• 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!

Hypex NCx500 Class D Amplifier Review

Rate this amplifier:

  • 1. Poor (headless panther)

    Votes: 3 0.6%
  • 2. Not terrible (postman panther)

    Votes: 6 1.3%
  • 3. Fine (happy panther)

    Votes: 56 12.0%
  • 4. Great (golfing panther)

    Votes: 401 86.1%

  • Total voters
    466

CleanSound

Major Contributor
Forum Donor
Joined
Apr 30, 2023
Messages
1,654
Likes
2,520
Bandwidth is simply the range of frequencies across which the measurements are made.
That doesn't make sense.

The range can be 0Hz to 1mHz, it's irrelevant because the value plotted is for a specific frequency, at a specific impedance, at a specific power output. At those specific points, that is the THD+N value.
 
Last edited:

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
2,552
Likes
3,435
Location
Detroit, MI
That doesn't make sense.

The range can be 0Hz to 1mHz, it's irrelevant because the value plotted is for a specific frequency, at a specific impedance, at a specific power output. At those specific points, that is the THD+N value.

THD+N depends on what bandwidth you are measuring over. The noise (N) component of THD+N greatly depends on what frequencies you are integrating noise over, the higher the bandwidth the higher the noise. For higher frequencies measurement bandwidth will impact THD as higher harmonics will be ignored if the bandwidth is too low.

Some examples illustrating this.

1 kHz test tone - 20 kHz bandwidth
MOTU M4 1K 0-20K BW.png


1 kHz test tone - 45 kHz bandwidth
MOTU M4 1K 0-45 BW.png


THD between the two is pretty much the same. 20 kHz bandwidth counts up to the 20th harmonic while 45 kHz bandwidth counts up to the 45th harmonic. However as more noise is integrated with the 45 kHz bandwidth noise is about 3 dB higher.

11 kHz test tone - 20 kHz bandwidth
MOTU M4 11K 0-20K BW.png


11 kHz test tone - 45 kHz bandwidth
MOTU M4 11K 0-45K BW.png


At 20 kHz bandwidth there are no THD components as the 2nd harmonic is at 22 kHz which is outside of the measurement bandwidth. With the 45 kHz bandwidth you get up to the 4th harmonic. As with the first example integrated noise is higher with the 45 kHz bandwidth.

Michael
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,117
Likes
11,018
Location
São Paulo, Brazil
Apollon NCx500ST listed. Possibly the cheapest assembled NCx500 in EU...
Yes, Apollon announced this in April.

And surprisingly 644W into 4 ohms both channels driven even using a single SMPS1200.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,952
Likes
16,819
Location
Monument, CO
That doesn't make sense.

The range can be 0Hz to 1mHz, it's irrelevant because the value plotted is for a specific frequency, at a specific impedance, at a specific power output. At those specific points, that is the THD+N value.
Read Michael's @mdsimon2 's excellent post. The bandwidth mentioned is the full bandwidth of the measurement system, not the narrow bandwidth around the test tone. If the measurement bandwidth is 20 kHz as in the Hypex spec, then the third harmonic (at 30 kHz) of a 10 kHz tone is already rolled off (reduced) by the low-pass filter before the THD measurement is taken. That will yield lower THD than if the third harmonic was included in the measurement. Reduced bandwidth for THD is more prevalent for class D designs IME due to their ultrasonic switching frequency, leading to inclusion of a low-pass filter on the amplifier's output. Note that most amplifiers include some sort of output LPF to reject noise and for stability, but they tend to have higher corner (roll-off) frequencies for A or AB designs, again IME.

Limited bandwidth is used more for noise measurements to improve stated SNR. A-weighted noise, following the ear's frequency response, is pretty commonly specified along with a wideband noise measurement.

HTH - Don
 

CleanSound

Major Contributor
Forum Donor
Joined
Apr 30, 2023
Messages
1,654
Likes
2,520
THD+N depends on what bandwidth you are measuring over. The noise (N) component of THD+N greatly depends on what frequencies you are integrating noise over, the higher the bandwidth the higher the noise. For higher frequencies measurement bandwidth will impact THD as higher harmonics will be ignored if the bandwidth is too low.

Some examples illustrating this.

1 kHz test tone - 20 kHz bandwidth
View attachment 293766

1 kHz test tone - 45 kHz bandwidth
View attachment 293765

THD between the two is pretty much the same. 20 kHz bandwidth counts up to the 20th harmonic while 45 kHz bandwidth counts up to the 45th harmonic. However as more noise is integrated with the 45 kHz bandwidth noise is about 3 dB higher.

11 kHz test tone - 20 kHz bandwidth
View attachment 293767

11 kHz test tone - 45 kHz bandwidth
View attachment 293768

At 20 kHz bandwidth there are no THD components as the 2nd harmonic is at 22 kHz which is outside of the measurement bandwidth. With the 45 kHz bandwidth you get up to the 4th harmonic. As with the first example integrated noise is higher with the 45 kHz bandwidth.

Michael
Maybe an explanation between tone vs. bandwidth would help.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,952
Likes
16,819
Location
Monument, CO
Maybe an explanation between tone vs. bandwidth would help.

There's a thread discussing THD and IMD on ASR: https://www.audiosciencereview.com/...armonic-and-intermodulation-distortion.25436/

The test tone is the signal under test, e.g. a 1 kHz tone is a 1 kHz signal injected into the amplifier. A tone represents a single signal (amplitude and frequency). You measure the THD from a range of basically DC (0 Hz) to 20 kHz or some other upper bandwidth limit. The 1 kHz tone will have harmonics (distortion) at 2 kHz, 3 kHz, 4 kHz, etc. multiples of 1 kHz if the amplifier is not perfectly linear. If your measurement bandwidth is 0 to 1 kHz, rolling off everything above 1 kHz to -infinity (0 output amplitude), then THD will be -infinity (perfect) because no harmonics are measured by the test meter within the measurement bandwidth. Expand the measured bandwidth to 10 kHz, and now you get nine harmonics (at 2, 3, 4, 5, 6, 7, 8, 9, 10 kHz) included in the measurement, and THD will no longer be perfect but something less (for a real amplifier).
 

Matias

Master Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
5,117
Likes
11,018
Location
São Paulo, Brazil
It is as if NCx500 has extra focus on reducing distortions on bass and mids (see IMD distortion at 1 kHz) a letting go earlier in treble (below 20 kHz), where musical signal is lower and the ear less sensitive. I wonder if that was a design goal.
A software engineer would tell you that this is a feature...
I see it as a consequence of a less punchy control loop compared to eigentakt and nilai.
Going back to this observations. On the July issue of audioXpress Frank Veldman says:

"The aim with Hypex amplifiers has always been to have excellent performance across the entire audio band, for UcD and NCORE this meant distortion levels were practically frequency independent. For the development of NCOREx and Nilai this has been reconsidered and while extensive (listening) tests obviously showed the importance of proper high frequency performance, it did not necessarily lead to full frequency independence. For new developments this meant high frequency performance has been as important as ever in Hypex designs, the philosophy of what is happening below that changed to some extent. (...)"

I guess it kind of confirms my theory.

NC500
1687278256276.png


Indeed on NC500 "distortion levels were practically frequency independent".

NCx500
1687278274020.png


And that part "extensive (listening) tests obviously showed the importance of proper high frequency performance, it did not necessarily lead to full frequency independence" can be seen as prioritizing to reduce distortion < 5 kHz, where the ear is more sensitive and the music signal is stronger, while keeping the > 5 kHz around the same levels as before. Very interesting.

But then if the new Purifi 1ET7040BA can maintain the insane low levels of low distortion in all frequencies then it's the best below and above 5 kHz. :)
 
Last edited:

CleanSound

Major Contributor
Forum Donor
Joined
Apr 30, 2023
Messages
1,654
Likes
2,520
Read Michael's @mdsimon2 's excellent post. The bandwidth mentioned is the full bandwidth of the measurement system, not the narrow bandwidth around the test tone. If the measurement bandwidth is 20 kHz as in the Hypex spec, then the third harmonic (at 30 kHz) of a 10 kHz tone is already rolled off (reduced) by the low-pass filter before the THD measurement is taken. That will yield lower THD than if the third harmonic was included in the measurement. Reduced bandwidth for THD is more prevalent for class D designs IME due to their ultrasonic switching frequency, leading to inclusion of a low-pass filter on the amplifier's output. Note that most amplifiers include some sort of output LPF to reject noise and for stability, but they tend to have higher corner (roll-off) frequencies for A or AB designs, again IME.

Limited bandwidth is used more for noise measurements to improve stated SNR. A-weighted noise, following the ear's frequency response, is pretty commonly specified along with a wideband noise measurement.

HTH - Don
@mdsimon2

Makes perfect sense now. Thanks for posting.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,952
Likes
16,819
Location
Monument, CO
Here is a plot showing how THD changes with measurement bandwidth. This is NOT a real amplifier! For illustrative purposes, all harmonics are -80 dB relative to a full-scale 0 dB 1 kHz signal tone. If the BW is less than 2 kHz, then THD is -infinity, off the chart (perfect). At 2 kHz bandwidth, we pick up the second harmonic at -80 dB, and THD jumps to -80 dB. The harmonic distortion does not add linearly in dB; you must RSS (root-sum-square) the harmonic signals to get the resulting THD. This means the plot is a gradually (logarithmically) rising line as you increase the measurement bandwidth and more harmonics are added to the THD level. By the time you reach 20 kHz bandwidth, the THD has worsened to -67 dB, without changing the actual harmonic levels from the amp but just increasing the measurement bandwidth to include more harmonics. In a real amplifier, higher-order harmonics would fall in amplitude until a certain high frequency is reached, then tend to flatten out, so the curve would not rise as rapidly. This is just to show how measurement bandwidth can impact the reported THD.

1687282608680.png


HTH - Don
 

Philipp

Active Member
Joined
Jan 30, 2023
Messages
130
Likes
157
Location
Trier, Germany

Philipp

Active Member
Joined
Jan 30, 2023
Messages
130
Likes
157
Location
Trier, Germany
I see this in step two of checkout.
Sorry if I am wrong, i do not want to win an argument, this would be just generally great news.
96F7BDEE-7883-4DEE-A366-992996844DA4.png
 

Philipp

Active Member
Joined
Jan 30, 2023
Messages
130
Likes
157
Location
Trier, Germany
1% battery remaining?
I have 5 kids, so its the same with me, even with 1% batterie left I still have to go on for ages ;-)
So my electronics better keep up or get replaced.
 
Last edited:

Philipp

Active Member
Joined
Jan 30, 2023
Messages
130
Likes
157
Location
Trier, Germany
Lucky me I live in Germany, toujours bien de vivre proche de la France!
 

DoesNotExist

Member
Joined
Nov 12, 2022
Messages
30
Likes
29
Location
Roma

amadeuswus

Active Member
Forum Donor
Joined
Jul 8, 2019
Messages
279
Likes
266
Location
Massachusetts
Apollon NCx500ST listed. Possibly the cheapest assembled NCx500 in EU...
If I am reading the cart numbers correctly, it seems competitively priced even for US buyers. (The version with standard buffer comes out to about $1300 shipped for me.)

On this forum, Apollon seems to garner praise for neatness and build quality. I also like the switchable balanced/unbalanced inputs. Almost tempted to take a chance that the NCx500ST won't ever need factory service....
 
Top Bottom