NAD M28 Seven Channel Power Amplifier Review

[cistercian]

Thanks for the review Amir! My take is that the amp is expensive and the performance is fine.
I would much rather not buy a 7 channel amp as it makes a great case for a rack of amplifiers instead!
But for those who don't geek out like I do on equipment I understand the appeal of one box.

I empathize with your concern over going backwards in spec too. Not what is wanted, especially at this price!
So...kinda meh I suppose. But certainly adequate.

 

[Chromatischism]

Our amp chart is getting pretty crowded. Can I make a suggestion to switch the axes like other review sites? It would be easier to read the names and it would be more phone-friendly, too.

https://www.anandtech.com/show/15787/amd-ryzen-5-3600-review-amazons-best-selling-cpu/4

 

[Kal Rubinson]

If NAD would have designed the amp boards themselves that would be pretty good.
But they took the excellent purifi modules and degraded their performance a fair bit.

NAD builds these modules themselves.

 

[Vasr]

NAD builds these modules themselves.

True. NAD builds these boards using the licensed Purifi design. Don't really know if that license includes the board design or just the circuit design.

Any OEM Asian company would be able to do it though. The circuit design is the hard part.

I think he meant the Purifi design in the second sentence (otherwise, it would be difficult to degrade it).
But it is not clear what else in the unit is entirely NAD's and where exactly the grounding issues may have cropped up.

 

[Vasr]

In my opinion it does score high in the price/performance of any amps, that was my point. I find it hard to fault a amp that measures in the top 10 since the introduction of this site.

This is where I differ from you and find it an unfortunate use of that single SINAD metric (and this is not even considering that it is an accidental collection of devices based on what were sent in), a valid criticism of ASR.

As a consumer, which was the context, any amp beyond a certain threshold (let us just say 96db as an example) is certainly not better performance in any way that would matter audibly. So being in the top 10 of that ranking is uncorrelated with any performance relevant to me at least in considering the potential candidates for purchase. But it doesn't mean I would buy anything or the cheapest in that group.

There would certainly be additional subjective criterion like looks and engineering. I can appreciate people buying a well constructed watch for its looks and the engineering that went into it which exceeds others in its class. There is a premium to be paid for it. But this is not a well engineered product objectively. It gets its ranking up there just because of the great licensed design used.

An analogy would be a car model where a brand licensed an impeccable performance engine from another company, put a nice chassis on it but did a mediocre job of the engine bay reducing the potential of the engine. But the engine was still good enough to be in the top 10 0-60mph acceleration ranking in its class. Is it worth paying a high price for it? Perhaps if you liked the looks and the bragging rights just because of one specific ranking (even if it made no difference to daily life) but that is like getting a "trophy wife" no?

If this amp module design was exclusive and the only way to get such a high performance amp was via a mediocre integration, then at least that would be understandable. But Purifi can be licensed by anybody and will likely make it into many more units in the future.

It is even weaker argument than the Dirac enabled units. At least Dirac makes the sound great but one wouldn't accept a mediocre engineered unit because it will be overshadowed by Dirac effects... hmmm noticing a pattern here with contemporary NAD products...

 

[PeteL]

True. NAD builds these boards using the licensed Purifi design. Don't really know if that license includes the board design or just the circuit design.

Any OEM Asian company would be able to do it though. The circuit design is the hard part.

I think he meant the Purifi design in the second sentence (otherwise, it would be difficult to degrade it).
But it is not clear what else in the unit is entirely NAD's and where exactly the grounding issues may have cropped up.

Sorry I might have miss some post in this debate. Did someone flagged an issue on how the grounding is implemented on the amp? What is the issue? With high speed switching lines the board design is key for performance, of course not as hard as the circuit, but I actually doubt that Purifi would show it's full schematics, they would give the gerbers and it would allow NAD to modify some more obvious parts. By specifications alone, I am quite certain that the input buffers are NAD design, more stuff to, protections, heat sensors, the brain. The power supply, it's hard to tell but if I base on M27 inside pictures, this one look like something that Hypex would have designed, but with some mods, mainly real estate management tough. M28 we'll have to wait and see. I still disagree that "it would be difficult to degrade it" For the reasons I already said. It's a more complex beast than the Purify ref design. The more you add stuff.... As mentioned before tough, I'd like to see the results with channel 1 8 loaded, rather than 7-8, just for fun, just this is likely to make a small difference, these modules are close together

 

[HTNut1975]

This is where I differ from you and find it an unfortunate use of that single SINAD metric (and this is not even considering that it is an accidental collection of devices based on what were sent in), a valid criticism of ASR.

As a consumer, which was the context, any amp beyond a certain threshold (let us just say 96db as an example) is certainly not better performance in any way that would matter audibly. So being in the top 10 of that ranking is uncorrelated with any performance relevant to me at least in considering the potential candidates for purchase. But it doesn't mean I would buy anything or the cheapest in that group.

There would certainly be additional subjective criterion like looks and engineering. I can appreciate people buying a well constructed watch for its looks and the engineering that went into it which exceeds others in its class. There is a premium to be paid for it. But this is not a well engineered product objectively. It gets its ranking up there just because of the great licensed design used.

An analogy would be a car model where a brand licensed an impeccable performance engine from another company, put a nice chassis on it but did a mediocre job of the engine bay reducing the potential of the engine. But the engine was still good enough to be in the top 10 0-60mph acceleration ranking in its class. Is it worth paying a high price for it? Perhaps if you liked the looks and the bragging rights just because of one specific ranking (even if it made no difference to daily life) but that is like getting a "trophy wife" no?

If this amp module design was exclusive and the only way to get such a high performance amp was via a mediocre integration, then at least that would be understandable. But Purifi can be licensed by anybody and will likely make it into many more units in the future.

It is even weaker argument than the Dirac enabled units. At least Dirac makes the sound great but one wouldn't accept a mediocre engineered unit because it will be overshadowed by Dirac effects... hmmm noticing a pattern here with contemporary NAD products...

Here’s another way of looking at it: what‘s the absolute cheapest way to get seven channels of Purifi?

 

[Doodski]

I'd like to see the results with channel 1 8 loaded, rather than 7-8, just for fun, just this is likely to make a small difference, these modules are close together

I also would like to see a AC current measurement with 1 channel driven to clipping. Then compare that to the actual fuse size at the rear panel.

 

[PeteL]

I also would like to see a AC current measurement with 1 channel driven to clipping. Then compare that to the actual fuse size at the rear panel.

I am quite certain that NAD built fully compliants and safe amps. I can't think of audio content or situations where all channels are driven at full rms power simultaneously... Actually Amir's test is quite a stress test, not real life.

 

[Doodski]

I can't think of audio content or situations where all channels are driven at full rms power simultaneously... Actually Amir's test is quite a stress test, not real life.

Testing to clipping is the usual stress test, has been for a long time in the industry and there's nothing unusual there.

 

[Vasr]

Sorry I might have miss some post in this debate. Did someone flagged an issue on how the grounding is implemented on the amp? What is the issue?

The current hypothesis to explain the rise in noise with power output before clipping (which I believe is usually the opposite of what you see)
https://www.audiosciencereview.com/...nnel-power-amplifier-review.15939/post-512117
by someone who designs circuits and corroborated soon after by another who seems to be hands-on as well. I have no technical knowledge of these things.

I agree with rest of your post. There is plenty of things in there for NAD to screw up.

Here’s another way of looking at it: what‘s the absolute cheapest way to get seven channels of Purifi?

That question makes no sense to me at this point in time. I am not really the type who asks what is the cheapest way to buy a computer with the latest Intel chip just when it is making its way into computers, typical consumer behavior I know.
But I do know people who do want to be the first kid on the block to own one...

I am the kind who asks why would you want to buy a Purifi based multi-channel now at a premium when there is just one of it, not the best integration of it and does not measure any better than a previous model nor would it sound any different. But that is just me.

 

[PeteL]

This is where I differ from you and find it an unfortunate use of that single SINAD metric (and this is not even considering that it is an accidental collection of devices based on what were sent in), a valid criticism of ASR.

As a consumer, which was the context, any amp beyond a certain threshold (let us just say 96db as an example) is certainly not better performance in any way that would matter audibly. So being in the top 10 of that ranking is uncorrelated with any performance relevant to me at least in considering the potential candidates for purchase. But it doesn't mean I would buy anything or the cheapest in that group.

There would certainly be additional subjective criterion like looks and engineering. I can appreciate people buying a well constructed watch for its looks and the engineering that went into it which exceeds others in its class. There is a premium to be paid for it. But this is not a well engineered product objectively. It gets its ranking up there just because of the great licensed design used.

An analogy would be a car model where a brand licensed an impeccable performance engine from another company, put a nice chassis on it but did a mediocre job of the engine bay reducing the potential of the engine. But the engine was still good enough to be in the top 10 0-60mph acceleration ranking in its class. Is it worth paying a high price for it? Perhaps if you liked the looks and the bragging rights just because of one specific ranking (even if it made no difference to daily life) but that is like getting a "trophy wife" no?

If this amp module design was exclusive and the only way to get such a high performance amp was via a mediocre integration, then at least that would be understandable. But Purifi can be licensed by anybody and will likely make it into many more units in the future.

It is even weaker argument than the Dirac enabled units. At least Dirac makes the sound great but one wouldn't accept a mediocre engineered unit because it will be overshadowed by Dirac effects... hmmm noticing a pattern here with contemporary NAD products...

You bring good points, it's a matter of philosophy I guess. By the way if as Amir suspect, it's the loading that could cause the performance compromise, I have to say I did the same with my own nc500 monoblocks built. I added 3 dB of gain to the suggested Hypex buffer design, works better for me, but My very ancient (calibrated) audio precision analyser won't get anything useful... So it is very likely that my amps don't give me the full performance, but no ways to know, certainly not something I can hear...

 

[PeteL]

Testing to clipping is the usual stress test, has been for a long time in the industry and there's nothing unusual there.

I'm talking at the one where 5 channel delivers 262W simultaneously

 

[HTNut1975]

I am the kind who asks why would you want to buy a Purifi based multi-channel now at a premium when there is just one of it, not the best integration of it and does not measure any better than a previous model nor would it sound any different. But that is just me.

Not sure if the price is at a premium. Take a typical 20 percent off NAD msrp and that gets you about $571 a channel. Not bad even with these tests, no?

 

[PeteL]

The current hypothesis to explain the rise in noise with power output before clipping (which I believe is usually the opposite of what you see)
https://www.audiosciencereview.com/...nnel-power-amplifier-review.15939/post-512117
by someone who designs circuits and corroborated soon after by another who seems to be hands-on as well. I have no technical knowledge of these things.

I agree with rest of your post. There is plenty of things in there for NAD to screw up.

Oh I see, It's a general observation, not an analysis of the design. A poorly tought trough grounding scheme usually leads to a noisier design yes, he is not wrong, I still stand that it's a sum of small things that ads up to create a noisy design, but again, only compared to reference, for it's own sake, It's a very low noise design. Also, I believe that there is a word missing in @JohnYang1997 comment. He ment to say:
The rised distortion is obviously due to loading. The ((NOISE)) issue is very likely to be ground connections.

 

[Doodski]

After @amirm hopefully does a tear down pics collage I suspect we will see a back plane of sorts connecting all this internal circuitry together. That's been used extensively in other modern NAD amps for the layout configuration.

 

[PeteL]

After @amirm hopefully does a tear down pics collage I suspect we will see a back plane of sorts connecting all this internal circuitry together. That's been used extensively in other modern NAD amps for the layout configuration.

I doubt it will be very different than the M27, we saw it, it's a good design, Personally I think this 3 dB SINAD difference is negligible, I expect a beefier power supply, we didn't see the multiple channel loaded test on the M27, so it's not clear that for noise and distortion alone, the M27 is the better product. I think we have to go beyond this test suite to fully asses Purifi.
Putzeys is all about maths and transfer function. If he took the already phenomenal nCore and tought that there was room for improvment and perform tweeks to lead to Purifi module, Personally, and it might not agree with the philosophy here, but I trust that Purifi is a "more perfect" amp module than nCore. I think it's a key difference, we all believe in mesurments, yes Putzeys too, but I think after he designed the amp that measured as perfect as he could, he still digged for the little stuff, His thing is to demonstrate that the maths are exact. You still get the Higher end product even if Nad allowed a bit more noise using it.

 

[pma]

Sometimes it is a one centimeter of a PCB track that shares a signal ground and supply bypass capacitor ground what makes a rise of distortion with load. I know what I am speaking about.

 

[JohnYang1997]

Oh I see, It's a general observation, not an analysis of the design. A poorly tought trough grounding scheme usually leads to a noisier design yes, he is not wrong, I still stand that it's a sum of small things that ads up to create a noisy design, but again, only compared to reference, for it's own sake, It's a very low noise design. Also, I believe that there is a word missing in @JohnYang1997 comment. He ment to say:
The rised distortion is obviously due to loading. The ((NOISE)) issue is very likely to be ground connections.

Distortion.

 

[JohnYang1997]

Sometimes it is a one centimeter of a PCB track that shares a signal ground and supply bypass capacitor ground what makes a rise of distortion with load. I know what I am speaking about.

Yep. However I believe purifi modules are connected using wires. Or the input board.