March Audio Hypex Ncore Power Amplifiers

[March Audio]

My understanding is that it's 1/8th power?

 

[Timbo2]

I believe there was a case (no pun intended), not unlike the famous McDonalds hot coffee case, which may have resulted in many of the brands now installing perforated high temp plastic covers over the outer steel cover, directly over the heat-sinks. I know Denon, Yamaha and Pioneer have done it on high powered AVRs for a while now.

I believe you solved a mystery for me! I have an entry level Onkyo AVR that is around 10 years old. It has a piece of ventilated plastic clipped to it above the ventilated steel case. I couldn’t for the life of me figure out why it was there.

I just realize I could pull it off and get the dust out from under it and that the case was finished under it. It went in the trash...

 

[DonH56]

For stereo amps in The Olden Days (not saying John and I are old, but...) the rule was 1/3 of full rated power. That is typically the worst-case bias condition for a class AB amplifier. Modified nearly twenty years ago (no, I did not remember when, it's on the document). See this amendment to the spec fo Home Entertainment Products (generally considered to be AVRs and multichannel amps):

https://www.ftc.gov/sites/default/f...roducts/001222traderegulationrulerelating.pdf

Excerpt: "... and reduce the preconditioning power output requirement from one-third of rated power to one-eighth of rated power. "

Brought about because manufacturers complained the higher-power pre-conditioning was unrealistic and too harsh (see notes in the PDF). I.e. they wimped out so they could make stuff cheaper (less robust circuitry and thermal management). For the real (consumer) world they have a point but I was weaned in a worst-case world.

I think Stereophile and others still use the 1/3 full-power level for one- or two-channel amplifiers and I would use that as well. For a class D amplifier it does (should) not make that much difference though efficiency is worse at lower power so maybe 1/8 would be better (?) It would be interesting to see a plot of efficiency vs. output power for those amps (yes, I am sure there are plenty around, but I need to go get breakfast).

HTH - Don

 

[March Audio]

I think it may have changed, this is the info I have.

The FTC establishes fair advertisement practices for home audio power ratings. This is described in the FTC document 63FR37233, 16 CFR, Chapter 1, Part 432.

It defines how the amplifier should be tested for power and signal distortion.

The FTC requires that the amplifier be pre-conditioned at one-eighth of rated total power output (for a multiple-output system, all channels are on) for one hour using a sine wave at a frequency of 1,000 Hz.

The power spectrum measurement is then collected with two channels at maximum rated power over the audio frequency range of 20 Hz to 20,000 Hz, in ambient still air of not less than 25°C, for the a duration of not less than 5 minutes.

The amp efficiency is in the specs, I will dig it out, but basically varies between 75 and 85% in the upper part of the curve.

 

[restorer-john]

Brought about because manufacturers complained the higher-power pre-conditioning was unrealistic and too harsh (see notes in the PDF). I.e. they wimped out so they could make stuff cheaper (less robust circuitry and thermal management). For the real (consumer) world they have a point but I was weaned in a worst-case world.

According to all the FTC stuff I read, they actually never implemented the changes after several periods (extended) of industry consultation- they reached no conclusion apparently. In the end, they left the rule as it was. There is a lot of conflicting information even on the FTC's site. I went down that hole last year looking for clarification.

 

[March Audio]

Found it for the P252. So as you can see the efficiency falls off at lower powers.

 

[March Audio]

Helloworld asked a question in the Hypex Nc400 review thread.

https://www.audiosciencereview.com/...ments-of-hypex-nc400-diy-amp.5907/post-132111

"Does OEM perform better than DIY?"

The Nc400 diy module is very high performance as the review shows.

The Nc500 OEM module is better performance again,( wow).
The Nc122, 252 and 502 are little behind in terms of some parameters, but still very much top teir. I will get together some comparison data and plots.

 

[DonH56]

Efficiency falls in class D amplifiers because the input, (pre-)driver, and control circuit stays active (natch) even as the outputs are switching less. Once the devices are switching nominally the efficiency flattens out with a small "standing" power plus losses in the drivers and switching devices (scales with power so efficiency stays relatively constant). That is also true for conventional amplifiers but, unlike class A, AB, B, G/H/etc. amplifiers, there is no standing (quiescent) bias current in the output devices.

They do not offer higher-power NCore modules, just the older UcD modules, for DIY'rs.

 

[pos]

The Nc500 OEM module is better performance again,( wow).

It is not: the better S/N is simply due to the lower gain as it does not have an integrated buffer.

 

[March Audio]

It is not: the better S/N is simply due to the lower gain as it does not have an integrated buffer.

It is. I will run through it later. Just as a starter significantly higher power output. Yes the final performance is dependant on the input buffer.

 

[restorer-john]

I'm really not sure now after going down the FTC site hole again.

It seems they have un-amended the 'rule' to be 1/8 power for an hour after previously backflipping on an amendment?

"Accordingly, the Commission is
amending section 432.3(c) of the Rule
by reducing the specified per-channel
power output during preconditioning
from one-third of rated power output for
one hour to one-eighth of rated power
output for one hour."

"(c) The amplifier shall be preconditioned by simultaneously operating all channels at one-eighth of rated power output for one hour using a sinusoidal wave at a frequency of 1,000 Hz " ([39 FR 15387, May 3, 1974, as amended at 65 FR 81240, Dec. 22, 2000] )

and this:

"Accordingly, in light of CEA’s
comment, the Commission has decided
to defer action on the proposed rule, but
keep open the rulemaking record in this
proceeding to allow sufficient time for
CEA to address the issues raised in the
SNPR, and encourage the exchange of
ideas between the Commission and the
industry.
Authority: 15 U.S.C. 41–58." (Federal Register /Vol. 67, No. 10 /Tuesday, January 15, 2002)

 

[DonH56]

Nearly 17 years and they're still addressing the issues? I give up...

 

[March Audio]

It is not: the better S/N is simply due to the lower gain as it does not have an integrated buffer.

@helloworld As mentioned here is a table for comparison. All things being equal SNR will increase with output voltage (power). Final NC500 performance will be influenced by the input buffer performance.

 

[rebbiputzmaker]

Efficiency falls in class D amplifiers because the input, (pre-)driver, and control circuit stays active (natch) even as the outputs are switching less. Once the devices are switching nominally the efficiency flattens out with a small "standing" power plus losses in the drivers and switching devices (scales with power so efficiency stays relatively constant). That is also true for conventional amplifiers but, unlike class A, AB, B, G/H/etc. amplifiers, there is no standing (quiescent) bias current in the output devices.

They do not offer higher-power NCore modules, just the older UcD modules, for DIY'rs.

It actually pretty easy to obtain all the ncores.

 

[pos]

It is. I will run through it later. Just as a starter significantly higher power output. Yes the final performance is dependant on the input buffer.

I would be curious to see a real implementation measured, and what the final residual noise ends up to be with a 26dB total gain.
Conversely I would also really like to see a nc400 measured with a lower gain (eg R141 removed gives a 13dB total gain).
I agree that the nc500 has significantly more voltage output capabilities, but current capability is about the same. As a consequence it gives more power on a 8 ohms load, but in the end it is only a 3dB difference.

If you compare the nc400 and nc500 distortion vs power curves at 4 ohm you clearly see a trend were the nc500 (sans buffer) has a lower residual noise, giving it an edge in the first watt(s), but distortion does not go as low as the nc400 when the power goes up.
If the buffer (and associated gain) cancels that noise floor advantage then the nc400 will be the better performing amp all around.

 

[March Audio]

I would be curious to see a real implementation measured, and what the final residual noise ends up to be with a 26dB total gain.
Conversely I would also really like to see a nc400 measured with a lower gain (eg R141 removed gives a 13dB total gain).
I agree that the nc500 has significantly more voltage output capabilities, but current capability is about the same. As a consequence it gives more power on a 8 ohms load, but in the end it is only a 3dB difference.

If you compare the nc400 and nc500 distortion vs power curves at 4 ohm you clearly see a trend were the nc500 (sans buffer) has a lower residual noise, giving it an edge in the first watt(s), but distortion does not go as low as the nc400 when the power goes up.
If the buffer (and associated gain) cancels that noise floor advantage then the nc400 will be the better performing amp all around.

Oh yes absolutely, I am all for independant testing and the final Nc500 numbers will depend on what the input buffer does to it. Just a note the gain of the Nc500 power section is still 13.5 dB I believe.

Overall? Depends where your emphasis is.

 

[rajapruk]

It actually pretty easy to obtain all the ncores.

Do you know where nc500 modules can be bought?

 

[pos]

Just a note the gain of the Nc500 power section is still 13.5 dB I believe.

Yes, similar to the nc400 minimum gain configuration, compared to the more typical 26dB gain of the default configuration as measured in the datasheet and by @amirm.

Forgot to mention the nc400 also has half the output impedance, which can be an advantage in some situations.
Idle power consumption is also marginally lower: 4.5W+8W, compared to 6.3W+9W, a 22% difference.

 

[DonH56]

Do you know where nc500 modules can be bought?

See the main Hypex website (not the DIY website); they only list them for OEMs at present.

https://www.hypex.nl/

 

[March Audio]

Yes, similar to the nc400 minimum gain configuration, compared to the more typical 26dB gain of the default configuration as measured in the datasheet and by @amirm.

Forgot to mention the nc400 also has half the output impedance, which can be an advantage in some situations.
Idle power consumption is also marginally lower: 4.5W+8W, compared to 6.3W+9W, a 22% difference.

I'm not sure the difference in output z is of any significance in any circumstance. The 500 is 1.5 mohms. What benefit will 0.7 mohms make when you have subsequent speaker terminal posts and speaker wire? As a reference 1.5mm wire resistance is 12.5 mohms per meter. That's 25mohms round trip per meter. 2.5mm cable is about 7.5 mohms so 15 mohms per meter. I'm ignoring any inductance component.

Power consumption....an additional 1.8w. Really? Take a look at a class A or AB. Not to mention the 500 is twice the power output so the design is different.