Benchmark ahb2 power amplifier

[barrows]

21 hours ago, sandyk said:

Kal

 Do you know how they manage to increase S/N to such a degree given limitations due to Johnson noise in resistors etc ?

 Signal/noise (A-weighted): 132dB (stereo) is much higher than most amplifiers are capable of .

Was this measured at the low gain setting ?

Regards

Alex

As i recall this is a low gain amp (I need to go back and look at the Stereophile measures), low gain is a sure way to reduce noise.  I suspect the rest of the engineering is good as well.  Of course nothing comes for free, with this low gain one will have to have higher gain in the source to reach the same levels in a given system context, higher gain in the source means higher noise there, so no free lunch.

This points out how there are benefits to be had by careful gain matching in a system.  With the right gain structure, one will have the lowest noise floor the components are capable of at the speakers.  

[barrows]

@Kal Rubinson, right, so the amp suits your system well.  Also, when bridged, and at the highest gain setting, the AHB achieves a more typical gain.

[barrows]

@Miska, yeah agreed.  We have this problem because there are no real standards for home audio.  Couple that with the high variability of average levels on recordings (loudness wars...), and gain matching and needs get complicated.  In many home systems people are throwing away performance by having awkward gain structures, requiring lots of attenuation.

i DIY my DACs, and so can adjust output capability to suit the system at hand, but even then I find I have to have a fair bit of 'extra" gain to accommodate many audiophile recordings with low average levels and high dynamic range (Reference Recordings, Channel Classics, etc)-then I have t apply quite a bit more attenuation for contemporary pop, like Lorde or something.

[barrows]

7 hours ago, Miska said:

 

Sure there are some differences and adjustment needs. But with suitable gain structure you can get very close to 0 dB attenuation with those highest quality recordings and get the best performance too. And still be safe with the louder lesser quality ones to rely on just small amount of digital adjustment for example from the player application.

 

Now I bet many audiophiles rarely if ever go to less than -20 dB attenuation. Same with my pre + power amps, quiet listening can be -50 dB volume and the loudest -20 dB. With loudspeakers that have way under 90 dB sensitivity.

 

Of the DACs best situation is with Holo Spring, since it puts notably lower output level with DSD inputs.

 

With headphone amps situation is much better, many HP-amps have switchable gain, so I can get better gain matching with headphones which is anyway what I mostly use for listening.

 

Yes, totally agreed.  In my case I set the gain of the DAC such that indeed I only use a few dB of attenuation with the lowest average level recordings.  But I apply a -5 dB reduction in software before my oversampling step (to DSD 256) to avoid clipping in the oversampling.  I do play music quite loud here as well (rural home, hurray!).  I end up with my DACs set for about 4 V output at 0 dB (balanced), and this works well for my my amp and speakers (my Amp reaches 200/400 watt output at 2 volt input).  I still end up applying quite a bit of attenuation for contemporary pop like Lorde.

Of course there are a lot of DACs with much higher output levels, 6-8 V is fairly common for balanced outputs.  And most folks not applying a reduction somewhere (like I do for oversampling) probably have too much gain and are using more attenuation than would be ideal.  A reasonable solution for computer audio folks would be too apply some reduction in software.  At 64 bits no one should be too concerned about applying, say, a 6 dB reduction in ROON or HQPlayer to reduce overall gain and allow their final volume control (even analog) to be operating more ideally. 

[barrows]

2 hours ago, Kal Rubinson said:

The value of the lower gain settings is not that it reduces the noise contribution of the amp but it, effectively, attenuates the noise contribution from its sources.

Kal, a more clear way of stating the above might be to say: the lower gain setting in my system reduces the amplification of the source noise (as it does not really attenuate it) such that it is no longer audible.  Or something to that effect.

 

Suffice it to say, proper gain matching in a system to where the user does not have to apply high levels of attenuation (using the volume control) can result in improved system performance, and it is likely that a lot of audiophiles are leaving some performance on the table in this regard.  If one's volume control never goes above -20 dB, then one could benefit from changing the gain structure/relationships between their components.

 

As to noise of the AHB amplifier itself, I am sure it is a very low noise design, but I would be careful directly comparing JA's measurements with those of other amplifiers with different gains, as this would be an unfair comparison.  I re-read the measurement section of the review and still could not find exactly what gain settings were used for which measurements.  But, I am not criticizing the AHB, from a measurements standpoint it is pretty clear that it performs very well indeed.  I would like to see a higher power amp from them though, as the increase in output impedance when bridged would bother me a little.

[barrows]

9 hours ago, Kal Rubinson said:

I was referring to a temporary situation but it reveals my personal preference for a physical knob control.  AFAIK, the only multichannel DACs that have a physical volume knob are the Merging (NADAC and Hapi) and the upcoming OKTO.  Do you know of any others?

Bricasti.