NAD M28 7 Channel Purifi Amplifier Teardown

[carlob]

So you want to buy an amp which is NAD but is made of one hypex spms and seven hypex modules (why not also the hypex buffer at this point) let alone if the smps is able to drive seven nc500 to clipping. Furthermore NAD is not allowed to develop it's own SPMS in its own product because the Hypex one is cheap and replaceable. So basically you're saying that NAD should be Apollon, or Nord or whatever.
At this point why all the other manufacturers don't do the same, just slap a couple of hypex modules and a hypex spms in a chassis and call it a day. Possibly crafted in a way so that a user can open it on day one and replace all the capacitors with ones of his liking. Or maybe they could sell their products with the option to mount the capacitors of your choice.

Stop the craziness please.

 

[kn0ppers]

It feels like some of you are too emotionally invested in this debate.

 

[HTNut1975]

If you don't want to understand thats fine...

One last try:
https://www.hypex.nl/img/upload/doc/smps/smps3k/Documentation/SMPS3KAx00_10xx.pdf
3000 Watt Output Power

You can connect as many amps as you want to it but you are limited to 3000 Watt Output.
That PSU is 290 € incl. tax single quantity end user price.

And please don't tell me the NAD M28 PSU can output 3000 Watts because it can't.

How do you know that the NAD custom power supply costs more than this Hypex? For all we know, it costs less for NAD. Is your complaint that NAD should have used something like this Hypex part so that DIY’er’s could fix it down the road? The complaint couldn’t be about cheap caps because this Hypex uses the cheap caps as well, no? (Forgive me if I’m misunderstanding.)

 

[Michael YYZ]

Carlo, had NAD used a Hypex in the M28, all this crowd would have been up in the arms saying that NAD was not able to design their own powers supply and all they did was buy OEM modules and fit them in a box. Can‘t please this crowd.

We don’t know why NAD went this route. Maybe it was cost; or, perhaps they needed to design a PS to their own specs and nothing else was available. If the guys here are not happy with NAD’s approach, tney should not buy their products. It’s as simple as that. The grapes that are out of reach feel sour.

For one, I really enjoy my M28 and I think I should probably spend less time here and more listening to music. So addictive!....

P.S. Love to Italy!

So you want to buy an amp which is NAD but is made of one hypex spms and seven hypex modules (why not also the hypex buffer at this point) let alone if the smps is able to drive seven nc500 to clipping. Furthermore NAD is not allowed to develop it's own SPMS in its own product because the Hypex one is cheap and replaceable. So basically you're saying that NAD should be Apollon, or Nord or whatever.
At this point why all the other manufacturers don't do the same, just slap a couple of hypex modules and a hypex spms in a chassis and call it a day. Possibly crafted in a way so that a user can open it on day one and replace all the capacitors with ones of his liking. Or maybe they could sell their products with the option to mount the capacitors of your choice.

Stop the craziness please.

 

[Vasr]

The grapes that are out of reach feel sour.

Obnoxiously presumptuous.

 

[Vincentponcet]

Uhmmm, no, not sarcasm. Why do you ask? It cost me ~$4,500 + tax, I probably won't be able to have it sold for anything more than $4,000 and there's a 40% consignment sale fee. Altogether, I may be able to recoup ~$2,500 if I'm lucky.

Casey

40% commissions for the store just to put it on the shelf?
Why not selling direct ?
A product which is very new, if you sell it direct, you should not take more than 30% loss, which is still not small.
In your case, like you, I would be not happy about the measurements, worse than the reference amp.
But I would most probably not change it. It is still a very good multi-channels amplifier.
I was going to buy it, but now I'm waiting for Apollon which announced a multi-channel version of their Purifi mini, they said they will send one for test to Amir.

 

[mocenigo]

@amirm
It's not the first time that nad does that, the channel difference is that 3 of them has absolute polarity inverted from what I understand.. It reduce the strain on the power supply.
And then invert the speakers connectors obviously.

Yep. Some channels are wired in AND out out of phase to reduce bus pumping.

 

[bobof]

@amirm
It's not the first time that nad does that, the channel difference is that 3 of them has absolute polarity inverted from what I understand.. It reduce the strain on the power supply.
And then invert the speakers connectors obviously.

If you read the Hypex SMPS3kxxxx PSU manual (a different unit to this one it seems, but I guess likely similar issue) they recommend this strategy as the PSU doesn't deal well with the pumping from half bridge amplifiers. Or alternatively, bridging two amp modules into a single amp (not economic or feasible here for 7ch in one box).

Of course, in the case where two modules are bridged, one half is inverted, and the input signal is identical, so there should be an almost perfect cancelling of the ill effects as far as the PSU might see.

However, in a 7 channel application, there is clearly an imbalance already (4 channels in one polarity vs 3 in the other), plus the signals may vary significantly. It's possible that a user might unwittingly decide to use just the 4 channels that are in one polarity, and ignore the other 3. Or put all their LCR on one polarity, and all the surrounds on the other, which would also be a likely imbalance.

So I wonder how well this actually works in practice, and if certain configurations of channel usage are much worse? It seems you'd likely get best test results if you tested two channels on opposite phases (simultaneously and with identical signals). And for worst results, 4 channels on the same phase all being driven with identical signals. I guess the test in the original review (6&7) is a halfway ground maybe.

 

[bobof]

Put the unit back together and tested channels 1 and 3 and results are almost the same:

View attachment 82861

SINAD remains the same so noise is the same. There is a bit less 2nd harmonic distortion (almost not seen above) but as I was preparing to show that comparison, it started to rise so this may be temperature dependent.

It would seem that both combinations tested were channel pairs in identical phase. Would have been interesting to test two channels that were out of phase and see if the performance was improved (though perhaps PSU pumping issues don't manifest at these levels?)

 

[Kal Rubinson]

It would seem that both combinations tested were channel pairs in identical phase.

FWIW, NAD's recommendation is:
• For optimum heat dissipation in an AV surround system, the following audio channel allocations are recommended:
Channel 1 = Front Left speaker
Channel 2 = Surround Left speaker
Channel 3 = Surround Back Left speaker
Channel 4 = Center speaker
Channel 5 = Surround Back Right speaker
Channel 6 = Surround Right speaker
Channel 7 = Front Right speaker

 

[bobof]

I guess that is just a common sense approach, with the likely most power hungry channels maximally distributed through the chassis, which also gives alternate phase to each half of an L/R pair.

I wonder if end users following that connection scheme may get slightly better results than measured here. Content dependent I guess.

 

[Descartes]

NAD is warranties for 5 years, and dealer cost is probably 40 to 45 points off retail I would probably buy it only if o can get a discount

 

[pseudoid]

Will you loose 2.5k because a few people on a forum do not like one capacitor brand? Not my business, but doesn't seem very reasonable...

Right on! I am surprised no one has complained about the cost (or the white color, for that matter) of the RTV used below the main caps on the power board...

 

[pseudoid]

...As an engineer it always pisses me off when I see other engineers do substandard work...

As a 'concerned' engineer [as; apparently you are, and, as; we all should be>> EEs or otherwise] getting 'p*ssed off' achieves nil for you and much better than getting 'p*ssed ON' by being a substandard designer. Let your work do the talking, while you continually aspire to achieve your best output, thanking those failures of others to further your achievements. Hopefully, before your retirement years because comparing yourself to others in your field is but a counter-productive endeavor.
There is no disrespect to you or what you wrote (or intent of this reply) and it is justified but never productive (i.e. when comparing to others).

 

[MakeMineVinyl]

Right on! I am surprised no one has complained about the cost (or the white color, for that matter) of the RTV used below the main caps on the power board...

If we are looking for a villain, we should look in the mirror because we are getting exactly what we want - products for as cheap as possible. Think that's sustainable? Its exactly the Harbor Freight Tools syndrome. Fair quality, but don't expect the tools to last beyond a couple of jobs.

There's absolutely no incentive for a manufacturer to use premium parts like capacitors if the cheaper ones are 'good enough' and the component has to be sold for as little as possible - and by the time they fail, people will just chuck it up to 'nothing lasts more than a couple years anymore' and buy a new component.

Obviously not all manufacturers use the cheapest possible components, reliability be damned - thankfully I work for just such a company - but something has to give somewhere. The extra quality costs more money which many people don't want to spend. So they spend the least possible money, the component fails in a couple years, and they bitch, and they move on.

How many threads have I seen on this very forum along the lines of 'what is the cheapest class D amp I can buy'?

 

[Kendo]

Im guessing nobody from Nad got back regarding the test results?

 

[danielsmi]

I recently reviewed the NAD M28 multichannel amplifier based on Purifi modules. As promised, this is a tear down of the unit with kind permission of the owner.

The amp was a delight to open with beautiful screws holding the attractive top down. Once removed, we are greeted to a nice and modular layout:

View attachment 82854

The right side is a unified power supply for all 7 channels. I searched quickly but could not find a matching Hypex power supply. The only markings on it are:

View attachment 82855

As you will see later, NAD marks their boards with their name so I suspect this was custom designed by someone and not NAD.

There was a question about the fuse size. The fuse is rated at 15 amps and 250 volts. Total volt-amps available then is 1800 assuming 120 volt mains in US. Assuming there is power factor correction, this translates into equivalent wattage minus some loss. Further assuming 90% efficiency we have 1600 watts available to all 7 channel amps. This is likely the upper bound as your 15 amp circuit won't tolerate that for log so maybe it is in the range of 1,200 to 1,500 watts of power supply juice.

The capacitor brand choice is disappointing in the power supply as it is the low tier Lelon brand and only rated at 85 degrees C despite being very close to the heatsink:

View attachment 82857

Moving on to the amplifier module, I was very relieved that it could be removed with just 5 screws:

View attachment 82856

The green board is the Purifi 1ET400A module. The capacitors have been replaced from Rubycon in the original reference design to United-Chemi-Con which is fine.

The Purifi is surrounded by the blue buffer board designed by NAD. The two op-amps are:

National/TI LM49720 "Dual High Performance, High Fidelity Audio Operational Amplifier " : https://www.ti.com/lit/ds/symlink/lme49720.pdf?ts=1599982501986&ref_url=https%3A%2F%2Fwww.google.com%2F

Microchip MCP6002 is a dual general purpose op amp: https://www.microchip.com/wwwproducts/en/MCP6002

Lower rate capacitors are used here from Samwha (old Samsung capacitor factory that was sold years ago):

View attachment 82859

Back tot he board, note the two tables above the text "channel differences." There are configuration differences for channels 1,2,3,4,5 and 5,6,7! The board I pulled out is channel 6 and is marked that way with a red dot. The channels I tested in the review were 6 and 7. What possible reason would there be for the right three channels to have different configuration? When I put it back together I will do a spot check of other channels to see if the performance is different.

Conclusions
Mechanically the NAD M28 is beautifully put together. Should any of the amplifiers fail, you can swap parts to find out if it is the buffer board or the Purifi module. And possibly source Purifi modules in the open market to fix.

The power supply is a different matter as it seems to be a custom product. And sadly it uses less than ideal parts for long term longevity. Someone was pinching pennies for no reason here. What is another $50 in material cost in the power supply with better branded/temperature rated capacitors?

------------
As always, questions, comments, recommendations, etc. are welcome.

Getting serious cabin fever with smoke filled air from forest fires keeping me from going out. Once it clears -- which it is supposed to do by tomorrow -- I am going to go wild and go buy something expensive to make myself feel better. Problem is money as always and I need good bit of it. So please donate what you can using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Here's some internal pics of the M23 I took. It looks like the M23 uses different capacitors in the SMPS than the M28. Anyone know they're any better than the M28's? I think the brand is DTDZ

 

[Descartes]

Here's some internal pics of the M23 I took. It looks like the M23 uses different capacitors in the SMPS than the M28. Anyone know they're any better than the M28's? I think the brand is DTDZ
View attachment 251423

View attachment 251424View attachment 251430View attachment 251429View attachment 251428View attachment 251425View attachment 251426View attachment 251427

Thank you for posting!

 

[pseudoid]

Here's some internal pics of the M23 I took.

Thank you for posting the internal pics.

What are these daughter-cards on 12pin headers: OpAmps? Pics?

 

[danielsmi]

I'm no expert but I think those are Op-amps