ExaU2I uses an FPGA to buffer I2S stream and read out with high quality clocks. But then the isolator is plonked right on the output I2S interface.
Any guesses on why that choice was made?
That was one of the better interfaces I've experimented with. However still nowhere near as good as the Phison PD2 implementation.

Really looking forward to measurements showing those differences. I assume that by the same logic, you will say that a WAV and FLAC file sounds different.
Yes they do depending on the decoding algorithm. The Resonessence Mirus does the FLAC decoding in the FPGA when using the SD card transport. It was the most difficult thing they had to do to make it sound good. However WAV still sounds better when comparing the 2 versions of the same track. Enough that I clearly know which version is playing without even seeing the format it's in. Anything that causes extra decoding degrades sound at some level. The simpler the better.
I don't really care. I learn by hands on experience. Not blinding parroting what others say.
My test methods are perfectly fine for my needs. They result in sales every time, as well as years of satisfaction for my clients. As I said the only reason to post measurements is for marketing purposes for online sales.
So, you don't really care if it is truth or a lie. This method is perfectly fine for your needs, as it result in sales... Sales is the only reason you post here, truth is not on your list.
So, you don't really care if it is truth or a lie. This method is perfectly fine for your needs, as it result in sales... Sales is the only reason you post here, truth is not on your list.
If 100% of my clients says it sounds better, it is the truth. Even if measurement data was to show degraded sound, the truth would still be that 100% of my clients prefer it.
Same with the SIL-994 in the NC-500 buffer. Even if the measurement results from
The LM4562 showed better performance, the truth is 100% so far who tried both in the rev C board, prefer the SIL-994.
I knew long before anyone tried the SIL-994 in a Hypex NC-500 buffer that they would all prefer it. And for some reason they do.
If 100% of my clients says it sounds better, it is the truth.
Even if 100% of a population who are not your clients say otherwise?
Even if 100% of a population who are not your clients say otherwise?
100% of the population who aren't my clients are who I'm talking about with the NC-500 buffer. I don't sell a NC-500 based amp.
If I owned a restaurant, and 90+ % of people told me they preferred green beans over Brussel sprouts, I would listen to them. I could show them data all day long that the Brussel sprouts were more nutritious, but they simply wouldn't care.
Last edited:
^ You, as usual, are avoiding to answer a simple but 'inconvenient' for you question and trying to change topic... The good thing is the answer is obvious to anyone familiar with the logics.
That was one of the better interfaces I've experimented with. However still nowhere near as good as the Phison PD2 implementation.
Are you saying that you can pick out difference between tracks playing on these? Other parts of the chain being high quality and transparent?
What kind of differences do you hear?
^ You, as usual, are avoiding to answer a simple but 'inconvenient' for you question and trying to change topic... The good thing is the answer is obvious to anyone familiar with the logics.
But 100% don't say it. If they did then I would definitely change my strategy. I prefer successful methods of doing things.
That was one of the better interfaces I've experimented with. However still nowhere near as good as the Phison PD2 implementation.
Are you saying that you can pick out difference between tracks playing on these? Other parts of the chain being high quality and transparent?
What kind of differences do you hear?
I'm saying with the Exasound, I could still hear a difference between different computers connected. Even George has now admitted this after boasting how effective his galvanic isolation was for years. Notice they are now selling a streamer called the Playpoint?
The Phison PD2 is the most immune to hardware change I have heard. I have 12 test computers in my arsenal now. I have tried everything. Every version of Windows. Even a stripped down version of Windows embedded 8. I have 3 Mac computers. A MacBook Pro with I7, MacBook Air with I5, and Macmini with I5. All 3 have Windows running on them with Bootcamp, as well as OSX. I have 5 different Intel NUC's, a Cubox, minnow board, HP tower with I7, and the custom made board that's going in my streamer.
So I do have a bit of experience with this.
Last edited:
I found the Rotel RB-1572 a fine class D amp (for those looking for that type of sound)
It is on special on Canuckaudiomart for 1150$ CAN, a real bargain
I don't think you can beat Rotel class D technology, look at those boards , and I heard them played almost perfect amps...
It is on special on Canuckaudiomart for 1150$ CAN, a real bargain
I don't think you can beat Rotel class D technology, look at those boards , and I heard them played almost perfect amps...

I don't think you can beat Rotel class D technology, look at those boards , and I heard them played almost perfect amps...
So what, in that picture, tells us something about how they sound? 🙂
the filtering inductors looks awesome, and the pcb boards and the supporting heatsinks. Real craftsmanship
I listen to them, they were paired with B&W 805D and 803 after. Let just say before that unit I never heard a good sounding Class D amp. All the ones on ebay and from china sounds like garbage or sewage water poured in your ears.
I listen to them, they were paired with B&W 805D and 803 after. Let just say before that unit I never heard a good sounding Class D amp. All the ones on ebay and from china sounds like garbage or sewage water poured in your ears.
the filtering inductors looks awesome, and the pcb boards and the supporting heatsinks. Real craftsmanship
Sure, but it is not at all guaranteed that that would translate into good audio performance. See the Bruno quote above. 🙂
So what technology does the Rotel use?Let just say before that unit I never heard a good sounding Class D amp. All the ones on ebay and from china sounds like garbage or sewage water poured in your ears.
I found the Rotel RB-1572 a fine class D amp (for those looking for that type of sound)
It is on special on Canuckaudiomart for 1150$ CAN, a real bargain
I don't think you can beat Rotel class D technology, look at those boards , and I heard them played almost perfect amps...
![]()
I count 9 boards. The stereo Nord only has 3 so the Rotel must be better. I hope those are audiophile fuses 🙂
Last edited:

The feedback on this thread should end all bickering anyways. It's been absolutely outstanding. Far more positive than I've seen on any subjectivist forum about class D. The Nord's are the ticket to kill the misconception that class D is cold and sterile 🙂
http://www.audioshark.org/showthread.php?t=9369&page=43&p=173262#post173262
http://www.audioshark.org/showthread.php?t=9369&page=43&p=173262#post173262
The feedback on this thread should end all bickering anyways. It's been absolutely outstanding. Far more positive than I've seen on any subjectivist forum about class D. The Nord's are the ticket to kill the misconception that class D is cold and sterile
I don't think any of the "bickering" (could we please try to avoid the unnecessary derogatives?) has had anything to do with class D per se. I think we all agree the hypex amps sound good - and for those who perceive transparency as "cold and sterile", there are lots of ways to add the coloration they prefer.
Indeed. Hypex Class D has been enjoyed and respected for a number of years now. No surprises with the NC500. Rather we have been trying to discuss bavmike's claims of shortcomings in the recommended buffer design and just how changing the op amp in this design might make a significant change in the sound of the amp, a change so significant that its audibility is clear through the entire chain. We've demonstrated that it isn't because the op amp in the recommended design isn't being driven too hard. It has been shown that it isn't distortion (except perhaps those who prefer the Burson). Bavmike seemed to be suggesting recently - or perhaps I am mistaken - that it is due to rather typical kick up in low frequency noise in IC op amps. It's unfortunate that there isn't comparable data in the data sheet for the SIL994. Other ideas?
Also, I'm not sure we really reached a satisfactory conclusion regarding DC coupling. I understood that while the introduction of a DC blocking cap in a line input adds noise - just make it large to reduce the noise contribution - DC coupling of amps was still "unwise" practice. Is this now considered unnecessary?
Also, I'm not sure we really reached a satisfactory conclusion regarding DC coupling. I understood that while the introduction of a DC blocking cap in a line input adds noise - just make it large to reduce the noise contribution - DC coupling of amps was still "unwise" practice. Is this now considered unnecessary?
- Home
- Vendor's Bazaar
- Hypex NCore NC500 build