Review and Measurements of Hypex NC400 DIY Amp

[DonH56]

How do you know this? Anyway, you guys are making me wonder. While I expect this industrial product to actually do what it says (it IS a mechanical bond system, that cannot dry out, and has very high sheer strength, in an application with truly little sheer force), still, I don't know anything about it's degradation over time. If it was bad, it shouldn't be in use anymore. But, I could, easily, secure a loop of SS wire around the assembly that would ensure that the little guys could not simply "fall off" and find the sparkingest places available. Anyway, enough about my choices. Back to your regularly scheduled NC400 sonic qualities programming!

I don't know anything about whatever specific glue you used, or much about glue in general, so could be completely off-base. If you trust it, that's fine; my experience is older but contrary (and fairly extensive over a lot of glued-on parts over the years). Part of the problem is the thermal cycling that happens. The metal and glue typically have different expansion coefficients so the glue is repeatedly stressed over time as you turn the amp on and off. That causes cracks that gradually get wider and more numerous, and the glue gets brittle over time from heat and oxidation, so eventually the parts fall apart (usually when you jostle or pick up the component).

Again, glue is not my thing, just relating my experience. Me, I'm from Missouri ("show me") and a skeptic, having been burned by failed glue joints many times over the years (and not just in electronic components; wood glue does not last forever, either). I was also taught to not trust adhesives alone for mechanical integrity over time and stress, but that may be more a function of high-rel systems for which I designed ICs (flight systems, space systems, shipboard and underwater sensors, etc.)

 

[VintageFlanker]

Perfect match with the ADI-2 DAC. The max input is rated at 1.66V (6.61 dbu), so I work with the +7dbu ref level. That way, I may deal with all the power available from - 96db to 0db!

View attachment 32101

Not exactly NC400s, but if anyone interested...

 

[casual_audio_nerd]

A few weeks in and my NC400's are running well. Not overheating, and there is a distinguishable difference in the "tightness" of the low frequencies. It's not life-changing, but it's a noticeable upgrade. The lack of trigger in got me experimenting with a WeMo smart plug so that I can use voice activation/scripts to power them on/off (the smart plugs are connected to a CyberPower Sinewave UPS. Now playing with Dirac Live on the NAD C658 to work around the challenges of living in an L-shaped loft.

 

[Matias]

The lack of trigger in got me experimenting

You can try one of those smart eco surge protector line strips that turns some receptacles on based on/off a "master" device connected to it. Works great. Cheap automation and saves money in the electricity bill.

https://www.amazon.com/Tripp-Lite-Protector-Right-Angle-TLP808NETG/dp/B003A816FM/ref=mp_s_a_1_12

 

[spaace]

Here is the square wave 10 watts before someone (John?) asks.

View attachment 19472

It is at 90 kHz bandwidth, with frequency of 1 kHz.

I know its a bit late - I am just getting to know and learn square waves. So when i was comparing the measurements over at stereophile of a hypex amp the resolution of the graph and hence the shape seems to be different

Thats coming from here

One thing i noticed was that the resolution is 20u in stereophile measurements and 200u in the measurement you have posted.
At what scale is the graph supposed to be looked at ?

Thanks

 

[HammerSandwich]

It's more about interpreting the scale that's presented.

The different timing is because Amir showed 1kHz while Stereophile showed 10kHz. Squarewaves are composed of odd harmonics, so the 1kHz shows a more square waveform, because it includes more harmonics (before the amp rolls off the HF). See Wikipedia's animation.

 

[DonH56]

The scale used depends upon the frequency and thus period of the waveform: frequency f (Hz) = 1/T where T = period (seconds).

Frequency = 1 kHz = 1000 Hz so period T = 1/1000 = 0.001 s = 1 ms = 1000 us.
Frequency = 10 kHz = 10,000 Hz so period T = 1/10000 = 0.0001 s = 0.1 ms = 100 us.

So you change the scale to show a complete period of the waveform you want to observe.

HTH - Don

 

[spaace]

Thanks that explains the scale and why its more square in the graph we have in this thread.

That still does not tell me whether the not-so-square wave seen in the stereophile measurements i linked above is ok or not. I ask since, i went in and compared a few other class a/ab amp measurements and they look neat and square.

Can you please explain why this is not a concern ? I would be really happy to know and can make a purchase, which i have been holding off worrying if it means anything bad.

Thanks

 

[DonH56]

Thanks that explains the scale and why its more square in the graph we have in this thread.

That still does not tell me whether the not-so-square wave seen in the stereophile measurements i linked above is ok or not. I ask since, i went in and compared a few other class a/ab amp measurements and they look neat and square.

Can you please explain why this is not a concern ? I would be really happy to know and can make a purchase, which i have been holding off worrying if it means anything bad.

Thanks

That square wave does not look too bad to me. It has a small amount of overshoot, probably from the output filter, but is very well controlled and quickly damped. The edges depend upon amplifier bandwidth and slew rate so you need to compare 10 kHz square waves and not 1 kHz to 10 kHz. While many amps may do better quite a few do much worse.

You might want to browse some of the technical articles on ASR. Here is one about square waves: https://www.audiosciencereview.com/.../composition-of-a-square-wave-important.1921/

 

[Julf]

The lack of trigger in

Both the nc400 and the smps have a facility to provide the trigger functionality.

 

[Hayabusa]

Amir,

I have build two mono blocks NC400/SMPS600.
I could hear a very faint tone when I switched them on. (I can hear this from the main listening position, 3.5 meters)
This sound is coming directly from the amplifier (or power supply).
So I measured the amp with microphone: there is a 450Hz (9th harmonic of the mains) tone at 21dB (@ 1meter).

Did you ever do such a measurement?
Its great to have -120dB noise levels on the speaker terminals, but it makes not much sense if the amplifier itself is making audible sounds clearly above that level.

 

[JimB]

Both the nc400 and the smps have a facility to provide the trigger functionality.

Can you say more about how you would implement such a feature?

 

[JimB]

... I have build two mono blocks NC400/SMPS600.
I could hear a very faint tone when I switched them on. (I can hear this from the main listening position, 3.5 meters)
This sound is coming directly from the amplifier (or power supply).
So I measured the amp with microphone: there is a 450Hz (9th harmonic of the mains) tone at 21dB (@ 1meter).

That is not normal in my experience, but I have heard something like it. Doing initial power-up checking of one of the NC400 amps I built, using test signals, with no speaker connected and with the case open in front of me, I could hear the test signals from the NC400 module, directly! Fortunately, the sound was drive-level dependent and tracking the program material, so I am unlikely to notice it with the case closed, 3 meters from the listening position, at normal playback levels. I suspect it is related to inductor vibration, but I haven't found time to address it with Hypex.

 

[Hayabusa]

That is not normal in my experience, but I have heard something like it.

Both amps have the same behaviour, the sound is there all the time, independant of input signal or load. I guess its the power supply.

 

[JimB]

Both amps have the same behaviour, the sound is there all the time, independant of input signal or load. I guess its the power supply.

If you want to pursue it further, unplug the power, remove the top cover, then power up and try to locate the source of the sound (carefully). If it is not obvious, you can use a cardboard tube to direct sound from a specific region to your ear. (Do not use a roll of aluminum foil!)

 

[restorer-john]

That is not normal in my experience, but I have heard something like it. Doing initial power-up checking of one of the NC400 amps I built, using test signals, with no speaker connected and with the case open in front of me, I could hear the test signals from the NC400 module, directly!

Under dummy load testing, many amplifiers will reproduce test tones from various parts- output transistors, heatsinks and sometimes the Zobel network inductors. It's pretty normal.

 

[JimB]

Under dummy load testing, many amplifiers will reproduce test tones from various parts- output transistors, heatsinks and sometimes the Zobel network inductors. It's pretty normal.

My case didn't concern me, overly, though it was without a load. But I would not live with what Hayabusa reports.

 

[Deleted member 2944]

Amir,

I have build two mono blocks NC400/SMPS600.
I could hear a very faint tone when I switched them on. (I can hear this from the main listening position, 3.5 meters)
This sound is coming directly from the amplifier (or power supply).
So I measured the amp with microphone: there is a 450Hz (9th harmonic of the mains) tone at 21dB (@ 1meter).

Did you ever do such a measurement?
Its great to have -120dB noise levels on the speaker terminals, but it makes not much sense if the amplifier itself is making audible sounds clearly above that level.

clearly something amiss with your units. Why can't you localize it to either the PS or amp module with a simple ear test????

Dave.

 

[amirm]

Did you ever do such a measurement?

No, I don't make acoustical measurements of noise as it is rare. When I hear it though I note it like I did in the Amazon Amp review (company confirmed that is what it does). In this case I did not hear such noise but I routinely wear headphones for testing so maybe it did it and I did not notice.

Coil whine which is the technical term is load, condition and sample dependent. Take a chop stick and carefully push on transformers with it and see if the noise goes away or changes. Some "goop" (silicone, etc.) may help. How the board is mounted to the chassis may also cause this.

 

[Julf]

Can you say more about how you would implement such a feature?

Right now on the road, and don't have the data sheets in front of me, but the smps600 has a "SMPS enable" input that switches the power supply (and of course the amps too) into standby (or out of it), and the nCores themselves also have an "enable" input for the same purpose.