MRC01
Major Contributor
That makes sense. The pot itself adds noise, but not distortion. The distortion must be caused by something else, perhaps PCB routing.
Correct.Am I reading the channel balance graph properly: the channels match within 0.5 dB from full volume to about -40 dB? Then from -40 and lower it gets all wonky.
Correct. I have recently separated the topic headings with "DAC Measurements" and "headphone amp measurements." The former is always with pre-amp out, not through the headphone jack.The HA-1's volume knob is a conventional analog pot and its fixed output setting bypasses the volume control. Do I understand how Amir tested -- both of the first 2 measurements were done using the line stage XLR outputs (not the headphone output)?
resistance = noise
There is no direct current measurements since all probing is done in parallel (voltage) mode. Watts is simply computed from giving the analyzer the resistance of the load. As noted, you can compute the current manually from the graph.@amirm Are you in general able to measure not only the watts and volts but also the current the headphone amps deliver on the common 30 and 300 ohm marks???
It didn't. Was LCD-X. Edited the review.
That looks like the formula for thermal noise. This excludes current noise, which makes it akin to the theoretical noise of an ideal resistor. Metal film or wire wound might get close to this. It's the same formula I used several years ago when building a stepped attenuator from metal film resistors. But IIRC, other kinds of resistors like potentiometers are noisier than this.Ok, made me look. Back-of-the-napkin estimate ...
That looks like the formula for thermal noise.
... the channels match within 0.5 dB from full volume to about -40 dB? Then from -40 and lower it gets all wonky.
That's actually pretty decent for a conventional potentiometer volume knob; I've seen much worse. But in my view, the volume knob is one of the weak points of the HA-1. At that Kilobuck and up price range there's no engineering excuse not to have a fine grained stepped attenuator or a 24-bit dithered digital control.Correct.
So the HA-1's line stage, which is opamp based, is cleaner than its headphone amp, which is discrete. I'm not surprised. I can see why they made the headphone amp discrete because if you want that kind of power (7 watts ) is it even possible with a typical opamp based headphone amp (something like an OPA209 or AD797 stiffened with BUF634)? It's just a shame they route the line output through the headphone amp when the volume control is being used.Correct. ... if you set the unit to variable, it routes the signal from the DAC completely through the headphone amp ...
Depends on your definition of typical opamps. opa189+tpa6120a2 composite will give 700ma current, if you parallel it you double that. 7 watts isn't a dream. For 8 ohm load it's voltage limited. For 4 ohm load it's fine. Ada4870 can work under +-20v so in that case it's even better.That's actually pretty decent for a conventional potentiometer volume knob; I've seen much worse. But in my view, the volume knob is one of the weak points of the HA-1. At that Kilobuck and up price range there's no engineering excuse not to have a fine grained stepped attenuator or a 24-bit dithered digital control.
So the HA-1's line stage, which is opamp based, is cleaner than its headphone amp, which is discrete. I'm not surprised. I can see why they made the headphone amp discrete because if you want that kind of power (7 watts ) is it even possible with a typical opamp based headphone amp (something like an OPA209 or AD797 stiffened with BUF634)? It's just a shame they route the line output through the headphone amp when the volume control is being used.
I need to still make them.....Amirm, do you have balanced cables yet?
I need to still make them.....
Whilst at this price range no doubt the single ended should be terrific as well still the balanced is more than likely better in leaps and bounds.