• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of Hypex NC400 DIY Amp

Buckeye Amps

Major Contributor
Audio Company
Forum Donor
Joined
May 28, 2020
Messages
1,671
Likes
6,571
Thanks for the replies! I have decided to go the NC400 route, with each module using it's own SMPS600, in a Ghent 3 channel case. Can't wait! Just wanted to make sure the power should be near the specified 400w as this DIY route is about the same price as a 3ch NC500 setup (which has more power but lower SINAD).
 

rvsixer

Addicted to Fun and Learning
Joined
Apr 6, 2018
Messages
612
Likes
413
Location
Somewhere at the base of the Rockies....
...(as the original review by amir fell short of the claims and all that could be speculated was the PSU limited it).
Apples to oranges I think:
Hypex spec = One SMPS600 PS driving one NC400 4 ohm load with single channel driven = 400W @ 1% THD
Amir's test = One SMPS600 PS driving two NC400's 4 ohm load with both channels driven = 225W @ 0.004% THD

Different system configuration, different number of channels driven, different THD limits, etc. I would say a single Hypex PS and NC400 will not fall short of their claims (Hypex has a history of publishing accurate numbers, assuming correct implementation by third party/OEM).
 
Last edited:

tparm

Senior Member
Forum Donor
Joined
Nov 5, 2019
Messages
352
Likes
223
Pardon my Class D ignorance, but all of this looks impressive, what is the input sensitivity of these modules? How will they perform with 2V? Let's say in the monoblock 600/400 configuration. Do you need to build in adjustable gain and therefore degrade performance at 2V? Intrigued by this technology compared to a high performing AB amp like a monolith 2X vs the ICE module architecture in the $500 XTZ Edge. I'm handy but not sure I have the time for DIY.

Thanks.

@Amir, what are the chances you have an XTZ or Monolith amp incoming?
 
Last edited:

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
3,745
Likes
3,031
Pardon my Class D ignorance, but all of this looks impressive, what is the input sensitivity of these modules? How will they perform with 2V? Let's say in the monoblock 600/400 configuration. Do you need to build in adjustable gain and therefore degrade performance at 2V? Intrigued by this technology compared to a high performing AB amp like a monolith 2X vs the ICE module architecture in the $500 XTC. I'm handy but not sure I have the time for DIY.

Thanks.

@Amir, what are the chances you have an XTC or Monolith amp incoming?
2V will give you the rated power output (400W into 4R nominal). Details in the datasheet. I think there's an application note giving details of how to change the gain, but you'd need to be able to change a surface mount resistor to do this.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,028
Likes
4,035
Location
Amsterdam, The Netherlands
Do you need to build in adjustable gain and therefore degrade performance at 2V?

Why would adjustable gain degrade performance?
 

tparm

Senior Member
Forum Donor
Joined
Nov 5, 2019
Messages
352
Likes
223
Why would adjustable gain degrade performance?
Simply looking at Amir’s reviews, it appears the two things that affect measurements are a reduction in voltage and an increase in gain.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,028
Likes
4,035
Location
Amsterdam, The Netherlands
Simply looking at Amir’s reviews, it appears the two things that affect measurements are a reduction in voltage and an increase in gain.

Reducing gain by increasing negative feedback might possibly increase performance. Increased input voltage level improves SNR, but only by shifting the reference point.
 

rodtor

Member
Forum Donor
Joined
Aug 21, 2019
Messages
97
Likes
180
Location
Toronto
Reducing gain by increasing negative feedback might possibly increase performance. Increased input voltage level improves SNR, but only by shifting the reference point.
Yes, Amir's tests for the Benchmark and Purifi amps report higher SINAD when the amplifier's gain is at the lowest level. So variable gain could help, as long as one can find a DAC or preamp that can deliver enough voltage to allow the amp to be used at that level.
 

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
Yes, Amir's tests for the Benchmark and Purifi amps report higher SINAD when the amplifier's gain is at the lowest level. So variable gain could help, as long as one can find a DAC or preamp that can deliver enough voltage to allow the amp to be used at that level.
Optimizing system gain structure does minimize net noise (in this discussion, minimizing the gain used in the power amp). One should take care not to do this to the point of increasing demands on a source component to swing more output voltage (and current) than is optimum for it - there can be a trade-off...
 

ehabheikal

Senior Member
Joined
May 22, 2020
Messages
442
Likes
161
This is a review and detailed measurements of a DIY power amplifier based on Hypex NC400 amplifier modules. It was built and is on kind loan from a member. I asked him to list the parts and this is what he supplied:
So putting aside labor, it is about USD $1,100. From outside, the unit looks very professional, sans branding:


The unit came with balanced inputs and speaker outputs and that is how I tested it.

The amplifier as noted is based on Hypex NC400 modules. If you don't know much about it, the chief designer Bruno Putzeys was the person that put high performance class-D amplifiers on the map, producing amplifiers that not only matched performance of classic AB amplifiers but bested them. I have been meaning to measure one for a long time and this is the first time I have had the opportunity.

Class-D amplifiers have the advantage of high-efficiency as the transistors are operated in switching mode: either full on or full off. They don't use power when fully off and when fully on, have the least resistance and hence lowest power consumption. The audio signal is used to modulated these pulses. Before outputting the signal though, the pulses need to be filtered as the thing will act like a little FM or TV station spitting out tons of high energy and high frequency noise. The filtering must be passive (e.g. like the one in your speakers). This can make class-D amplifiers load dependent.

Class-D amplifiers are quite a bit more complex than traditional amplifiers due to operating at high speeds (well above audio band). Creating a reliable one is also challenging. This is the motivation behind creating modules around them so that the design can be done once and reused without chance of screwing things up. Countless audio companies as a result buy these modules and use them in their products. These modules are also available to DIY people in both kit and individual form as is the case here.

I opened the unit and was pleasantly surprised by the clean layout and nice work the owner had done to put it together (click on image for larger size):
View attachment 19461

The two round PC boards are the NC 400 modules which as the name produces, is able to generate 400 watts of power. They are powered by a switching power supply also from Hypex (SMPS600). As you will see later, the power supply limits the amount of power these two hungry amps can generate.

I was so pleased to see proper and safe grounding lug in green and sanding off the anodizing finish on top right and bottom of the chassis and likewise the mating position on the lid to make sure it is grounded. Both of these are missed in some products such as created by Schiit.

Personally I would have given up the aesthetics and put the speaker plugs close to their modules to make a symmetrical system. The wiring colors for the speakers follow mains voltage convention in US with black being "hot" and "white" not (neutral). In DC wiring though, we like to use red for positive and white/black for negative. Alternatively white can be positive and black negative. Fortunately that is a visual nuisance as they are wired correctly to the amp and the terminal.

The only major concern I have is not regarding the assembly but the design of the switchmode power supply by Hypex. A number of power transistors are mounted to the bottom clear aluminum heatsink which is nothing but bent piece of metal. It gets quite warm and is sitting within a millimeter or two of the large electrolytic capacitors below it. The two caps were getting quite warm even in limited use and will surely result in them drying out prematurely and failing. This case is plenty large so personally I would attempt to straighten the bent aluminum so that it is more vertical.

Just casually feeling the heat generated by the amps and power supply, the later was much warmer. The case on the other hand, has vents for the amp modules and not the power supply. If the lid can be reversed, that may be something I would do to keep the power supply cooler.

Let's get into measurements and see if we finally have a performant power amplifier as our previous units have failed us and failed us badly.

Measurements
As usual, let's start with our dashboard with 4 ohm load and 5 watts of power with AES-17 40 kHz filter:

View attachment 19464

Wow! Finally we have a power amplifier that matches the performance of mid-tier DACs and outperforming previously tested amplifiers by tens of dBs. The third harmonic is actually well below -110 dB. And if I turn one channel off, we get near perfect performance:

View attachment 19465

The second harmonic is now down to -130 dB!!! Pretty remarkable.

Of course that is at 5 watts. Let's sweep the input level and see the performance across the range:
View attachment 19466

Performance actually improves up to about 30 to 50 watts and then starts to climb some. At near full clipping, we get healthy amount of power (225 watts both channels driven) with 0.004% distortion. It completely blows out of water the previously tested amps in both power and distortion/noise.

Using near max power, we get very nice signal to noise ratio:
View attachment 19467

Frequency response is also far better than digital amps we have tested:
View attachment 19468

I need to develop more complex loads to better determine the response into real speakers.

Broadband FFT shows what exists above audio band (post AES-17 filtering):

View attachment 19469

We clearly see the switching frequency of the amplifier (around 460 kHz) despite the filtering in the amplifier and the AES-17. We can see the same in our waveform at very, very low power levels:

View attachment 19470

Harmonic distortion versus frequency is exceptionally well behaved and good:
View attachment 19471

Notice the NC400 (in blue) relative to the abomination that was Behringer A500 (in dashed red). Note that this is a wideband test so THD+N is higher than the dashboard (it would not otherwise include all the harmonic distortions at 20 kHz).

Conclusions
Folks, welcome to adult amplifier world! :) Previous amplifiers that we tested were awful in measured performance and I was starting to think good amps are a thing of past. Fortunately Hypex NC400 amplifier modules show that very well engineered amplifiers exist in class-D configuration no less! Not only we have stellar performance here, but also get to have high efficiency and cool running amplification to boot. With nearly quarter of kilowatt of power available, there should be no issue driving even inefficient speakers to very loud levels.

For better longevity I would use a different and possibly more powerful power supply to get even more performance.

As it is, I can wholeheartedly recommend hypex NC 400 modules for power amplification. Yes, the pink panther is happy!

This seems to be a very easy DIY project although I caution you to NOT engage in it if you are not comfortable and knowledgeable about mains wiring and voltages. You can easily get yourself shocked or killed. So buy an already made one if you are not able to build one safely.

I am pleased to have a reference for testing a few more power amplifiers that have been piling up in my lab. :)

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

If you like this review, or even if you don't but wish for me to escape the rain in Seattle and go somewhere sunny, please consider donating 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).



I am comparing this to the 1et400a measurements especially the graph i attached below comparing effect of frequency on distortion at different power levels and comparing it to the nc400 i see that only one graph was done at 5W, is the nc400 immune to increase in noise, ie did you test it and thus did not need other graphs or was this not tested as deeply only and the NC400 could have increasing distortion in high frequencies as the power is raised?
 

Attachments

  • Purifi 1ET400A Class-d Amplifier Module THD vs Frequency vs Power Audio Measurements.png
    Purifi 1ET400A Class-d Amplifier Module THD vs Frequency vs Power Audio Measurements.png
    24.5 KB · Views: 152

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,028
Likes
4,035
Location
Amsterdam, The Netherlands
I am comparing this to the 1et400a measurements especially the graph i attached below comparing effect of frequency on distortion at different power levels and comparing it to the nc400 i see that only one graph was done at 5W, is the nc400 immune to increase in noise, ie did you test it and thus did not need other graphs or was this not tested as deeply only and the NC400 could have increasing distortion in high frequencies as the power is raised?

From Hypex
Screenshot from 2020-07-22 08-21-08.png
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,028
Likes
4,035
Location
Amsterdam, The Netherlands
Seems like nc400 is affected by increasing distortion at higher frequencies it from the graph on the left.

Do you think 0.02% at 10 kHz is audible?
 

ehabheikal

Senior Member
Joined
May 22, 2020
Messages
442
Likes
161
Total harmonic distortion
One question, when we measure thd at for example 10K are we measuring the THD produced by the 10K or the distortion at 10K that can be
It's not.

If 0.02% is not audible would i feel the difference between an nc252 or so at sinad of 94 or so and an nc400 with sinad of 104?
I would rather not pay more if i will not hear the difference ( and i am about 50 old so my hearing above 12k is not that great )
 

tmtomh

Major Contributor
Forum Donor
Joined
Aug 14, 2018
Messages
2,728
Likes
7,989
Total harmonic distortion
One question, when we measure thd at for example 10K are we measuring the THD produced by the 10K or the distortion at 10K that can be


If 0.02% is not audible would i feel the difference between an nc252 or so at sinad of 94 or so and an nc400 with sinad of 104?
I would rather not pay more if i will not hear the difference ( and i am about 50 old so my hearing above 12k is not that great )

Just keep in mind that one also needs to consider other factors - gain, intermodulation distortion, perhaps damping factor, and I'm sure others I'm forgetting about. But yes, while one cannot say for certain, I would say there's a good chance a 252 and a 400 are going to sound the same (when driven to volumes that are within the power handling capacity of each one, given that the 252 is less powerful).
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,028
Likes
4,035
Location
Amsterdam, The Netherlands
One question, when we measure thd at for example 10K are we measuring the THD produced by the 10K or the distortion at 10K that can be

We are measuring the distortion produced by a 10 k tone, so the first distortion product is at 20k.
 
Top Bottom