Marchaudio P501 Mono Block Power Amplifier Review

[Doodski]

I certainly agree with ^that.^

Is that 1% distortion actually clipping though?

I thought we were debating and pate rubbing about 0.1% versus 1%… and that clipping would likely be in the 10s of % or higher, pr maybe more like 50 to 100%.
(I admit I am bit lost.)

I think clipping should be as soon as flattening of the waveform is noticed. If that entails using a O-scope to dial it in then that might help.

 

[amirm]

Is that 1% distortion actually clipping though?

It is. Either current or voltage is maxed out causing that huge rise in distortion.

 

[Holmz]

It is. Either current or voltage is maxed out causing that huge rise in distortion.

Right.

One would not hear that in the subwoofer.

However the clipping sprays the harmonics out to the right, where they would be more easily heard.
so the 20Hz curve in and of itself is not a drama… it is the other frequencies truing to ride on top of that that create the listening drama.

I usually run a HPF between the preamp and amp(s), so the low freq is subdued... kind of exactly for the same reasoning To avoid clipping and not havethe amp lifting the entire load.

 

[amirm]

Agree. 500 Hz is probably the worst case for audibility of distortion as its harmonics easily travel into the region we are most sensitive. 20 Hz is there to fully cover the audible range and for those folks wanting to drive subs.

 

[PeteL]

What? This is absolutely worst case for this amp:

The sweep ended at that point, registering the highest clipping point. Heck, it is even after the clipping occurred with a second kink in the graph. There is no way you can defend this practice as it is not consistent whatsoever. If you changed the input sweep level, it could stop at a different point. Ditto for driving it even harder. And different amps have different points where this graph stops. So this does not help your argument one bit unless you want inflated power ratings for amps to sell them.

Sorry I don’t understand what you mean, what does it mean, the highest clipping point. The graph stopped there because the generator stopped there? no? It’s not a characteristic of the amp, is it? What doees that second kink mean? it doesn’t clip as hard past 400 Watts? measurment error? In all case what does it have to do with the amp, the highest clipping point is when an amp enter protection, not where the generator stop his sweep, no? what am I not getting?

 

[restorer-john]

Sorry I don’t understand what you mean, what does it mean, the highest clipping point. The graph stopped there because the generator stopped there? no? It’s not a characteristic of the amp, is it? What doees that second kink mean? it doesn’t clip as hard past 400 Watts? measurment error? In all case what does it have to do with the amp, the highest clipping point is when an amp enter protection, not where the generator stop his sweep, no? what am I not getting?

What you need to understand is all amplifiers behave differently in and around the point of clipping.

We have all sorts of behaviour. Current limiting can commence, control of the input level mechanisms, power supply rails can collapse either by themselves or design.

There is not a one size fits all approach. Some amplifiers can go backwards in terms of power output once the clipping point is reached in some cases.

These class Ds hang on until they hit the wall and then they hit it very hard. Other classes/designs behave totally differently.

 

[Doodski]

These class Ds hang on until they hit the wall and then they hit it very hard.

Apparently they hit hard early too so if one needs a 150w class AB then they better get a 200W class D because when they distort it sounds pretty bad. No issues there the class D amps have muchO power generally speaking.

 

[PeteL]

What you need to understand is all amplifiers behave differently in and around the point of clipping.

We have all sorts of behaviour. Current limiting can commence, control of the input level mechanisms, power supply rails can collapse either by themselves or design.

There is not a one size fits all approach. Some amplifiers can go backwards in terms of power output once the clipping point is reached in some cases.

These class Ds hang on until they hit the wall and then they hit it very hard. Other classes/designs behave totally differently.

Thanks, yes I understand that. That don’t expllain « Heck, you are even past clipping with a second kink » We are into clipping, I understand that, I don’t get the notion of « highest clipping point » by Amir’s explanation.

 

[JasonWells]

Bu

Did you not bother to read my post which you quoted? I do exactly that:

The sweep response is doing more than determining power. These bar graphs however, are dedicated to that and use consistent criteria and have high precision.

You are contradicting yourself. why is 1% distortion OK here but not in the sweep?

 

[JasonWells]

What? This is absolutely worst case for this amp:

The sweep ended at that point, registering the highest clipping point. Heck, it is even after the clipping occurred with a second kink in the graph. There is no way you can defend this practice as it is not consistent whatsoever. If you changed the input sweep level, it could stop at a different point. Ditto for driving it even harder. And different amps have different points where this graph stops. So this does not help your argument one bit unless you want inflated power ratings for amps to sell them.

How can taking a measurement at a fixed distortion point of 1% be inconsistent?

 

[don'ttrustauthority]

As you already know, 1% is one of the industrial standard. so most of Amp commercials are at 1%.
(Like https://www.tpdz.net/productinfo/821195.html)

1% is a commonly used level of distortion that many manufacturers use to advertise performance.

That does not make it an "industrial standard".

 

[JasonWells]

Well then you need to learn about amplifier design. An amp either runs out of voltage or current. In either case, it can no longer produce more power and pushing it further just produces deformed waveform indicated by rapid rise in distortion. No one should operate an amplifier in that region if they want excellent fidelity.

Of course, it produces distortion. what is your point?

Define "excellent fidelity". is it 0.0001% distortion, 0.001%, or 0.1 % or 1%. your happy with 1% for burst tests.

 

[Doodski]

@JasonWells what is your position on this % figure. Do you have a opinion. What is your back ground on the matter?

 

[JasonWells]

Thanks, yes I understand that. That don’t expllain « Heck, you are even past clipping with a second kink » We are into clipping, I understand that, I don’t get the notion of « highest clipping point » by Amir’s explanation.

I'm sure the amp would produce more power at 2%. just need to drive it harder. 1% is not the highest clipping point. It's a non made up term. it's just where thin blue stopped simply because 1% is a common reference point in amplifier testing.

 

[JasonWells]

Once more, this is done with the bar graph measurements. It is impossible to do with the sweep since you can't predict the distortion levels with fixed set of input levels. The only way to get there is to increase sweep resolution by 100 times, making the measurements take that much longer, and stress the amplifier that much. Even then you may not be able to get the same number in the exponential part of the curve.

The solution is the bar graph methodology where the analyzer hunts for specific fixed distortion level and when it gets there, it reports it. Not to ask for it in the sweep.

This doesn't make sense and again you are contradicting yourself. you seem quite able to read a value off the sweep and print it on the graph labelled as "max power". So it's not impossible as yiu state.

So present the graph on its own without the max power statement and use the bar graph method determine max power at a fixed level, 0.01%, 0.1% or whatever.

 

[MAB]

So you agree that your method gives variable max power readings for each amp. so you would agree that makes it difficult for people to compare one test to another. so you would agree that the headline "max power" you print on every graph can be misleading as it's not made under the same conditions each time.

So why don't you change your methods to determine power at a consistent level? if you feel it should be 0.001% thats fine. it doesn't really matter, although I would argue thats too low. Consistency however does matter.

Second point about the 1% level. it's not "the worst clipping point". that's just dramatising things for effect. it is simply 1% distortion level. its no less valid than the random points you choose that are far below any audible level of THD.

The research shows 1% is unlikely to be audible with music. you frequently test speakers that exceed this level, so don't ever egg the impact.

I don't even know what the "design parameters" of the amp means. it's simply a point where distortion starts to increase. The only relevant aspect of this is where on that knee the distortion starts to become audible. if you want to be conservative then research shows that with sine testing, the worst case scenario, it becomes audible around 0.1%. I don't see any justification for a lower point, but I would like to hear your view on that.

Also, If you object to 1% so much, why do you include the CEA busrt test which has its measurement point at 1%? you are contradicting yourself.

Basically I just don't understand your reluctance to choose a fixed distortion level, be it 0.001%, 0.01% or 0.1%. Variable levels on each test is the wrong thing to do. I don't know how you expect thin blue to conform to your methods when essentially there is no defined parameters for him to follow.

You should be able to answer every question you pose here from the graphs Amir provides. The fact that you ask means you either can't understand the graphs, or you are arguing for the sake of arguing.

 

[PeteL]

I'm sure the amp would produce more power at 2%. just need to drive it harder. 1% is not the highest clipping point. It's a non made up term. it's just where thin blue stopped simply because 1% is a common reference point in amplifier testing.

I think also. but I just wanted to understand what Amir meant.

 

[JasonWells]

You should be able to answer every question you pose here from the graphs Amir provides. The fact that you ask means you either can't understand the graphs, or you are arguing for the sake of arguing.

I can. This is exactly the point thin blue made. people can read the graph, but its Amir who objected to it. Many readers don't understand the graphs, they just see headlines.

 

[JasonWells]

There is no such "research."

Nonsense.

OK, are you up for some abx testing with sine waves? let's see what distortion you can hear. we can even try some of the levels you have defined max power at in recent tests, 0.001%, 0.085% etc.

 

[Holmz]

Agree. 500 Hz is probably the worst case for audibility of distortion as its harmonics easily travel into the region we are most sensitive. 20 Hz is there to fully cover the audible range and for those folks wanting to drive subs.

I am probably not the brightest fellow, but even I doubt I would get this particular amp to drive a sub.

But there are people with difficult, inefficient, full range speakers that likely need 400W.,., however I usually use a Vandersteen powered sub, or some other sub, and hence the inline XO to relieve the need for that much power for the first 1-1/2 octaves.
Luckily a 50-100W amp is then all I need.

Of course the amount of power is also heavily influenced by the gendre of music and amount of bass (EQ) etc, in addition to the raw SPL they want.


@thin bLue & @amirm - Maybe showing the rail voltage on the o-scope plot would help some people?
Or is it flattening from lack of current, and the rail is sagging down?

The power supply is arguably one of the more important parts of an amp. Some get around with feedback, and more “open loop”, low feedback amps, often have the rails locked down a bit tighter/stiffer.