Could you give a bit of background info on why the 10th harmonic has been the standard? As I understand class d amp hf noise is caused by the PSU and is not harmonic distortion from a frequency in the audible band?
-
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
- Thread starter amirm
- Start date
Hf noise in class D is part of class D and normally out band and filtered. Ps noise is another construction area.
The IEC and IHF decades ago defined THD using ten harmonics, and many analyzers like the AP use ten as well. I've forgotten the rational; my vague memory is that only low-order harmonics were significant (for example, an old ADI app note says only the first five are usually significant) and ten was deemed to be more than enough. Higher harmonics tend to be buried in the noise floor and thus difficult if not impossible to separate from noise. These days, noise floors are lower, DSP allows much more noise averaging to reduce noise, and so higher harmonics are more readily extracted. But, ten is still reasonable from what I have seen. Higher harmonics, e.g. 5-9, are large enough to rise above the noise floor and, rather than rolling off, tend to flatten out in amplitude as feedback rolls off. When -60 dB was excellent, you'd rarely see more than the second and third harmonics, but now with -100 dB and better common those higher-order harmonics are more readily seen.
As an interesting aside, the IEEE power standard (519) for power lines uses 50 harmonics to define THD, but that is for 60 Hz stuff.
The HF noise from a class D amp is not harmonic distortion per se but is still HF energy that is both signal-modulated and can cause tweeters and such to misbehave so remains of interest. Higher switching frequencies are more prevalent now so the energy is further above the audio band but watching the ultrasonic energy they produce is still a relevant measurement in my mind.
FWIWFM - Don
As an interesting aside, the IEEE power standard (519) for power lines uses 50 harmonics to define THD, but that is for 60 Hz stuff.
The HF noise from a class D amp is not harmonic distortion per se but is still HF energy that is both signal-modulated and can cause tweeters and such to misbehave so remains of interest. Higher switching frequencies are more prevalent now so the energy is further above the audio band but watching the ultrasonic energy they produce is still a relevant measurement in my mind.
FWIWFM - Don
Here is the Stereophile measurement of the AHB2 19/20 kHz:
It's hard to tell with these two graph scales, but Hypex NCx500 and AHB2 appear to be equivalent in the audible range, roughly around -112 dB.
- Rich
Last edited:
Once again - if we have a generator with 40 ohm output impedance and we shall load it with 1.8 kohm impedance with 1.8 ohm nonlinearity (-60 dB of original impedance), we shall measure -93 dB nonlinear distortion. Similar to post #1 results. And, we coul see that the buffer has improved overall linearity at higher voltage level. So I think it is clear.
Especially we cannot compare 2 distortion measurements that were done at different output power!!
My power was 5W. (NC500OEM bridged/Marantz)
The Marantz PM-11s2 might do even better than the Pm-10. Will measure and update.
Not sure if you are asking specifically about a stereo config of the NcoreX or just generally. I have tested Hypex bridging previously and seemed just fine. This was using NC502MP. Although Hypex only recommends 8 ohm loads when bridging it, they show 4 ohm performance in the data sheet...
My takeway was they more likely mean to avoid a complex 4 ohm load that may drop even lower at some frequencies.
Hm hard stuff after 49 messages here. Is it possible to formulate the things without to be personaly.
Maybe dial it down? John may be strongly opinionated but at least talks about the products, not the persons...
Most definitely yes! Please
paulraphael
Active Member
Why not? You'd just need a 1/4" in and a speakon out.
I'm waiting to see if the bass amp makers start using these newer generation modules in their heads. They tend come out with new models (using new modules) at more or less the same time. IcePower has been the most popular since the beginning of the class-d revolution. My amp uses an older generation module that's rated at 500W into 4 ohms. It won't do 2 ohms. Newer versions are rated ~800W into either 4 or 2 ohms. I believe these all have lousy distortion specs at full power. But bass is a bit forgiving of this.
Would be cool to see little amps that could the put out 1K or more into 2 ohms, without noise or distortion problems. This would lead to lots of flexibility (ability to use 1 to 4 speaker cabinets depending on your needs. I suspect price is still a problem here. And the demand for lots of power seems to be going down as more venues get modern PA systems.
Remember, it would be a double-edged sword. If you or I start such thread, it would be barely visited and will be forgotten in 2 days. The topic of generator output impedance value and possible non-linear low input impedance of the non-buffered Hypex module is related to this thread and to another similar thread with the same issue of distortion rise above 5 - 10W. It is a universal issue, people just are not experienced with such technical scenario. Who knows how the input of NCx500 without buffer looks like? Why is it not appreciated to find real reason of the distortion behaviour?
He won’t be back anytime soon……..or later even.
AndreaT
Addicted to Fun and Learning
I enjoyed the review immensely. I am happy the module and the PS were tested as they are the heart and lungs of future amplifier. I commend Amir, Rick and Hypex for this preview review of an outstanding new module: it clearly spells out what it is and what it is not. I am very happy about ASR.
Fred, it is not about driving 2k linear resistor, I will try to explain again. It is about driving 2k possibly nonlinear impedance. Nonlinear means it changes value with voltage level. Even 2 ohm of nonlinearity would bring distortion rise to about -90 dB, when driven from 40 ohm zero distortion generator. Do you know how the input circuit of non- buffered NCx500 look like? Can you assure it behaves like a pure 2k resistor? I am trying to be serious, this is a qualified debate.
Similar threads
- Poll
