifier-heaphone system. This paper will show that a high amping factor reuces istortion at the heaphone inputs, improves phase response, an flattens frequency response. Damping factor has long been recognize as an important parameter for auio power amplifiers, but has been ignore when builing an testing heaphone amplifiers. An amplifier has improve control over the physical movements of transucers when the amping factor is high. This control is essential for reucing the istortion cause by mechanical resonances in the transucers. A high amping factor also improves time-omain response by quickly accelerating an amping the mechanical motions of the transucers in response to musical transients. Finally, a high amping factor insures that transucer impeance vs. frequency variations have minimal effect on frequency an phase response. Damping factor is simply the ratio of the loa impeance to the source impeance. An amplifier with a high output impeance will have a poor amping factor. For example, a traitional heaphone amplifier with a 30-Ohm output impeance has a amping factor of only 2 when riving 60- Ohm heaphones. In contrast, enchmark's "0-Ohm" HPA2 heaphone amplifier (incorporate into many enchmark proucts) has an output impeance of approximately 0.01 Ohms. This means that the HPA2 has a amping factor of approximately 6,000 when riving 60-Ohm heaphones. The following tests show how a high amping factor improves auio performance.

2 x= Heaphone Output at 1 mw into Sony MDR-V6 Heaphones (106 SPL) THD+N vs Frequency (w/80 k LPF unweighte) y= k 2k 5k 10k 20k Sweep Trace Color Line Style Thick Data Axis Comment Cursor1 Cursor2 1 1 lue Soli 4 Anlr.THD+N Ratio Left 0-Ohm Drive Re Soli 4 Anlr.THD+N Ratio Left 30-Ohm Drive * * HPA THD+N vs Frequency 60 Ohm Heaphones at 0-Ohm vs 30-Ohm at 106.at27 Figure 1 emonstrates ramatic ifferences in istortion as the amping factor is change while riving heaphones. The plot shows THD+N vs. frequency. The blue trace shows minimal istortion at a amping factor of about 6,000. The re trace shows significant istortion at a amping factor of about 2. The two traces show a 50 ifference in low-frequency istortion. The re trace also shows evience of transucer resonance problems near 150, 300, an 1.2 k. The blue trace in Figure 1 was prouce using a enchmark HPA2 heaphone amplifier (0.01-Ohm output impeance) riving Sony MDR-V6 heaphones (60-Ohm input impeance). The re trace was prouce by inserting a 30-Ohm series resistor between the HPA2 an the heaphone input. When the resistor is inserte, the amping factor is reuce to 2. In both cases, the auio analyzer is monitoring the input to the heaphones. All of the istortion shown in the re trace is evelope across the 30-Ohm series resistor as a result of the mechanical motion of the heaphone transucers. Figure 2 shows a spectrum analysis of the istortion prouce at 100. The test setup is ientical to the setup for Figure 1. The blue trace was prouce at a amping factor of about 6,000. The re trace was prouce at a amping factor of about 2. The re trace shows a 64 increase in 2n harmonic istortion an a 38 increase in 3r harmonic istortion. r A x= Heaphone Output at 1 mw into Sony MDR-V6 Heaphones (106 SPL) THD+N vs Frequency (w/80 k LPF unweighte) y= k 2k 5k 10k 20k Sweep Trace Color Line Style Thick Data Axis Comment Cursor1 Cursor2 1 1 lue Soli 4 Fft.Ch.1 Ampl Left 0-Ohm Drive r A r A 2 1 Re Soli 4 Fft.Ch.1 Ampl Left 30-Ohm Drive * r * r HPA FFT 100 Tone 60 Ohm Heaphones at 0-Ohm vs 30-Ohm at 106.at27

3 Figures 1 an 2 clearly show that auible increases in istortion may occur when the amping factor is too low. Our own listening experiences suggeste that heaphones are cleaner when riven from a very low source impeance. We foun the ifference most noticeable when listening to solo piano recorings with the MDR-V6 heaphones. At a amping factor of 2, the harmonic istortion prouce by the heaphones seeme to beat against the inharmonic overtones of the piano giving the impression of a poorly tune instrument. This effect seeme to iminish when the amping factor was increase to It was this listening test that inspire enchmark's "0-Ohm" HPA2 heaphone amplifier in r A Heaphone Output at 6.65 mw into Sennheiser HD 650 Heaphones (106 SPL) THD+N vs Frequency (w/80 k LPF unweighte) k 2k 5k 10k 20k Sweep Trace Color Line Style Thick Data Axis Comment 1 1 lue Soli 4 Anlr.THD+N Ratio Left 0-Ohm Drive 2 1 Re Soli 4 Anlr.THD+N Ratio Left 30-Ohm Drive x= HPA THD+N vs Frequency 300 Ohm Heaphones at 0-Ohm vs 30-Ohm at 106.at27 Heaphone Output at 6.65 mw into Sennheiser HD 650 Heaphones (106 SPL) THD+N vs Frequency (w/80 k LPF unweighte) y= k 2k 5k 10k 20k Figure 3 is the same as Figure 1 except that the MDR-V6 heaphones have been replace by 300-Ohm Sennheiser HD650 heaphones. The blue trace is 30,000. The re trace is 10. The re trace shows slightly better high-frequency performance than in Figure 1. This improvement is ue to a 5 to 1 improvement in amping factor (10 instea of 2). Figure 4 is the same as Figure 2 except that the MDR-V6 heaphones have been replace by 300-Ohm Sennheiser HD650 heaphones. The blue trace is 30,000. The re trace is 10. This plot shows that the higher impeance of the HD650 heaphones i little to improve performance at 100. Sweep Trace Color Line Style Thick Data Axis Comment Cursor1 Cursor2 1 1 lue Soli 4 Fft.Ch.1 Ampl Left 0-Ohm Drive r A r A 2 1 Re Soli 4 Fft.Ch.1 Ampl Left 30-Ohm Drive * r * r HPA FFT 100 Tone 300 Ohm Heaphones at 0-Ohm vs 30-Ohm at 106.at27

4 Figure 5 shows the frequency response variations that occur when the amping factor is too low. The blue trace is 6,000. The re trace is 2. The re trace shows a 0.7 variation in frequency response ue to variations in the input impeance of the Sony MDR-V6 heaphones. A high amping factor is essential for achieving a flat frequency response. r A Heaphone Output Frequency Response 0-Ohm vs 30-Ohm Drive, Loa = Sony MDR-V6 Heaphones k 2k 5k 10k 20k Sweep Trace Color Line Style Thick Data Axis Comment 1 1 lue Soli 4 Anlr.Level A Left 0-Ohm Drive 2 1 Re Soli 4 Anlr.Level A Left 30-Ohm Drive HPA Frequency Response 0-Ohm vs 30-Ohm.at27 Note the peak in the response at 300. This correspons to the 300 resonance problem shown in Figure 1. This anomaly is not an issue when the heaphones are riven from 0.01 Ohms. e g Heaphone Output Phase Response 0-Ohm vs 30-Ohm Drive, Loa = Sony MDR-V6 Heaphones k 2k 5k 10k 20k Sweep Trace Color Line Style Thick Data Axis Comment 1 1 lue Soli 4 Anlr.Phase Left 2 1 Re Soli 4 Anlr.Phase Left Figure 6 shows the phase response variations that occur when the amping factor is too low. The blue trace is 6,000. The amping factor for the re trace is 2. The re trace shows variations in phase response ue to variations in the input impeance of the Sony MDR- V6 heaphones. The blue trace shows that phase response is well controlle when the amping factor is 6,000. A high amping factor is essential for achieving a flat phase response. HPA Phase Response 0-Ohm vs 30-Ohm.at27

5 Please note that all of the istortion, phase, an frequency problems shown in Figures 1 through 6 (re traces) are ue to the mechanical transucers in the heaphones working against a 30-Ohm output impeance. If we replace the 60-Ohm heaphones with a 60-Ohm loa resistor, all of the errors isappear. Heaphone amplifiers are traitionally teste with a resistor loa, or no loa at all. Ieal resistor loas hie most of the problems that occur when traitional 30-Ohm heaphone amplifiers rive heaphones. For this reason, heaphone specifications are often very misleaing. All measurements shown in this paper are taken at the heaphone inputs. These are not measurements of the heaphone acoustic output. Acoustic measurements may show significantly higher levels of istortion. ut, the acoustic measurements can never show less istortion than the measurements taken at the heaphone inputs. The 0.01-Ohm rive provie by the HPA2 shoul ampen some of the istortion create by the transucers, an acoustic measurements may show this. Our listening tests suggest that this may be the case. These are areas of ongoing stuy in our research lab. At a 0.01-Ohm input impeance, both sets of heaphones receive a clean electrical signal (as show by the blue traces). Nevertheless, in our listening tests, the HD-650 heaphones always seeme to be cleaner than the MDR-V6 heaphones. This implies that only some of the transucer istortion can be remove with a low-impeance heaphone amplifier. Again, this is an area of ongoing stuy. The frequency response an phase response variations shown is this paper shoul be reproucible with acoustic measurements. For example, the 0.7 change in the frequency response of the MDR-V6 heaphones (shown in Figure 5) shoul be reproucible with acoustic measurements. In summary, we have shown that a high amping factor improves auio reprouction through mechanical transucers. The formula at the beginning of this paper shows that amping factor is the ratio of the loa impeance to the source impeance. Heaphones typically have 60-Ohm to 300-Ohm input impeances. We use two popular heaphones that are at opposite ens of this range. oth heaphones showe performance improvements when the source impeance was reuce from 30- Ohms to 0.01-Ohms. The enchmark "0-Ohm" HPA2 heaphone amplifier use in these tests has an output impeance of 0.01 Ohms at 1 k. The HPA2 achieves a amping factor of 6,000 to 30,000 at 1 k when riving typical (60-Ohm to 300-Ohm) heaphones. The 0.01-Ohm rive impeance of the HPA2 prouce ramatic performance improvements when compare to a 30-Ohm rive impeance. Distortion at the input of the heaphones was reuce by as much as 64, 10 egree phase response variations were remove, an frequency response was improve by as much as 0.7. When using a 30-Ohm heaphone amplifier, the amping factor can be increase slightly by selecting 300-Ohm heaphones instea of 60-Ohm heaphones. This change increases the amping factor from 2 to 10. Figures 3 an 4 showe that a amping factor of 10 was insufficient. Significant performance

6 improvements were obtaine by raising the amping factor to 6,000. No aitional improvement was obtaine by increasing from 6,000 to 30,000. A low output impeance places significant emans on the heaphone rive circuits. The HPA2 is well suite to this task. We have shown that the HPA2 remaine nearly istortion free while riving heaphones (see the blue traces in each of the figures above). In a previous white paper we examine the performance of two other "0-Ohm" heaphone amplifiers an ocumente the ifferences. The HPA2 showe significant avantages over the other two units. Test Conitions to Prouce a Damping Factor of 2: 1) Sony MDR-V6 60-Ohm Heaphones 2) 30-Ohm metal-film series resistor (to egrae amping factor) 3) enchmark HPA Ohm heaphone amplifier (in DAC1 HDR) 4) Normal listening level (106 peak SPL) Test Conitions to Prouce a Damping Factor of 10: 1) Sennheiser HD Ohm Heaphones 2) 30-Ohm metal-film series resistor (to egrae amping factor) 3) enchmark HPA Ohm heaphone amplifier (in DAC1 HDR) 4) Normal listening level (106 peak SPL) Test Conitions to Prouce a Damping Factor of 6,000: 1) Sony MDR-V6 60-Ohm Heaphones 2) No series resistor 3) enchmark HPA Ohm heaphone amplifier (in DAC1 HDR) 4) Normal listening level (106 peak SPL) Test Conitions to Prouce a Damping Factor of 30,000: 1) Sennheiser HD Ohm Heaphones 2) No series resistor 3) enchmark HPA Ohm heaphone amplifier (in DAC1 HDR) 4) Normal listening level (106 peak SPL) Test Equipment: Auio Precision System 2722 Copyright 2011 enchmark Meia Systems, Inc.