Review and Measurements of Nord One NC500 Amp

[NielsMayer]

IMHO, both of these builds are total fails from an RF perspective. They don't even follow some of the simple hints put out by Bruno P. in constructing these amps. Even though the "apollon" looks cleaner, it is actually far worse from an RF perspective.

(1) long input leads "in case" (inside case, each long balanced wire shields is a giant antenna to pickup the switching frequency).
(2) no attention paid to "pin 1 problem" (long pin-1 ground -- probably going all the way from XLR plug and grounding at the board -- is a giant RF antenna).
(3) output leads crossing and near input leads. (i'd expect problems on the right channel of the apollon where both inputs cross directly near right channel output).
(4) long and asymmetrical routing (on LHS channel of apollon) of power cabling.

The entire physical layout of the apollon unit is about as stupid as you can get. Asymmetrical for one and designed more for external appearance than electrical performance.

For details, see http://www.rane.com/note165.html "Pin 1 Revisited"
"Figure 3. Internal conductors (including PCB traces) that connect to pin 1 can act as antennas. Radio frequencies flowing on the cable shields are re-radiated inside of the unit. In reverse, radio frequencies generated inside of the unit (DSP clock, etc.) can exit the unit via pin 1 and contaminate surrounding equipment." (and with a class-D amp that is bascially an AM radio transmitter inside the case that's a lot of potential contamination)

https://www.rane.com/note151.html ] (Grounding and Shielding Audio Devices)

The issue is that with a wide-band amp, you have to be very careful about what kind of noise and RF is being fed to the inputs.
https://www.diyaudio.com/forums/class-d/321632-hypex-ncore-nc400-input-anti-alias-filter.html

(consider the "aliasing" problem as switching frequency from once channel leaks into the input of the other in a non-monoblock build -- imho these amps should never be setup in anything other than a monoblock configuration or use two separately shielded sub-cases inside a stereo configuration -- and they should never share a power supply for that precise reason).

Apollon NC800SL
Picture optimized

View attachment 27068


Nord One NC500 Amp

The choice is evident to me.

 

[Burning Sounds]

Apollon NC800SL
Picture optimized

View attachment 27068


Nord One NC500 Amp

The choice is evident to me.

One of the reasons I chose Apollon for my multichannel Hypex-based amp (ASR thread here) was because of their tidy wiring and the potential for EMI etc from mutiple modules in the same case. Regarding the perforated inner case on the Apollon - its effectiveness will depend on what air vents are on the outer case at the bottom. Mine only has one row of vents at the bottom (in the rear), although the top is fully ventilated.

 

[maty]

I just sent an e-mail to Apollon:

Hi from Spain,

I think you must read this comment about your implementation.

https://www.audiosciencereview.com/...ements-of-nord-one-nc500-amp.7704/post-181633

- maty -

 

[audioBliss]

Who needs more than 400W into 4ohm?

Despite advertised amp power specs, most speakers can’t handle more than 150W into 8ohm, many can’t even handle 100W. Power recommendation specs are meaningless and are different than wattage handling specs.

Well it's not 400W 4Ohm both channels driven. It's 225W both channels driven in 4Ohm. My speakers can handle 280W in 4Ohm continously and 5000W peaks over 400Hz. Not that I'd ever want to play that loud. It probably nothing to worry about but on good dynamic recordings the power draw can be insane. On normal crappy masters I probably never use more than 1W.

When I think about it the NC400 should be enough and as someone already stated above it does seem the most safe to just use real monoblocks like the original kit on hypex website. Now I'm just a bit worried about any potential problem using XLR to RCA cables(my Arcam only has RCA out). Anything to think about here?

 

[Samoyed]

This is a review and detailed measurements of the Nord One Hypex NC500 class-D module based power (speaker) amplifier. It is on kind loan from a member who recently purchased it. Nord is a British company so lists the cost of the Nord One as £ 979 or US $1,250 as of this writing. For that cost you get a pair of NC500 class-D modules from Hypex with dual PS1200 power supplies. I *think* this cost includes the stock input "buffer" (amplifier) from Hypex. You also have a choice of fancier, audiophile-type input buffer.

The Nord One comes in a very large enclosure (full 19 inch rack width):

View attachment 27025​

A hard push button turns the unit on and off. Even with overdriving the unit, I could not get it to go into protection mode so not sure if such a feature is there or not.

The back panel has the bare minimum plus trigger sockets:

View attachment 27026​

The unit as it came had an issue. Can you spot it?

View attachment 27027

Look at the right channel: the red cable is connected to black terminal instead of red. So the two channels were out of phase from each other. I can just imagine someone buying this and raving about soundstage and such, only to realize that it is a defect, not a feature. Anyway, I fixed it for the measurements you are about to see.

The speaker terminals are massive by the way.

While we are looking inside, here is what it looks like (pardon the mobile phone picture):
View attachment 27028

Not a whole lot in there as all the heavy lifting is done by Hypex in producing the switching power supplies, the NC500 amplifier modules and an "input buffer" board which seemingly does more. Main power is routed through it rather than directly to the amplifier modules. Speaker output also goes through it.

The large enclosure allows the components to be easily laid out. Wiring is generally neat and is better than most non-skilled people could do. It seems like good value to me for all the functionality that is in there and hand assembly of the unit.

Power Amplifier Audio Measurements
As usual, we start with our 5 watt output into 4 ohm load for the dashboard:
View attachment 27029

I was a bit unhappy about the channel in blue which seemed to dance around with those extra spikes. The levels are low but not nice engineering.

Overall THD+N as expressed with SINAD is in very good category:

View attachment 27030

At 96, it is well above our geometric mean of all the amplifiers tested so far (80). It does lose by a big margin compared to the NC400 though which ironically was a DIY build with a single power supply.

Let's see what the power output looks like relative to distortion:
View attachment 27031

I worked hard to get the distortion as low as I could as Hypex data is slightly better than this. I realized that when the unit is dead cold, it performs a bit better. I suspect Hypex has tested the amp some other way than with that eval input buffer board as distortion seems to set it very early in my testing at around 2 watts whereas Hypex doesn't show this effect:

View attachment 27032

Overall power versus distortion curve though does match what they show with that increasing distortion from 5 to 30 watts and then settling down to a lower number. It is unfortunately to see some non-linearity in that range where a lot of listening will be done.

Anyway, the good news is that the dual power supplies do their job, helping the NC500 amplifiers to produce copious amount of power to the tune of 445 watts. Combine this with excellent signal to noise ratio and you have a lot of clean power:

View attachment 27033

Frequency Response is uneventful:
View attachment 27034

By the way the above measurements were made without any AES-17 filter. I was asked to evaluate this amp with and without it so I thought I show that in the power measurements at 8 ohm:

View attachment 27035

The graphs with and without AES-17 filters are essentially the same. Below 1 watt, the output level is low enough that the analyzer gets a bit confused with the ultrasonic noise and glitches up and down a bit. So no, AES-17 filter (with 40 kHz flat response) that I use do not give any advantage to class-D amps.

Speaking of ultrasonics, here is the output without an AES-17 filter:
View attachment 27036

The switching noise is very high at just -12 dB below signal level. Would have liked to see this lower. The Nord One has no CE mark so would be interesting to see if additional filtering would have been necessary to pass certification. Of course it is not audible so it is OK in that regard.

I was surprised that there was less intermodulation distortion at high frequencies than our best measured yet Benchmark AHB2:

View attachment 27037

I will have to double check to make sure the Benchmark was also tested at 5 watts. There is however harmonic distortion that is much higher than the Benchmark starting at 35 kHz.

We can see that in this graph of THD+N versus Frequency at a few different power levels:
View attachment 27038

The red graph is at lowest power (probably less than 1 watt) and the pink is at max (440 watts). The rise in distortion above 1 kHz is too steep in my liking. In that regard, the 1 kHz measurements we usually perform are the best case scenario.

Conclusions
I went into this review expecting better results than I achieved. Despite higher designation, the NC500 seems to be a lower performing module than NC400. How much the input buffer board adds to this, is unknown until we test other variations.

That aside, you are getting huge amount of clean power here that outdoes all of the bargain amplifiers we have tested so far (switching or not). At nearly 500 watts into 4 ohm, you should have ample power to drive anything from subwoofers to main speakers. The extremely high signal to noise ratio should be silent in use. Price is reasonable assuming shipping is not too expensive and it includes the input buffer.

In other words, we have found another power amplifier that I can recommend.

----------------------
Questions, comments, critique, etc. are welcome.

My late grandfather-in-law used to say, "it is hell getting old." To wit, I don't remember if I asked you all to send me money so I can buy a boat and enjoy the beautiful weather we are having. Regardless, send me the money anyway using:

Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).

It is odd that a limited production hand assembled amp would leave the manufacturer with its channels out of phase. I would have thought a listening test would be part of the final QC process. On the other hand, perhaps the builders are objectivists, who have been scientifically proven to be unable to hear this difference.

Colin had an employee, now replaced, who wasn’t up to snuff...maybe he was at fault. My revision D, with Sparkos, is amazing!

 

[Deleted member 2944]

I just sent an e-mail to Apollon:

You are quite the little busybody, aren't you?

Dave.

 

[maty]

I just had a nice fresh espresso coffee. And the company deserves to be aware of that comment and we know the answer.

 

[Samoyed]

Christ be Jesus, I hope the guy comments but the tenor of your email is, well, edgy.

 

[maty]

I am not nervous because I do not have a class D amplifier . Tweeter and class D amp... I like more as KEF LS50 wireless: class D to the woofer (and subwoofer) and class AB to the tweeter but...

In real life, I usually contact someone with the capacity to make decisions, so I try to do the same in the virtual world.

I still wait for an answer to my old emails to some audio companies. That they do not respond is significant.

 

[NielsMayer]

FYI although there are no "continuous power" rating on the Hypex NC400, a similarly-powered NC502 module lists 100W Continuous power output per channel at 25C per Hypex NC502MP Datasheet pp. 6.

Also, people building these amps should read the instructions first (especially so-called "professional builder" examples pictured in this thread which are absolutely heinous from an RF engineering perspective).

Hypex NC400 Datasheet

It is very important to use twisted pairs for the output cabling, at the very least anywhere near the PSU cabling. The distortion of the NC400 is so low that it is easily exceeded by magnetic coupling between the supply wiring and the audio wiring.

pp.10

9.1Cable dressing
The NC400 module has exceedingly low distortion. This makes it very easy for extraneous causes to add much more distortion and colouration than the amplifier itself. The first major cause of such distortion is direct magnetic crosstalk from the supply cable into the audio input or the loudspeaker output. This is minimized in several ways:
1) Run the audio and power supply cables away from each other.
2) Tie-wrap the supply cable to form a tight bunch.
3) Tightly twist all loudspeaker cabling inside the chassis.

9.2Input configuration

The second major cause of extraneous noise and distortion is common-impedance coupling between power wiring and signal connections. This is caused by the use of ground-referenced signalling (aka single-ended or unbalanced transmission), which unfortunately is the predominant method of interfacing consumer-grade audio equipment. As a result the ground circuit becomes an inseparable part of the signal path and any currents flowing through the ground circuit affect the audio signal.Ground loops cannot fundamentally be avoided. Star grounding only works at low frequencies. The longer the wires to the “star”, the lower the frequency at which supply decoupling becomes ineffective. The maximum permissible trace length of decoupling capacitors in class D amplifiers is measured in millimetres and the value of the decoupling capacitances is large. To make a long story short: you cannot connect the audio signal using a single wire and hope to solve the ensuing drama using star grounding. It won’t work.

pp. 16.

 

[MZKM]

Well it's not 400W 4Ohm both channels driven. It's 225W both channels driven in 4Ohm. My speakers can handle 280W in 4Ohm continously and 5000W peaks over 400Hz. Not that I'd ever want to play that loud. It probably nothing to worry about but on good dynamic recordings the power draw can be insane. On normal crappy masters I probably never use more than 1W.

When I think about it the NC400 should be enough and as someone already stated above it does seem the most safe to just use real monoblocks like the original kit on hypex website. Now I'm just a bit worried about any potential problem using XLR to RCA cables(my Arcam only has RCA out). Anything to think about here?

No, it’s 400+ Watts into 4ohm. As previously pointed out, the version Amir tested a while back was using a monoblock power supply in stereo configuration, giving almost 1/2 the wattage.

I can also almost gaurentee their’s not a (consumer) speaker on Earth that can handle 5kW on power.

 

[Samoyed]

FYI although there are no "continuous power" rating on the Hypex NC400, a similarly-powered NC502 module lists 100W Continuous power output per channel at 25C per Hypex NC502MP Datasheet pp. 6.

Also, people building these amps should read the instructions first (especially so-called "professional builder" examples pictured in this thread which are absolutely heinous from an RF engineering perspective).

Hypex NC400 Datasheet

pp.10

pp. 16.

Well, subjectively, my Nord is wonderful. So, maybe my ears aren’t up to snuff....

 

[audioBliss]

No, it’s 400+ Watts into 4ohm. As previously pointed out, the version Amir tested a while back was using a monoblock power supply in stereo configuration, giving almost 1/2 the wattage.

I can also almost gaurentee their’s not a (consumer) speaker on Earth that can handle 5kW on power.

Yeah I'm actually not really sure what to make of that specification of my speakers. But they can handle a lot of power since they are constructed to be crossed over to subwoofers. They are not meant to be played alone full range which seems to have the benefit of them being able to play incredibly loud at very low distortion above the crossover frequency of about 80Hz. Anyway on to topic.

So basically there no know deal breaking faults with the nc400 monoblocks?

Edit: Btw my speakers are not really cosumer speakers. I've written about them in the members area.

 

[levimax]

Lets do the maths (I hate this as I am somewhat mathematically inept so please check the numbers )

Just went to Madisound and at random picked a Scanspeak D2604 tweeter. It parameters are below.

View attachment 27065

So we want to calculate the power dissipated by the tweeter caused by the switching frequency. First we need to calculate the impedance of the coil at the switching frequency. We will use 400kHz. We will ignore a few things but it will be ballpark.

Impedance Z = 2 * Pi * F * L

Z= 2*3.14*400000*0.00002
Z= 50.24
+R = 53.14 ohms

This value is ignoring the effect of any speaker cabling or crossover in circuit which will increase the impedance further.

So current = V / Z
I=0.35 / 53.14
I= 0.0066 A
I=6.6mA

Power = V * I
P=0.35*0.0066
P= 0.0023 W RMS

Power dissipated by the coil = 23 mW RMS

A total non issue, and this is in a directly connected scenerio without a crossover circuit or speaker cables.

Look, I have no doubt a bad class D amp somewhere has fried a tweeter, but in properly designed amps its not a problem.

Since an ideal amp is a wire with gain from an engineering standpoint then the class D amps, while having many advantages, do have issues with ultra sonic distortion which has not yet been "solved". It seems that "outside of audio frequency" distortion gets a free pass on this site because it can't be heard while no effort is made to qualify different "inside audio frequency" distortion ( 2nd vs 3rd vs 7th etc.) because "all distortion is bad and we are looking for less distortion because it shows better engineering whether it can be heard or not".

I am taking a risk here doing math so if it is wrong please correct. Taking the tweeter example above and assuming the ultra sonic distortion is constant and a 2 volt 3 KHz signal is fed to the tweeter with 4 ohm impedance (1 Watt) by something like the Bench Mark amp and the THD is -100 db that would mean that 0.00002 volts X 0.0001 amps = 0.00000002 Watts (not sure if I have RMS Watts or not) of distortion is being created by the amp which is 115,000 times less distortion than the class D example above.

To say that the class D has 115,000 times more distortion than the best non switching amps does not add anything to the discussion but to say that ultra sonic distortion does not matter at all because it can't be heard seems inconsistent to me. If the closest to a wire with gain is what defines SOTA then class D is not there yet.... but if you factor in cost, size, weight, efficiency and audio band distortion class D makes a very compelling case.

 

[Deleted member 2944]

I still wait for an answer to my old emails to some audio companies. That they do not respond is significant.

Yeah, I would say so.

Dave.

 

[March Audio]

Wouldn't it generate heat in the tweeter putting a stress on it that a class AB amp would not?

No. The calculations above show the power dissipated us utterly trivial.

 

[NielsMayer]

Well, subjectively, my Nord is wonderful. So, maybe my ears aren’t up to snuff....

Based on the pictures, and modulo the incorrect polarity output mistake, if I were to get a pre-assembled unit, i'd pick the Nord over the Apollon. Although the left channel of the Nord posted earlier has power supply too close to input leads, and could trim an inch or two off the output leads. But at least the layout is somewhat rational w/r/t wiring lengths and proximity and wire-crossings.

And any company that took as much time and care to assemble their amps as I did mine would go out of business. Probably took me as long to get the stupid oversized speaker leads into the NC400 outputs "the right way" (so that there's no gap between the wires at the output) than it does for some people to build an entire amp.

 

[maty]

Yeah, I would say so.

Dave.

A good example, PS Audio and class D ICEpower 700ASC.

https://www.audiocircle.com/index.php?topic=156250.msg1673278#msg1673278

-> https://www.audiocircle.com/index.php?topic=156250.msg1673308#msg1673308

15 Mar 2018

-> -> https://www.audiocircle.com/index.php?action=post;msg=1673402;topic=156250.220

Response received: they send the message to the expert.

He must still be meditating the answer. One year and two months.

 

[Samoyed]

I think you missed the inference.

 

[March Audio]

Since an ideal amp is a wire with gain from an engineering standpoint then the class D amps, while having many advantages, do have issues with ultra sonic distortion which has not yet been "solved". It seems that "outside of audio frequency" distortion gets a free pass on this site because it can't be heard while no effort is made to qualify different "inside audio frequency" distortion ( 2nd vs 3rd vs 7th etc.) because "all distortion is bad and we are looking for less distortion because it shows better engineering whether it can be heard or not".

I am taking a risk here doing math so if it is wrong please correct. Taking the tweeter example above and assuming the ultra sonic distortion is constant and a 2 volt 3 KHz signal is fed to the tweeter with 4 ohm impedance (1 Watt) by something like the Bench Mark amp and the THD is -100 db that would mean that 0.00002 volts X 0.0001 amps = 0.00000002 Watts (not sure if I have RMS Watts or not) of distortion is being created by the amp which is 115,000 times less distortion than the class D example above.

To say that the class D has 115,000 times more distortion than the best non switching amps does not add anything to the discussion but to say that ultra sonic distortion does not matter at all because it can't be heard seems inconsistent to me. If the closest to a wire with gain is what defines SOTA then class D is not there yet.... but if you factor in cost, size, weight, efficiency and audio band distortion class D makes a very compelling case.

No this is completely wrong. Class d does not get a free pass. The switching frequency is not just unheard, it's simply not reproduced. It's called putting it into context. You need to come up with a cogent argument as to why it's a problem.

Your example above is completely erroneous. The signal isn't there, there is no distortion.