Why does this matter to my ear.
I hear great highs from NC500's class-D as good as good linear amps when driving say a nice even benign flat 4 or 6ohm ribbon tweeter with low phase angles.
But when the same amp is driving a speaker with not such a benign load I change my mind from nice to so so about the sound, and linear amps then sound way better.
Cheers George
I hear great highs from NC500's class-D as good as good linear amps when driving say a nice even benign flat 4 or 6ohm ribbon tweeter with low phase angles.
But when the same amp is driving a speaker with not such a benign load I change my mind from nice to so so about the sound, and linear amps then sound way better.
Cheers George
I think Class D output LC filter is tuned for a particular impedance (4ohms, 8ohms etc). If that changes too much, the filter is not optimal and perhaps there is more HF leakage or higher distortion. What does your speaker impedance curve look like - that is, what is “not benign” in your opinion?
This (not so good sound from the Hypex NC500's Class-D's) was heard with my ML Monolith III, my friends B&W 802 diamonds, and another friends normally beautiful sounding refurbished Quad ESL63's, yet these NC500's sounded great with the Raven tweeter'ed two way.
Cheers George
I see what you mean. The 802D has a very non flattering impedance curve. Downright nasty. That peak near 2.5kHz in particular will wreak havoc. This is from Stereophile’s article.
This is why, to me linear amps can do it all, but Class-D's have problems with many speakers that need to be addressed.
And why I like the idea that Technics had if I'm on the right track, with the (GaN) SE-R1 of raising the switching frequency three times higher to 1.5mhz and accordingly the output switching noise filter as well three times higher, then the speakers that cause trouble will be less likely to do so.
Cheers George
It is not so clear that having GaN is the panacea. From what I have seen and heard, the technology is not quite there yet. The switching is so fast, that it generates a lot of higher order harmonics, while technically at a low level, taken as a whole, can sound fatiguing, brittle, and not pleasant after 15 minutes of listening.
To get 3 x higher switching frequency (1.5mhz) than what it is at present (600khz), and therefore have a low order low pass output filter that is also 3 x higher.
As not to cause this sort of phase shift down into the audible hf and mf band (red trace). https://ibb.co/ZKgkyb6
From what I understand you do need GaN technology to get it. That why Technics did it with the SE-R1 and from all reports from those privileged few that have heard it, the highs are sweet/delicate and extended, and maybe also the answer to driving without distortions, HF that has large variations in impedance and phase angle.
Cheers George
Last edited: