Review and Measurements of Benchmark HPA4 Headphone Amp/Pre

[RichB]

If you select the DAC ON INPUT1 option, you can send the remote code for any digital input and this will set the HPA4/LA4 to input 1.

DAC on Input 1 could work for the balanced inputs if the Analog button directly selected Input 2.
Since it does not, there is no good way to select Input 2.

Once DAC on Input 1 is Yes, all other input buttons work and also select Input 1.

Alternately, if there were a DAC on Input 1 AND the Analog button selected Input 2, that would work.

However, it makes more sense to have the D1 - D4 buttons select inputs on the LA4/HPA4 when DAC on Input 1 is NO.

- Rich

 

[Dealux]

One HPA4 feature that is often overlooked it the fact that it uses a 256-step fully-balanced relay-controlled attenuator to adjust the output volume. The result is that the measured performance is available at a wide range of volume settings, not just where the volume control is maxed out (as it always is when these headphone amplifiers are being measured). It would be tedious and time consuming to perform measurements at various volume settings, but this is where the HPA4 would really shine.

Isn't that just this measurement?

 

[Tom C]

Isn't that just this measurement?

I could be wrong, of course, but I believe that graph is generated by varying the input voltage, not by adjusting the volume control on the device under test. To get this, you’d have to set the volume control at max, and leave it undisturbed throughout the test.

 

[Dealux]

That makes sense because you couldn't control the output so precisely with the volume knob.

 

[RichB]

That makes sense because you couldn't control the output so precisely with the volume knob.

This means we do not really test the performance of the volume controls because it is difficult.

- Rich

 

[MC_RME]

It is not difficult, but time consuming. And producing tons of data that need interpretation. If you then get into the details (there are several advantages of a balanced relay attenuator compared to the mostly used unbalanced ones, some not obvious at all) then you could make a whole website about this topic...or countless YouTube videos...

 

[Rja4000]

If you then get into the details (there are several advantages of a balanced relay attenuator compared to the mostly used unbalanced ones, some not obvious at all) then you could make a whole website about this topic...or countless YouTube videos...

Which would be very interesting to me.
But probably not to thecmajority of reader here.

 

[John_Siau]

Which would be very interesting to me.
But probably not to thecmajority of reader here.

The balanced attenuator/gain stage allows a variable gain differential amplifier which optimizes the SNR over a wide range of gain and attenuation settings.

All of the stepped attenuator topologies work well at unity gain, but many start to add excessive noise and/or distortion when the gain moves above or below unity gain.

The impedance of the resistors will determine the amount of Johnson noise added by the resistor network. We keep the impedances very low to minimize the Johnson noise and then we use a fully-balanced topology to gain an additional 3 dB noise advantage. The cost of the balanced topology is that we have to use twice as many resistors and relays. The balanced topology also allows us the ability to control the left and right gain independently (implementing a balance control). It also means the channels can be completely isolated because they are not sharing any relays.

 

[RichB]

It is not difficult, but time consuming. And producing tons of data that need interpretation. If you then get into the details (there are several advantages of a balanced relay attenuator compared to the mostly used unbalanced ones, some not obvious at all) then you could make a whole website about this topic...or countless YouTube videos...

When an Amplifier/Preamp/AVR/ASVP is tested only at unity gain, there volume control measurement may not be indicative of performance in use.

- Rich

 

[MC_RME]

To add to @John_Siau 's post: IMHO the biggest advantage of balanced relay attenuators is that the attenuation operates ground-free. Due to the low impedance/resistance and ground current flowing at input and output, on the same reference as the attenuator, unbalanced relay attenuators show weird side effects when doing their job (attenuating), from simple level deviations to crosstalk and others. That behaviour changes then with the quality of the connection cables at input/output, the source impedance etc..

From my experience if RME ever makes a unit that has such an attenuator, it must be a balanced version, or it's not worth it. For a similar reason the RefLevel switching in our units is done in symmetrically built (balanced) circuits.