Deal with this off forum. Posting private correspondence without the writer's permission is a violation of forum policy. Read and familiarize yourself with our rules.I understand .
Huh? THD+N is higher at lower powers because of the noise floor, but THD is roughly lowest (of the 3 power levels shown) at 1 Watt. Why THD drops as freq rises above several kHz is at once wonderful and freqy
. Actually it makes me wonder if it's an artifact of the measurement gear. These amps are linear...too linear!
Last edited:
At 1W at all impedances the analyser sees only noise. The plot is called THD+N for a reason. Measuring THD when noise is greater than the distortion is difficult. So, as soon as distortion has dropped into the noise floor, the THD+N graph actually shows noise-to-signal ratio. I hope the attached pdf helps clarify matters.
The noise voltage is constant, but the voltage corresponding to 1W depends on the impedance. So, noise-to-signal is lowest at 8 ohms.
@Erland, the fact that you have not sold your Ncore modules shows your ears and your mouth disagree. Your behaviour shows that you have a personal gripe against Hypex, godnoze for what infraction, and you have decided to take that out on a public forum. Unfortunately for you, you have posted photographs of your power supply and your listening setup (and your wonderful midrange which proves your speakers must be world's best) so everyone knows exactly how serious to take you. Your continued presence not so much shows your fighting chops as your inability to hear the smirking.
The noise voltage is constant, but the voltage corresponding to 1W depends on the impedance. So, noise-to-signal is lowest at 8 ohms.
@Erland, the fact that you have not sold your Ncore modules shows your ears and your mouth disagree. Your behaviour shows that you have a personal gripe against Hypex, godnoze for what infraction, and you have decided to take that out on a public forum. Unfortunately for you, you have posted photographs of your power supply and your listening setup (and your wonderful midrange which proves your speakers must be world's best) so everyone knows exactly how serious to take you. Your continued presence not so much shows your fighting chops as your inability to hear the smirking.
Last edited:
I could see all that from the original THD + N vs. power plots (although your revised plot makes it quite clear without tilting your head on the side)
What I don't understand is why in the THD vs freq plot, the 10W distortion is shown as higher than the 1W distortion. This doesn't match up with any of the THD+N vs. power plots
If the right hand plot is @4R, it would appear to be both the 10W and 100W curves that have been placed where the 1W should be and vice versa.
According to the left hand plot, 10 and 100W should both be @ 0.00045% where the 1W seems to sneak just above 0,001% -if all is @4R of course
200W makes a lot of sense, but the blue (10 or 20 W) and green (1 or 2 W) curves at the right hand plot doesn´t seem to add up to what the left plot tells.
BUT if as what ChirsPa says, that the blue and green have been mixed up, then everything seem to fall into place.
2W gives just beneath 0,001% THD+N (the blue) and 20W gives around 0,00045% THD + N (green)
That said, its all about how it sounds (to me at least)
BUT if as what ChirsPa says, that the blue and green have been mixed up, then everything seem to fall into place.
2W gives just beneath 0,001% THD+N (the blue) and 20W gives around 0,00045% THD + N (green)
That said, its all about how it sounds (to me at least)
NC400 vs NC1200 data sheet
Bruno, the data sheet for the NC400 seems to show much better distortion figures then the data on the NC1200. I realize that the figures are low as they are and I realize that the NC1200 has more watts but I would have thought seeing how the NC1200 is the top line OEM module its figures would be better (if not equal to) then the NC400.
Bruno, the data sheet for the NC400 seems to show much better distortion figures then the data on the NC1200. I realize that the figures are low as they are and I realize that the NC1200 has more watts but I would have thought seeing how the NC1200 is the top line OEM module its figures would be better (if not equal to) then the NC400.
The NC1200 has a bigger and hence slower power stage. That determines how good distortion can get. As it is, I think the distortion of the NC1200 is low enough for any use it can be put to. The interpretation that the NC1200 should be higher up the ladder compared to the NC400 is yours. The NC400 has fancy discrete audio circuitry which the NC1200 doesn't have. They're different modules for different markets. I certainly don't see one above or below the other.
That said, its all about how it sounds (to me at least)
I agree! I'm still impressed when I listen. Just fantastic IMO.
Oh, OK. I was applying the noise-as-floor concept to the left graph, but treated the right graph as an absolute level instead of a percentage... in other words, I saw what I wanted to see. I had looked sideways at the graph until punting: the shape of the bottom, green "1 Watt" trace had that nice flatness to it so I ignored percentage vs. level. Those low swoops on the graphs make me dizzy
. Thanks for explaining the problem and inciting a correction.When Bruno refers to Erlend's amp build, it helps to see another linear supply (UcD, not ncore) amp which is equally roomy but neatly laid out here:
http://www.virtalahde.com/0306_vahv1.JPG
Virtalähde Mastering
It looks like more wires should be twisted, and everything should be closer together, but Jakko is pretty careful. He has a great combination of electronic design and construction skills plus an ME's ear.
Last edited:
Great looking power supply, but I would hate to see a compact, efficient amp such as nCore saddled with a huge, bulky, inefficient unregulated PS...
c'mon... so you mean you would use a switching amp with a switching supply? that doesn't make any sense.
- Status
- Not open for further replies.