The question is how is this beefy low impedance PSU made. IC regulators, discrete or opamp regulators? What kind of discrete circuit? Which transistors, voltage references or opamps?
And also... it's really weird to me that the "ground channel" opamp should influence the sound (and the stability) of the amp...
I really prefer to "only" have the opamps that amplify the stereo channels to take care of...that's enough, I say... call me a simpleton
Finally, in my limited experience with these 3 channel headphone amps, I didn't entirely like the sound they produced. The 3 channels somehow seemed to make things thinner, and less natural...both using BUF634's (though this was preferable) and with the AD8397. I remember I experimented a lot in order to find something in G that went well with the AD8397 or AD45048 in RL, and I was never quite happy with the results.
But...I don't pretend to say the final word on the whole active ground matter... not at all. I just have my sentiments
If you use BUF634 as it is, not in the loop of an opamp, the high output impedance gives you an amp with lot's of crosstalk and artificially wide center information (like vocals). That sound is erroneues. I've tried open loop buffers without success as ground channel amps several times.And also... it's really weird to me that the "ground channel" opamp should influence the sound (and the stability) of the amp...
I really prefer to "only" have the opamps that amplify the stereo channels to take care of...that's enough, I say... call me a simpleton
Finally, in my limited experience with these 3 channel headphone amps, I didn't entirely like the sound they produced. The 3 channels somehow seemed to make things thinner, and less natural...both using BUF634's (though this was preferable) and with the AD8397. I remember I experimented a lot in order to find something in G that went well with the AD8397 or AD45048 in RL, and I was never quite happy with the results.
But...I don't pretend to say the final word on the whole active ground matter... not at all. I just have my sentiments
AD8397 sounds like it does. I think neither you nor I could be happy with such an amp. Besides it's not meant to be used at unity gain. It's not that the ground channel amp has to be run at unity gain, but it usually is.
The ground channel has the same influence as the L/R channels. I think all opamp rollers knows this.
I've just done a test with one of my three channel amps. I rebuilt it so that the ground channel was used as a railsplitter, handling all grounds, and then rebuilt it to active ground again. It sounds better when the grounds are separated and when measured the crosstalk is better and the distortion is lower.
I've compared LM317/337 with ground channel and I prefer the latter. I've compared resistors/capacitors à la CMOY to ground channel and the latter is without question better to my ears. Active ground sounds different. At first it can be experienced as thin and voices doesn't sound as wide, but to my ears this is how it should be.
No, that was in a commercial amp, and the BUF634 was in the AD8610 feedback loop.
The opamp that was working in unity gain, was not the AD8397. I tried various...
And yes, that's the problem! It had an important influence on sound it shouldn't have had! It makes me think of some odd form of distortion...? Or simply of an unnecessary complication...
To me, the presence of a further chip opamp/buffer, which is in the signal path too, can only double the "thinning" effect typical of chips being used to drive transducers.
Maybe a 3 channel opamp with hefty discrete buffers... I haven't heard that.
Wait...yes I have! I had the iBasso P1, that had complementary transistor buffers in L, R, and G. That one, thinking of it, sounded really good and wasn't thin or ethereal sounding.
Last edited:
The AD8397 sounds better without the (opamp) active ground, to me, since it tends to be thin sounding per se...
Anyhow... back to our duty... The King Cobra's amazing! The more it burns in, the better it feels. From the start, there was a substantial increase in sonic composure & coherence, and in microdynamics, compared to the already very respectable Van Den Hul The Bay C5.
Anyhow... back to our duty... The King Cobra's amazing! The more it burns in, the better it feels. From the start, there was a substantial increase in sonic composure & coherence, and in microdynamics, compared to the already very respectable Van Den Hul The Bay C5.
You have a way with words Andrea. To bug you a bit I could say that it's a matter of projection.
Your hearing is really different from mine. If you find AD8397 "thin" I have no chance to follow your descriptions on how things sound. I find it's one of the faults it doesn't have.
Back to ground channels. Something has to deal with the return currents, and whatever this is it will influence the sound whether it's transistors, monolithic regulators, capacitors, transformer or opamps. With a ground channel you have an easy access to tweak the sound. I only find this welcome.
Your hearing is really different from mine. If you find AD8397 "thin" I have no chance to follow your descriptions on how things sound. I find it's one of the faults it doesn't have.
Back to ground channels. Something has to deal with the return currents, and whatever this is it will influence the sound whether it's transistors, monolithic regulators, capacitors, transformer or opamps. With a ground channel you have an easy access to tweak the sound. I only find this welcome.
When used for headphone driving. It was thin (in vocals) yet didn't lack midbass...strange.
It might be the opposite to the LT1028 in this regard!?
Last edited:
True...but an active ground channel made with opamps influences the sound unusually...
Depends. When your LR opamp is compromised in the first place, OK. But not when the opamp is good enough not to need these funny alchemies
The Audioquest King Cobra has really transformed the sound of my system for the better... more than just an opamp change for sure!
There's such a rich tonality now... such an impactful and taut bass... midrange body and refinement, treble accuracy without any emphasis... this cable has real class.
"Vivadixiesubmarinetransissionplot" by Sparklehorse (their wonderful 1995 debut album) is so much more captivating now..
There's such a rich tonality now... such an impactful and taut bass... midrange body and refinement, treble accuracy without any emphasis... this cable has real class.
"Vivadixiesubmarinetransissionplot" by Sparklehorse (their wonderful 1995 debut album) is so much more captivating now..
Last edited:
I would relace those after the DAC with the LT1028ACN8 (which is DIP and gain of -1 stable). It's really very transparent and controlled, and it's very good at tonality...
Then you can use the LME49710HA (TO99) as the ubalanced to single ended converter, and buffer. Or also the LT1468CN8 (DIP) as suggested, or (I think I prefer this) the LT1357CN8 (also DIP). Better not use the LT1028 in that position because source impedance is high and I'm not sure it would be stable..
Hi AndreaThe Audioquest King Cobra has really transformed the sound of my system for the better... more than just an opamp change for sure!
There's such a rich tonality now... such an impactful and taut bass... midrange body and refinement, treble accuracy without any emphasis... this cable has real class.
"Vivadixiesubmarinetransissionplot" by Sparklehorse (their wonderful 1995 debut album) is so much more captivating now..
Now that you are using a new cable, you probably need to retest all those opamps
Thanks! Actually, in my naivete I had already placed OPA637s after the DAC (and tweaked the gain so they'll stay stable), with OPA627s in the summing stage... but I'm not worried about those since I don't use the SE outputs anymore. This was done a long, long time ago - but I've been wondering if the OPA637s were actually better than the old NE5534s they replaced, in this application. In any case, the gain is too high for my liking and the tonality isn't very good either, so they'll have to go.
So now, I'll have to choose between the LT1468 and LT1028...
So now, I'll have to choose between the LT1468 and LT1028...
Interesting... having replaced 5534's myself in a DAC (Micromega Stage 2) I know that OPA604 and AD845 work well for me. I use AD845 for the final buffer. Noise isn't an issue here, and at those impedances the NE5534 is probably about as good as it gets.
There is some evidence to suggest that FET input opamps work better for I/V conversion, possibly because Bipolar input stages "interact" with any HF hash coming from the DAC in a non linear way... hence the OPA604.
There is some evidence to suggest that FET input opamps work better for I/V conversion, possibly because Bipolar input stages "interact" with any HF hash coming from the DAC in a non linear way... hence the OPA604.
Hi Andrea
Now that you are using a new cable, you probably need to retest all those opamps
Hahaha, not quite, because the change in sound, while significant, would be the same for all opamps.
Besides, I surely used good cables before, not just the VDH, but also the Sommercable classique, and others.
If that were true, I'd have to retest all those opamps also due to the digital cable change to the Supra TRICO-RCA, which was as significant as that of interconnects.
But no, that's not quite the case. Except that maybe the LT1028 would fare slightly better with the Audioquest, compared to the OPA1611.