NC252MP (class D) vs. A250W4R (classAB) burst measurements into 4ohm//2.2uF load

[DonH56]

Is there a graph of output impedance someplace? Curious...

 

[PeterOo]

I have already done it several times, so again -

View attachment 292267 View attachment 292266
It is a design flaw. Good amplifier must handle 2uF in parallel with 4ohm. Period.

Perhaps fair to report that the impedance curve for the A-1 is when you set the tone control to +32dB at 23kHz. Why would you use that as representative of actual loudspeakers?
There was no impedance curve for the flat setting they used for the further measurements, but there the impedance at 20 kHzwas somewhere between 4 Ohm and the 1.3 Ohm.

Sound-Lab A-1 electrostatic loudspeaker Measurements

Sidebar 6: Measurements

www.stereophile.com

Also how did you extrapolate the impedance up to the switching frequency?

 

[pma]

More appropriately, we're talking about an extremely tiny niche of speakers, none of which will be driven at those powers at those frequencies. Assuming the measurements are correct (and I have a few doubts), an amp that does fine with 99.99% of speakers is certainly a good amp. I have three different Hypex (or Hypex licensed nCore) amps here and they do very well driving both dynamic and electrostatic speakers.

However, they took a lesson at Hypex when designing the Nilai ....
You measured at 1kHz, Stuart?

P.S.: Image says 4R//2uF, text says 4R//4uF. What is true??

 

[ahofer]

Literally I am doing that. So I guess that is one... Price and rated speaker power handling are but two of many factors in determining what amplifier and speakers to buy and pair.

I've matched a sub-$1000 amp (March audio p252) with $11k speakers (Revel F228be), but they are an easier load than many less expensive speakers.

 

[SIY]

However, they took a lesson at Hypex when designing the Nilai ....
You measured at 1kHz, Stuart?

View attachment 292310

P.S.: Image says 4R//2uF, text says 4R//4uF. What is true??

2uF. Sorry, I missed the typo. For that power sweep, I used 1kHz. One figure that was cut for length was THD+N vs frequency with and without the 2uF load- the overlay was perfect.

I got similar results with the NCx500.

 

[pma]

2uF. Sorry, I missed the typo. For that power sweep, I used 1kHz. One figure that was cut for length was THD+N vs frequency with and without the 2uF load- the overlay was perfect.

I got similar results with the NCx500.

Thanks. In case of NC252MP, there is an interesting dependence of the amp stability/distortion on capacitance value. I tried the load 4R7 in parallel with 2.2uF, 4.4uF, 22uF, 33uF at 1kHz. Though with high capacitances 22uF and 33uF the module behaves nicely, with low noise and low distortion, with 2.2uF and 4.4uF it goes crazy. Please see below. There will be a border line somewhere when the FB loop gets stable again. Test frequency is 1kHz, BW22kHz.

 

[SIY]

Thanks. In case of NC252MP, there is an interesting dependence of the amp stability/distortion on capacitance value. I tried the load 4R7 in parallel with 2.2uF, 4.4uF, 22uF, 33uF at 1kHz. Though with high capacitances 22uF and 33uF the module behaves nicely, with low noise and low distortion, with 2.2uF and 4.4uF it goes crazy. Please see below. There will be a border line somewhere when the FB loop gets stable again. Test frequency is 1kHz, BW22kHz.

I don't have a NC252MP on hand to see if I get the same- if one crosses my path, I'll try your experiment.

 

[pma]

I don't have a NC252MP on hand to see if I get the same- if one crosses my path, I'll try your experiment.

The same I can say about the Nilai .

BTW, 4R7//2.2uF to 4R7//33uF is no especially difficult load at 1kHz as a test tone. Any amp should be able to pass it and any good class AB amp passes it. What it in fact tests is a FB stability, loopgain design.

 

[pma]

Thanks. In case of NC252MP, there is an interesting dependence of the amp stability/distortion on capacitance value. I tried the load 4R7 in parallel with 2.2uF, 4.4uF, 22uF, 33uF at 1kHz. Though with high capacitances 22uF and 33uF the module behaves nicely, with low noise and low distortion, with 2.2uF and 4.4uF it goes crazy. Please see below. There will be a border line somewhere when the FB loop gets stable again. Test frequency is 1kHz, BW22kHz.

I don't have a NC252MP on hand to see if I get the same- if one crosses my path, I'll try your experiment.

I have repeated the exactly same test with my A250W4R, class AB amplifier 2x250W/4R. I consider it and average, good, robust and reliable amplifier, but not a SOTA. It is a good old Douglas Self's dominant pole topology.
However, it behaves, under difficult load conditions, miles and miles ahead of the NC252MP. As you can see, parallel loading (with 4R7) of 2.2uF - 33uF makes no problem to this amp and gives almost same results, only reflecting the capacitor parallel impedance at 1kHz. It is even a better result than in your Nilai Audioxpress test, above some 10-20W. It is a robust, stiff amplifier. Class D still is unable to compete. Class D boom is all about business + marketing.

As a comparison - NC252MP

 

[SIY]

Class D still is unable to compete. Class D boom is all about business + marketing.

Uh huh.

 

[pma]

Unfortunately, the NC252MP, and I am quite sure it applies to any other amp from Ncore line, does not like much any kind of complex load, talking about distortion. It only likes pure resistive load.
I have just finished the test with a series complex load 3.4ohm + 2mH, which makes 5.08 impedance magnitude and +48° phase at 300Hz. A comparative measurement with pure resistive load 4.7 ohm was done. Inductor 2mH is an air core type without any iron parts.

One can see that with the complex load the distortion above 6V output (7W approx.) starts to rise compared to resistive load. So it goes, with light 5.08ohm/+48° load. No excuses to unrealistic load.

So, we can see in the reviews nice distortion plots that reviewers take with purely resistive load that never appears with a real speaker. And nobody knows how the amp under test would behave under realistic conditions, regarding distortion. I know that compared to speakers the nonlinearity shown is low, however, what is the significance of those ultralow numbers measured with resistors?? There is absolutely no consequence with distortion into complex load, especially in case of class D amplifiers like Ncore.

 

[PeterOo]

Unfortunately, the NC252MP, and I am quite sure it applies to any other amp from Ncore line, does not like much any kind of complex load, talking about distortion.

If this goes for all Ncore amps, then how is your measurement different from @SIY ’s who said he did not see any effect of 2uF // on THD+N vs frequency for Nilai en NCx500 in post #805?
He measured at too low output for the difference to show?

EDIT: typo vs frequency instead of vs power.

 

[NTK]

If this goes for all Ncore amps, then how is your measurement different from @SIY ’s who said he did not see any effect of 2uF // on THD+N vs power for Nilai en NCx500 in post #805?

SIY's graph does show some degradation (but totally inconsequential) in the performance. For reference, below are the datasheet performance and SIY's measurements.

 

[valerianf]

NCx and Nilai are new hardware designs.
The detected complex load issue may have been fixed.
According to a web interview NCx is the budget version and Nilai the SOTA result.
It seems that now the older NC design need to be avoided.

 

[Buckeye Amps]

It seems that now the older NC design need to be avoided.

Guess I need to replace the NC502MP/NC252MP setup I've been using.

 

[valerianf]

"Guess I need to replace the NC502MP/NC252MP setup I've been using."
Only if you find any audio quality issue .
I have a ICE class D amp inside a Yamaha WXA-50 and I am fully satisfied for the surrounds.

 

[SIY]

It seems that now the older NC design need to be avoided.

Really?

 

[Buckeye Amps]

Only if you find any audio quality issue .

That goes against what you said about the older NC needing to be avoided

 

[valerianf]

"That goes against what you said about the older NC needing to be avoided"
Class D amplifiers are electronics products.
Why to want to buy an old design when the manufacturer just launched a new improved product?

A few years ago I wanted to buy a Yamaha class D amp: ICE module was the only choice.
Why Yamaha: to get the warranty from them and the pre-amp/remote features.

After reading this thread, I would now choose a NCx or Nilai.
@pma clearly proved that there is a design issue with the NC design connected to a complex load.

 

[Buckeye Amps]

Why to want to buy an old design when the manufacturer just launched a new improved product?

After reading this thread, I would now choose a NCx or Nilai.

I can think of two big reasons right away:
1) Large speaker counts for HT usage
2) Wanting great sound but on a budget