Review and Measurements of Hypex NC400 DIY Amp

[amirm]

Curiously, on AP's site they say this:

"Note that only the 2700 Series analyzers accept optional AES-17 filters. The APx Series and ATS-2 don’t need them—their standard low-pass filters are implemented digitally and already have a brick-wall roll-off that meets the AES-17 specifications."

There are actually two types of AES-17 filters for Audio Precision analyzers. One is internal and is designed to give you the standardized bandwidth for general work. That filter is NOT sufficient for class-D amplifiers, nor delta-sigma DACs with a lot of noise shaping. These products have such high slew rate that they screw up the front-end of the analyzer, resulting in erroneous THD measurements and may even confuse the auto-scalar. For these measurements, you must use the external, passive AES-17 filter as I have been using.

The part you quoted above is for the internal AES-17 filter. I have not been using that as I am using the ADC filter in AP to limit bandwidth.

 

[Jimster480]

Wow, the performance is pretty awesome.
I caught the panther and it was a nice Segway into the review

 

[restorer-john]

The part you quoted above is for the internal AES-17 filter. I have not been using that as I am using the ADC filter in AP to limit bandwidth.

That's what I thought, it was the curious wording they used.

 

[March Audio]

Here is that:

500 Hz:

View attachment 19590

There is almost no phase differential with zero crossing almost the same.

20 kHz:
View attachment 19589

There is 2.6 microseconds of phase shift now. If my math is right, that is about 20 degrees of phase shift.

Running AP without AES filter makes it struggle to measure phase but going along with it anyway, we get:

View attachment 19593

Thanks Amir, that's great. So seems to confirm the ADC filter is adding additional phase shift.

 

[dragonspit4]

Power amp is used for speakers right?
How much power can this unit output?

 

[vitalii427]

Power amp is used for speakers right?
How much power can this unit output?

Did you read the first post?
1. Yes
2. From measurements: two channels 225W each into 4Ohm load

 

[amirm]

Power amp is used for speakers right?
How much power can this unit output?

It is in the review:

 

[MZKM]

Power amp is used for speakers right?
How much power can this unit output?

This unit is slightly customized I believe, and is limited by the power supply (the 600 model he bought it for the monoblock unit, the 1200 model is for stereo use), but the super simple DIY kit that Hypex (monoblocks for <$1600/pair I believe) is a might more powerful, Datasheet with specs and graphs.

If you want an assembled unit, James Romeyn sells a stereo unit for $1500 I believe and a monoblock pair for $1900 I believe.

 

[MZKM]

After watching the tutorial on the same page, it looks a very easy and ready to put together DIY kit, indeed.
Great lead @amirm , amazing !

Keep in mind the current kits don’t even require soldering (as stated in the linked video’s description), it’s basically plug/screw and play.

 

[gvl]

Keep in mind the current kits don’t even require soldering (as stated in the linked video’s description), it’s basically plug/screw and play.

Dang, I was a good boy all year and Santa just dropped off a new soldering station. Oh well.

I'm keen to build a near field desktop friendly SOTA class-D amp, is there anything similar to the Hypex modules out there but with less power and for less $$$?

 

[edd9000]

I'm keen to build a near field desktop friendly SOTA class-D amp, is there anything similar to the Hypex modules out there but with less power and for less $$$?

Icepower modules turn up on ebay. Minidsp sell a plate amp for active speakers with 2x125w https://www.minidsp.com/products/plate-amplifiers/pwr-ice125

 

[gvl]

Icepower modules turn up on ebay. Minidsp sell a plate amp for active speakers with 2x125w https://www.minidsp.com/products/plate-amplifiers/pwr-ice125

Thanks, looks interesting. Do they sell to general public? Can't seem to be able to see pricing or add to cart, etc.

 

[jmmaher]

If you want an assembled unit, James Romeyn sells a stereo unit for $1500 I believe and a monoblock pair for $1900 I believe.

I purchased one a few years ago from James, he does an excellent job.

 

[andreasmaaan]

Thanks, looks interesting. Do they sell to general public? Can't seem to be able to see pricing or add to cart, etc.

ICEpower modules are sold direct to the public. I’m on the phone atm but I have the British distributor bookmarked on my laptop. You should be able to find a local distributor through the ICEpower website. I’ve also seen lots of modules on eBay, although given how low they’re priced I’m not sure of their authenticity.

 

[badboygolf16v]

Dang, I was a good boy all year and Santa just dropped off a new soldering station. Oh well.

I'm keen to build a near field desktop friendly SOTA class-D amp, is there anything similar to the Hypex modules out there but with less power and for less $$$?

If Class D is not essential, maybe consider a Neurochrome chip amp? Not sure of the $$$...

 

[restorer-john]

So here's the truth on the actual long term capability of the amplifier setup Amir tested (2xNC400+1xSMPS600) according to Hypex themselves:

At least they are honest enough to say their PSU can't deliver its rated power continuously. That's a first.

Note these are 1KHz results too, not 20-20KHz.

So, we have a continuous 75Watt per channel amplifier @1KHz with a THD <1% that has a decent headroom to be able to deliver peaks in excess of 220W (THD 1%) for a minute and a half. Whoopee.

That's hardly groundbreaking, exceptional, or even remotely exciting, especially at the prices of those little modules.

In 1977, (yes, 41 years ago) we had the Sony TAN-88, a PWM (Class D) power amplifier that delivered 160W per channel continuously, both channels driven from 20-20KHz at <0.5%THD (at full power!), driven by a SMPS supply all housed in an 80mm high, 11kg box.

If you wanted ball-tearing dynamic headroom, the Proton D-1200 showed that in spades in 1986 (yes, 32 years ago) at a low $599.

"The D1200's output clipped at 155 watts per channel into 8 ohms, 175 watts into 4 ohms, and 350 watts into 2 ohms. We measured the dynamic power response of the amplifier with bursts of 20, 100, and 200 milliseconds repeated twice per second. The 8-ohm output with a 20-millisecond burst was 593 watts, for a dynamic-headroom measurement of 7.73 dB. The power increased to 1,187 and 1,800 watts into impedances of 4 and 2 ohms, respectively. During the longer bursts the available power was lower, of course, but even at 200 milliseconds the 8-, 4-, and 2-ohm power readings were 478, 700, and 1,055 watts.

It appeared that only after the signal duration exceeded 1 or 2 seconds did the power output fall to its steady-state value.
The 1,000-Hz harmonic distortion was between 0.0025 and 0.005 percent from 1 to 130 watts into 8 ohms. The 2- and 4-ohm readings were in the range of 0.008 to 0.003 percent from 1 watt to their clipping levels. With both channels driving 8-ohm loads, the distortion was typically about 0.003 percent from 50 to 2,000 Hz at power outputs from 10 to 100 watts, and the maximum reading was 0.024 percent at 20,000 Hz and 100 watts."

It seems that the Emperor has no new clothes, just a really ugly shirt.

 

[amirm]

That's hardly groundbreaking, exceptional, or even remotely exciting, especially at the prices of those little modules.

That is limitation of the power supply though, not the amplifier modules.

 

[Sancus]

Assuming you use the SMPS1200 instead, then all those numbers are doubled. As far as I can tell, there's not much price difference between the two power supply modules. I do agree it is absurd to call power supplies 600W and 1200W when they max out at 150W and 300W continuous respectively.

Since it does seem to be a thermal limitation, perhaps you could increase continuous output even further by adding heatsinking or other cooling solutions. Which is not unreasonable for a DIY solution. It does seem like some actual manufacturers even use one SMPS1200 per channel with their NC500 implementations.

 

[restorer-john]

That is limitation of the power supply though, not the amplifier modules.

Not quite true Amir. The PSU is inadequate, that is a given, but the modules themselves cannot dissipate the heat for continuous operation at even their rated maximum output. Their continuous rating is significantly de-rated accordingly, and that is with adequate heatsinking.

So we have power supplies and their attached amplifiers which are both playing funny numbers. What a combination.

Let's not make excuses for 600W supplies that are really only 150W supplies and power amplifiers that are claiming absurdly high output powers at high THD levels, only to really be medium powered when (properly) continuously rated. We have decades of stringently measured amplification to compare to these things. There is zero justification whatsoever for changing the testing parameters to help this stuff appear better than it is.

Test them properly. Preconditioned for 1 hour, and 5 minutes full power into their rated loads. Single and both channels driven. The power-cube/active load setup is a ridiculous expense and unless we all contributed for you to buy it, it's not justifiable IMO.

 

[Deleted member 2944]

John,
Why is this surprising to you? Much of the concept of these Class-D/SMPS-PS amplifiers is to reduce thermal issues and decrease PS size and cost.
They're designed to play music in your system and not heat the room of your lab. The engineering trade-off is obvious here and has been understood for many years.
Hypex makes no claims otherwise in either their amplifier module datasheets or the SMPS supply data sheets.
Dave.