Hypex Nilai500DIY Amplifier Review

[Scott Lanterman]

Value is always the issue in my book. It is complex, what you value is different than what I value. The equation is a standard value engineering equation and it is used by companies to help define there value proposition, but it might not be obvious because cost and performance are user defined and often complex.

If you want a high performance amp then you define what that "high performance" means to you. Are the low noise and distortion numbers at 500 watts that Nilai has divided by the $890 ea. purchase price larger than some alternative? Your actual choice is the largest V when doing an all things considered deliberation before you put down your money. The costs might be the price of the product, but maybe you are ok with your "cost" of a DIY build, and using the product without it being tested, which is a risk or another cost to you. If you value an electronic component being tested that cost translates into a cost you are willing to pay, but as the consumer you decide. I trust Hypex being a solid brand and if something went wrong I think they will respond so I decided the untested path was worth the risk and I was ok with it since I value the 500 clean watts in a small box more.

The businesses I have worked with often were not aware of this equation, you get it some economics classes that talk about value engineering. I find it useful but not everybody likes the equation. I like the equation so I use it.

To say the greatest value in the cheapest option I think confuses value with price. If you must have instrument grade tools than you are probably not looking at the cheapest. If you value VU meters than you might not want this amp at all. Value is just not that simple and I offer V = P/C with some caution because it can be used as a shortcut as opposed to helping a person clarify that the price does not relate to performance but is its own independent variable. Value is getting the performance you want at the lowest total cost. If the cheapest meets your performance desires than you have a big value that you are considering. It's not so simple in my view.

 

[MCH]

The Nilai five warranty from an established leader in the field is part of the value calculation for me as a buyer of Nilai Mono 500 diy kit.

I don't know if the 5 years warranty on the nilai is a launch offer or if it is supposed to be a permanent offer, because the general warranty terms in their website state the mandatory 2 years. Maybe @Hypexsales can comment. Sorry if already discussed!
5 years warranty is indeed very good value nowadays.

 

[Scott Lanterman]

5 years on a brand new product with very high specs and the most power! That is quite a bit with something measuring as good as anything but with more power. The actual boards seem better than average and the power module has a polished heat sink because this thing does get warm.

 

[Whatsthephuss]

HI there.

Hi amirm, thanks for the response and again thank you for conducting all of these informative tests.

As noted, the THD+N vs frequency is over 45 kHz bandwidth and with class D amps, this reduces their measured performance fair bit. The 45 kHz bandwidth came from much testing and arguments. If set much higher than this, it completely screws up class D amp due to increased noise above audible band. To set it lower as Hypex/Purifi do, would be wrong as well. So I have settled at 45 kHz where we can see the differentiation with class AB designs well.

To say that setting the bandwidth lower is “wrong,” is an opinion. Just like whether or not an amp is “good” or “bad.” That leads me to ask, wrong for what purpose? You’ve stated that the purpose of your measurements is for comparison. i still think it would be valuable information for comparison to measure an amp both with a 22kHz bandwidth and also a much higher bandwidth like has always been done. No one said you had to pick one, and the only way to make everyone happy is to give both sets of data. This would be very useful to compare how an amp would perform in a commercial environment with low pass filters, as opposed to the home environment where I assume there aren’t historically a lot of hi cut filters…

Choosing to continue with 45kHz bandwidth seems to be setting a new standard that makes it harder to compare to manufacturers data sheets and provides less than ideal data for how a product performs in real world conditions. Expanding the measurements seems to be a logical evolution, so I hope you consider this a suggestion for improvement more than a criticism.

 

[Sokel]

5 years on a brand new product with very high specs and the most power! That is quite a bit with something measuring as good as anything but with more power. The actual boards seem better than average and the power module has a polished heat sink because this thing does get warm.

5 years warranty seems more like a trend at this time,it's not only Nilai as I see:

Order the SoundImpress HY122-2CH Stereo Amplifier | Ncore® | Powered by Hypex

Buy the HY122-2CH online at SoundImports ✓ Same day shipping ✓ Low shipping rates

www.soundimports.eu

Order the SoundImpress PU400-2CH Stereo Amplifier | Eigentakt | Powered by Purifi

Buy the PU400-2CH online at SoundImports ✓ Same day shipping ✓ Low shipping rates

www.soundimports.eu

Order the SoundImpress ICE1200-2CH Stereo Amplifier | Powered by ICEpower

Buy the ICE1200-2CH online at SoundImports ✓ Same day shipping ✓ Low shipping rates

www.soundimports.eu

 

[HarmonicTHD]

I'm new to this forum: does value even properly come into play for product reviews? I don't see how it can. After all, the greatest value in amplifiers is found in the cheapest ones. They take you from silence to music, at the least cost. And probably satisfy the needs of a majority of music lovers (with their smart phones and earbuds). At the other end of the spectrum are folks who spend $1,000/foot for speaker cable because they are not limited by price and are easily persuaded.

Personally, for a review I just want to know how the device measures compared to other products, and what effect those differences might have on how I hear music played through them. I can draw my own (nonlinear) efficiency curves.

Correct. Value is something very individual and personal. It is futile to argue about it. We are therefore best left with the facts which are provided here and then decide based on our own value perception.

As a side note, that’s why I also find these price discussions quite ludicrous as what is cheap to one person might appear expensive to another.

 

[Shadders]

Hi amirm, thanks for the response and again thank you for conducting all of these informative tests.

To say that setting the bandwidth lower is “wrong,” is an opinion. Just like whether or not an amp is “good” or “bad.” That leads me to ask, wrong for what purpose? You’ve stated that the purpose of your measurements is for comparison. i still think it would be valuable information for comparison to measure an amp both with a 22kHz bandwidth and also a much higher bandwidth like has always been done. No one said you had to pick one, and the only way to make everyone happy is to give both sets of data. This would be very useful to compare how an amp would perform in a commercial environment with low pass filters, as opposed to the home environment where I assume there aren’t historically a lot of hi cut filters…

Choosing to continue with 45kHz bandwidth seems to be setting a new standard that makes it harder to compare to manufacturers data sheets and provides less than ideal data for how a product performs in real world conditions. Expanding the measurements seems to be a logical evolution, so I hope you consider this a suggestion for improvement more than a criticism.

Hi,
Here is a link to Hifi News Nad M23 power amplifier that uses the Purifi module implemented in their own Nad design.

NAD M23 Class D Power Amp Lab Report

Lab Report This, to all intents and purposes, is the power amp stage from NAD's fully-fledged digital/analogue BluOS-equipped M33 integrated [HFN Aug '20], the Purifi ET400A Class D module delivering the raw specification of the more affordable C 298 power amp [HFN Oct '21] but with 'Masters...

www.hifinews.com

Hifi News use their own proprietary equipment which uses a measurement bandwidth of 80kHz, since the Audio Precision has problems with class D at that measurement bandwidth.

Using a 45kHz filter gives you more of a picture of how the amplifier performs harmonically/noise past 22kHz. It is up to you how you assess the figures, but the 45kHz bandwidth does allow you to compare class D with class AB with a more even playing field.

My preference is to keep the 45kHz for the above mentioned reasons, since the manufacturer will provide the 22kHz bandwidth test results anyway.

Regards,
Shadders.

 

[Extreme_Boky]

Hi,
Here is a link to Hifi News Nad M23 power amplifier that uses the Purifi module implemented in their own Nad design.

NAD M23 Class D Power Amp Lab Report

Lab Report This, to all intents and purposes, is the power amp stage from NAD's fully-fledged digital/analogue BluOS-equipped M33 integrated [HFN Aug '20], the Purifi ET400A Class D module delivering the raw specification of the more affordable C 298 power amp [HFN Oct '21] but with 'Masters...

www.hifinews.com

Hifi News use their own proprietary equipment which uses a measurement bandwidth of 80kHz, since the Audio Precision has problems with class D at that measurement bandwidth.

Using a 45kHz filter gives you more of a picture of how the amplifier performs harmonically/noise past 22kHz. It is up to you how you assess the figures, but the 45kHz bandwidth does allow you to compare class D with class AB with a more even playing field.

My preference is to keep the 45kHz for the above mentioned reasons, since the manufacturer will provide the 22kHz bandwidth test results anyway.

Regards,
Shadders.

To make sure that AP readings look good, just limit the bandwidth to whatever you want... as long as the specifications look nice, who cares how class D sounds... amps are light, cheap, have 0.5kW of power, and great specifications... it's so good that you won't even have to listen to them (amps).

And, if there's a creep in THD + N... just apply more feedback - that will fix the issue, so that AP is happy peppy again

 

[tmtomh]

To make sure that AP readings look good, just limit the bandwidth to whatever you want... as long as the specifications look nice, who cares how class D sounds... amps are light, cheap, have 0.5kW of power, and great specifications... it's so good that you won't even have to listen to them (amps).

And, if there's a creep in THD + N... just apply more feedback - that will fix the issue, so that AP is happy peppy again

That's a lot of FUD in a short post. Very efficient.

 

[Julf]

Using a 45kHz filter gives you more of a picture of how the amplifier performs harmonically/noise past 22kHz. It is up to you how you assess the figures, but the 45kHz bandwidth does allow you to compare class D with class AB with a more even playing field.

Yes, it would give a more even playing field in terms of a pretty insignificant aspect.

 

[Shadders]

To make sure that AP readings look good, just limit the bandwidth to whatever you want... as long as the specifications look nice, who cares how class D sounds... amps are light, cheap, have 0.5kW of power, and great specifications... it's so good that you won't even have to listen to them (amps).

And, if there's a creep in THD + N... just apply more feedback - that will fix the issue, so that AP is happy peppy again

Hi,
Here is a link to the Philips UcD amplifier from a third party document :

http://www.handsontec.com/pdf_files/UM10155_2_Discrete_ClassD_Amp.pdf

The patent referenced in section 7.3 is by Bruno Putzey.
Isn't what you describe already implemented by Hypex/Purifi, the high level of feedback to reduce the THD, resulting in a reduced overall gain ?

For the measurement bandwidth, it is what it is. Past 10kHz you probably will not hear the THD, but on paper, the use of 22kHz could make the class D look better than a class AB amplifier measured using 80kHz (in general).

Regards,
Shadders.

 

[pma]

Yes, it would give a more even playing field in terms of a pretty insignificant aspect.


For the measurement bandwidth, it is what it is. Past 10kHz you probably will not hear the THD, but on paper, the use of 22kHz could make the class D look better than a class AB amplifier measured using 80kHz (in general).

Regards,
Shadders.

Just posted a thread on this:

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters It has been quite long lasting discussion here on measuring bandwidth for THD+N plots vs. amplifier power, measured at various frequencies. Especially for class D amplifiers, THD+N results get rapidly worse, at...

www.audiosciencereview.com

It is about engineering consistency, not about cherry-picking parameters that suit to this or that design concept. It is funny how technical questions easily become emotional when it comes to “preferences” of any kind.

 

[Shadders]

Just posted a thread on this:

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters It has been quite long lasting discussion here on measuring bandwidth for THD+N plots vs. amplifier power, measured at various frequencies. Especially for class D amplifiers, THD+N results get rapidly worse, at...

www.audiosciencereview.com

It is about engineering consistency, not about cherry-picking parameters that suit to this or that design concept. It is funny how technical questions easily become emotional when it comes to “preferences” of any kind.

Hi,
Agree on the consistency approach. The problem is that only Hifi News uses a measuring bandwidth of 80kHz for both class AB and class D, since they do not use Audio Precision.

As such, using 45kHz for the measurement bandwidth by this site, somewhat addresses the inconsistency, and hence my preference for it.

Regards,
Shadders.

 

[tmtomh]

Just posted a thread on this:

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters

Effect of measuring bandwidth on class D amplifier measured THD and THD+N parameters It has been quite long lasting discussion here on measuring bandwidth for THD+N plots vs. amplifier power, measured at various frequencies. Especially for class D amplifiers, THD+N results get rapidly worse, at...

www.audiosciencereview.com

It is about engineering consistency, not about cherry-picking parameters that suit to this or that design concept. It is funny how technical questions easily become emotional when it comes to “preferences” of any kind.

Agree on the importance of consistency. But I think you are not factoring in an important aspect of the discussion: the additional frequencies in an 80kHz bandwidth signal compared to a 45kHz bandwidth signal are irrelevant to the actual performance of an audio amp. Testing to 80kHz or 100kHz is no more or less proper or arbitrary conceptually than testing to 45kHz. Class AB amps don't need to have bandwidth to 100kHz either. The fact that it's been possible to test them with high-bandwidth signals without unduly contributing to their measured THD has no special meaning. It's simply been a tradition that got codified into a standard, and from that the ultrasonic bandwidth capability of an amp became, for many people, a marker of engineering quality. There's no essential reason for that association - it's an "emotional" attachment some folks have to bandwidth.

In that vein, if we want to look for the most "emotional" comments here about this issue, we can look no further than @restorer-john , who regularly gets into highly rhetorical froths about how outrageous, deceptive, and appalling it is that we are letting standards slide by accepting lower-bandwidth signal testing of Class D amps. You voice a less emotional version of the same idea here and in countless other comments.

The point is that testing to 45kHz does indeed suit the "design concept" of Class D. But that's an appropriate choice that properly reveals the performance and capabiiities of these units. It's not "cherry picking" unless it obscures or hides performance flaws that actually correspond to the unit's performance for its intended application.

 

[pma]

"Typically"? Is it specified in any standard, or is it just a convention by a specific manufacturer of measurement gear?

You asked the question on standards, so yes, there is a standard IEC 60268-3

IEC 60268-3:2018 | IEC Webstore

IEC 60268-3:2018 Standard | Sound system equipment - Part 3: Amplifiers

webstore.iec.ch

Sound system equipment - Part 3: Amplifiers

Link to the preview: https://webstore.iec.ch/preview/info_iec60268-3{ed5.0}b.pdf

and this is the intro to THD measurement

So, it is about "highest significant harmonic". And in case of these Ncore and similar, highest significant harmonic is definitely not limited to 3rd only. Sometimes, 5th may be highest. Limiting BW to 22kHz or so is done only for manufacturer's marketing benefit.

 

[Rick Sykora]

So, while not disputing the technical differences, quoting a THD spec with 20 Hz to 20 kHz bandwidth is NOT unique to Class D. A quick check of Class AB amp manufacturers (ATI, Emotiva, Rotel and others) are all claiming THD specs at the same 20 to 20 kHz bandwidth).

 

[Skeeter]

This is a review and detailed measurements of the Hypex Nilai500DIY class D amplifier module with PS500DIY power supply. Parts were provided by the company and member @Rick Sykora was kind enough to assemble it:
View attachment 262137

As indicated, the single channel amplifier is on the left and the power supply on the right. The amplifier has built-in preamp providing three levels of gain (12, 22 and 28 dB). A jumper selects the appropriate level. At 12 dB, you need about 12 volts to reach full power which requires a pro audio interface or a pre-amp with that level of output. The sole input in this configuration is balanced.

In use the amplifier was exceptionally solid even when pushed hard. It kept going and stayed very cool. It is rated to 2 ohm (see measurements below).

Edit: in the original review I had the top open so I could change the gain settings. Alas, it turns out that doing so allowed some noise pick up around 10 kHz. I re-tested with the lid closed and that issue went away. As a results, we have slightly better performance especially in high-gain.

Hypex Nilai500DIY Measurements
Let's start with lowest gain setting:
View attachment 262735

That is superb performance with distortion at or below -140 dB! Here is the response at higher grain settings:
View attachment 262139

View attachment 262736

With all thee gain settings, the amp lands in our top 20 best amplifiers ever reviewed based on combined noise and distortion rating (SINAD):
View attachment 262737

Noise performance is excellent but just shy of state of the art:
View attachment 262738

But gets there at full power:
View attachment 262143

As we expect from Hypex, frequency response is load impedance independent:
View attachment 262144

Multitone performance is excellent:

View attachment 262739

View attachment 262740

Amplifier is stable on power up:
View attachment 262741

Let's look at available power at 4 ohm:
View attachment 262149

Looks like we no longer have to decided between superbly low noise and distortion and available power! State of the art performance is once again achieved in low gain while still producing lots of power (342 watts). Allowing distortion to climb to 1% naturally gives us more power (and matches company spec):
View attachment 262150

We still have healthy power even with 8 ohm load:
View attachment 262151

Since the amp is rated at 2 ohm, I measured that as well:
View attachment 262152
View attachment 262153

There is a bit of penalty but still great performance. Seems power supply is limiting power so you don't get more juice.

I was a bit surprised at the ballooning of the distortion at higher frequencies:
View attachment 262154

EDIT:
There is more of a noticeable pop on power off than power on:
View attachment 262742

As requested, here is the switching response:
View attachment 262743


Conclusions
It is clear Hypex is taking competition in class D amplifiers seriously responding with a DIY friend amplifier solution that has a ton of power and wonderful performance. While we could nitpick here and there, but in the larger context of the amplifiers out there, this is a superb offering of great performance and lots and lots of power.

FYI I don't have pricing for the kit. If someone has it, please post. I see online ads saying around $900 to $1,100 for a single channel.

I am happy to add Hypex Nilai500DIY amplifier and its companion PS500DIY power supply to my recommended list.

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

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

By any measure, the HF distortion on this amplifier is very high. I’d like to see a 19+20 kHz IMD test.

 

[pma]

By any measure, the HF distortion on this amplifier is very high. I’d like to see a 19+20 kHz IMD test.

Yes it is, and it is a system principle attribute. You, and anyone interested, may check this soundstage set of measurements of NAD C 298 (Purifi)

SoundStageNetwork.com | SoundStage.com - NAD C 298 Stereo Amplifier Measurements

www.soundstagenetwork.com

Especially THD+N plots

that are measured properly with appropriate measuring bandwidth.

 

[tmtomh]

Limiting BW to 22kHz or so is done only for manufacturer's marketing benefit.

You keep making this assertion, and it keeps being somewhere between misleading and flat-out false.

If one were to publish THD figures for the Audiophonics Purifi amp Amir tested using a signal of greater bandwidth, that would increase the overall THD, calculated as an average. Amir reported that distortion components - not SINAD, but just the distortion since that's what you're obsessed with - of that unit were down around -125dB. That's 0.00006%. So let's say a wider-bandwidth signal, which would produce higher distortion at ultrasonic frequencies, would increase the THD by a factor of 5x. That would push the distortion up to a whopping 0.0003%, or -90.5dB.

But of course, that's not what you are concerned about. Instead, you keep saying these amps have "poor high frequency linearity" as if they start to distort in significant ways within the audible range. They don't, and you know it. And we know they don't because that allegedly "poor high frequency linearity" produces lower distortion between 5-20kHz than the amplifier module that you hold up as superior.

In fact, a THD rating of a Purifi-based amp and a Class AB amp, both fed a high-bandwidth signal, can be just as deceiving if not moreso, because the amps might have similar THD but the AB's distortion figure will come from a higher overall level of distortion throughout the audible range, while the Purifi's will come from a disproportionate amount of distortion in the ultrasonic range combined with vanishingly low distortion in the audible range. That's an important difference - in the 2-5kHz range where our hearing is most sensitive, the distortion of the Purifi amp is *25-33dB lower* than your "superior" UcD180. But that's not a difference you care about.

It's your prerogative not to care about that difference, but by the same token it's my prerogative to say that you're being silly and that you're intentionally ignoring and omitting crucial information to try to make your point - and that you're a hypocrite for ignoring key engineering factors while claiming the high ground of engineering quality and ethics.

Your argument is fatally flawed.

 

[Skeeter]

The HF distortion profile is more like that of a zero global feed back amp. Simple question: why is this?

A two tone IMD test will expose any inherent HF non-linearity and does not require bandwidths of >21 kHz since the intermod components are folded down into the audio band.