March Audio P122 Class D Amplifier Measurements

[SIY]

Tell me more

There are many things I'd like to be doing with a speaker manufacturer. I will of course not mention borrowing amps from time to time.

 

[March Audio]

There are many things I'd like to be doing with a speaker manufacturer. I will of course not mention borrowing amps from time to time.

I will PM you later

 

[RayDunzl]

Perspective on this power business...

@amirm quotes 103W at the knee
Spec says 125W at 1%

Difference is less than 1dB

 

[March Audio]

And shipping costs from the land of Oz to Europa...

Hi @Xulonn

We do have a UK based distributor.

Keith at Purite Audio

https://www.puriteaudio.co.uk/

@Purité Audio

 

[Xulonn]

Hi @Xulonn

We do have a UK based distributor.

Keith at Purite Audio

https://www.puriteaudio.co.uk/

@Purité Audio

 

[ChrisPa]

The process for determining power at a set percentage is very stressful on the amplifier. The analyzer keeps hunting around trying to get to that near exact value, meanwhile pushing the amp into and out of clipping. It is also unreliable in that some amps simply can't go low enough in distortion. Even 1% THD+N causes problem with some amps and I have had to adjust that to 2% or even higher.

So for now, I like to stay where I am. The knee is the conservative assessment of the design limit of the amplifier. In a perfect world, every company would rate their amps this way and reviewers would measure the same way. A few watts lower or higher there is not material. Then, I have continuous and burst at 1% THD+N. Adding yet another test is more work and I am not sure it adds much value.

I think having distortion values for both the knee and a predetermined distortion value is the right way to do it.

Personally I've only really ever looked at the knee values when looking at/selecting amps. If I'm choosing a low distortion amp then I want to run it as a low distortion amp. If I'm going to run it into its higher distortion regions then I could get a higher distortion/higher power amp instead.

It's a shame the predetermined value is 1% as this does feel very high by today's performance capabilities, I tend to "eyeball" the distortion/power curve and makes a judgement on how much additional power I'd want to allow the amp to give - usually none.

Based on your comments of testing effort versus getting measurement results, I think you've got it about right.

 

[McFly]

Yes, I will send in another P502 soon for re-test.

Please do

 

[amirm]

It's a shame the predetermined value is 1% as this does feel very high by today's performance capabilities, I tend to "eyeball" the distortion/power curve and makes a judgement on how much additional power I'd want to allow the amp to give - usually none.

It is tricky to determine power at lower values. Some amps never hit smaller numbers. Indeed I often have to go up to 2% just to get some data.

If you look, my 1% wattage is not that much higher than the knee of the curve for many amps. So it is a good limit I think.

 

[mika91]

Hi,
It seems that the nc122mp modules performs pretty well.
I initially wanted to buy a nc252mp based amplifier to pair with my RME DAC, but not so sure now if I need to pay extra bucks for extra power. (
my current 2x50W amp knob was never set above -15dB.)
What do you think ?

 

[Sal1950]

What do you think ?

High quality amplifers have been fully transparent for quite a while now. IMO if you looking at amplifier purchases today, think of them as a lifetime purchase. Buy as much power as you can afford, you never know what you might need for your next set of speakers, which is about all you should ever need to upgrade again.

 

[Billy Budapest]

Different people's requirements do vary quite a lot. It does depend upon speakers, application; size of room, listening style etc. Even 75 watts (8 ohms rating) into typical 85dB/w/m speakers can generate decent sound levels, circa 103dB

I currently use a 25 watt per channel amp with my speakers (Triangle Celius)—granted, they are very efficient speakers. I never understood people who insisted on 100 watts per channel plus. It’s all marketing unless you have extremely inefficient speakers and listen at extremely high levels.

 

[McFly]

I use ATC scm40s which measure around 81db sens. I need every watt i can get in my lounge. I currently have about 300 watt per channel but might move to 500+. During movies im sure its clipping.

 

[March Audio]

I currently use a 25 watt per channel amp with my speakers (Triangle Celius)—granted, they are very efficient speakers. I never understood people who insisted on 100 watts per channel plus. It’s all marketing unless you have extremely inefficient speakers and listen at extremely high levels.

Well, as most things in life its rarely that simple.

I have found that more power does tend to mean that things sound less strained and more dynamic when you do turn the wick up. Its not really about maximum SPL levels because you need to double power to increase by just 3dB. To put that into context generally people consider an increase of 10dB to sound twice as loud. For reference disconnect one of your speakers, that is now 3dB quieter.

Also speaker sensitivity has a *big* impact. As mentioned above double the power is required to increase SPL by 3dB. So an 85dB/w/m speaker needs twice the power to sound as loud as a 88dB/w/m speaker.

Another consideration is the parameters of the speaker load. If it has extreme phase angles the amplifier has to cope with dissipating (wasting) far more power than is being usefully used in the load.

 

[March Audio]

I use ATC scm40s which measure around 81db sens. I need every watt i can get in my lounge. I currently have about 300 watt per channel but might move to 500+. During movies im sure its clipping.

wow that really is low.

You need our p701, 1200 watts in 2 ohms

 

[McFly]

I need 1200 into 8 please ATC state they can take 300, but then go on to say max SPL 112db. Which one would assume is at 1m. So taking that into account with 81db sens i think i could use 1200W... I dont believe it haha

 

[March Audio]

I need 1200 into 8 please ATC state they can take 300, but then go on to say max SPL 112db. Which one would assume is at 1m. So taking that into account with 81db sens i think i could use 1200W... I dont believe it haha

Oh just bridge 4 amps

Actually I have always wanted an excuse to build an amp based on one the Hypex NC2K modules

 

[tw99]

ATC spec them at 85db/w not 81...

 

[Billy Budapest]

Well, as most things in life its rarely that simple.

I have found that more power does tend to mean that things sound less strained and more dynamic when you do turn the wick up. Its not really about maximum SPL levels because you need to double power to increase by just 3dB. To put that into context generally people consider an increase of 10dB to sound twice as loud. For reference disconnect one of your speakers, that is now 3dB quieter.

Also speaker sensitivity has a *big* impact. As mentioned above double the power is required to increase SPL by 3dB. So an 85dB/w/m speaker needs twice the power to sound as loud as a 88dB/w/m speaker.

Another consideration is the parameters of the speaker load. If it has extreme phase angles the amplifier has to cope with dissipating (wasting) far more power than is being usefully used in the load.

Yes, all true. The Celius has an efficiency of 92 dB and in terms of impedance is a very easy load to drive, so my experience might not be typical. However, I think that a large part of focusing on wpc is marketing, especially given at “normal” listening levels most of us are only using 5-10 watts of power.

 

[kaka89]

ATC spec them at 85db/w not 81...

I am a SCM 40 user too.
I have seen 3rd party review/measurement said it is 84db, which not far from the spec 85db.

I wonder is Purifi 1ET400A good enough? or should I go with hypex 500 series

 

[Certainkindoffool]

Irrespective of fluctuating impedence and overhead requirements to avoided clipping, I think that if loudness curves are accomodated for, your going to be using (potentially significantly) more power at any given volume level below reference than your speakers efficiency rating would imply.

I would also suspect this is the case for many speakers after equalization(but before loudness compensation) as well.

Edit: Didn't intend to quote Billy's post there. Removed.