Table of Contents
Advertisement
The whole signal path from DAC to the headphone speaker is balanced (!) Signal inversion
happens fully transparent and lossless within the digital domain
Signal inversion is completely transparent and lossless
The entire signal path from the DAC to the headphones is balanced
There is no switching of signals (often with cable harnesses) at the output jacks
In True Balanced mode the maximum output level rises to +7 dBu for IEM, +13 dBu for Low
Power and +25 dBu for Hi-Power. The signal to noise ratio of IEM and L-Power modes rises by
about 1 dB.
While the higher output voltage might be useful for some older exotic headphones, a four times
higher output power (around 8 Watts, per channel) does not make sense. Fortunately the current
limit circuit in the ADI-2/4 Pro SE takes care of that and prevents the output power to rise higher
than 4 Watts, and below 2 Watts at impedances below 20 Ohms. See chapter 34.18 for details
about the available output power at the phones outputs.
34.7 Crosstalk at the TRS headphone outputs
In the Tech Specs, the crosstalk (also called channel separation or crosstalk) at the TRS head-
phone output is given as values between 60 and 80 dB, while at the line outputs often more than
110 dB are reached. How is this possible? An explanation can be found in an article by Solder-
dude, Resistance, Impedance and other issues.pdf:
https://diyaudioheaven.wordpress.com/tutorials/power-impedance-etc/
Starting on page 11, Solderdude describes and calculates crosstalk as a problem of the common
ground line in headphones. What is missing is the contact resistance of the common ground as
soldered to the TRS, across the TRS itself to the jack socket (contact resistance), and this then
to the PCB (printed circuit board).
But that's a maximum of 0.00x ohms, right? Correct, but there is the problem with the usual log-
arithmic approach. This is a simple voltage divider, which logarithmically simply does not advance.
For a headphone with 32 Ohms, a ground transition with x Ohms causes y crosstalk:
x Ohm
y dB
0.1
-50 dB
0.01
-70 dB
0.001
-90 dB
0.0001
-110 dB
0.0001 Ohm is not possible in the real world. 0.001 Ohm is not possible either, because the design
of the jack plug simply does not allow it. In practice, one measures significant differences in dB
depending on the jack plug used, which is due to both its respective construction and the used
material. 0.01 Ohm or just below is therefore a much more realistic value.
Using a 4-pin XLR or Pentaconn will give better values, even if they are used unbalanced. Two
separate contacts result in halved contact resistance for ground. Add an ideal ground connection
after the socket and one receives more stable, slightly higher values.
Even if the values at the TRS are disappointing at first (measured 80 dB instead of expected 120),
one better not measures headphones with 3 meters of common ground wire (0.5 Ohm cable
resistance = 36 dB crosstalk attenuation). But such extension cables are rarely to be recom-
mended for other reasons as well.
User's Guide ADI-2/4 Pro SE – v 1.0
81
Advertisement
Table of Contents
