• 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!

Marchaudio P501 Mono Block Power Amplifier Review

Status
Not open for further replies.

Buckeye Amps

Major Contributor
Audio Company
Forum Donor
Joined
May 28, 2020
Messages
1,655
Likes
6,514
I would be very surprised if 1% of trained listeners can identify the Purifi 7040SA versus the Purifi 400A in a level matched blind listening test. You might find a rare system that offers a tell but I would really have to dig to find a speaker configuration where the difference is easily identifiable.
Of course within operating range, I agree there is almost no audible difference.

What I'm referring to is the scenario of someone with 2ohm speakers (or even a subwoofer) where the 400a is being pushed to its limit or even past, the 7040sa would be am advantage and that's when you may hear a difference (since the 400a in this scenario may be clipping but the 7040sa isn't).

Again, a niche scenario for sure.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Would you agree a typical hifi speaker is probably about 85dB/2.83v/m? A 8 ohm speaker is 1 watt at this level. most speakers fall below this impedance and are less efficuent. 6 ohms a fair average?
So at 1 watt maybe 84dB/1m.
At 250mW this is 78dB at 1m. if you listen at 3m this falls to about 71dB. all rough numbers, but I don't think 71 dB is loud. doesn't Amir test at 5 watts because he considers it a typical listening level?

I'm listening right now at the level I listen when doing other stuff - in this case browsing ASR. I just turned on rew with my Umik1 and the dB meter. Average over a minute was about 68dB, Max (over about a minute) was 75dB

I'm listening near field about 1m (or less) from each speaker. Speakers are 88dB and 6Ohm, so 1.5W for 88dB and 1W gives just over 86dB

So that is about -18dBW average, and -13dBW max: so about 16mW and 50mW.

If I turn it up to a "critical" listening level you can add less than 10dB to that - so 78dB average and maybe 85 Max. That is pretty loud for me, and I'd probably have to be in the house on my own. Still less than a watt at loudest parts (ignoring transient peaks), and around 158mW average.

Even if I was listening at 3m my "loud" listening will be less than 1W average.
 
Last edited:

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
3,303
Likes
3,838
3. I do run power vs frequency tests. Here it is:
index.php


This is superior to picking one power output and running the sweep vs frequency. You see full data for complete power sweep, albeit at discrete frequencies. Notice how picking the "knee" helps here to show the limited power at 20 Hz.
It's a very good example, in this example, 20 Hz appear to be clearly more power limited than all other frequencies, still it have the knee at the same place, 135 Watts as all the others. The graph tells us that it's clipped harder than the spectrum, but the power value at the knee don't give this information. You have to take into account the steepness of the curve. A knee, does indeed tells us that the amp stopped acting linearly, no debate there, but in this graph, where is it actually where the amp stop being linear? (allowing for tolerance for the number of captured values, just general concept) Is it 30? 200? approx 350?

1670104484872.png
 

PeteL

Major Contributor
Joined
Jun 1, 2020
Messages
3,303
Likes
3,838
I'm listening right now at the level I listen when doing other stuff - in this case browsing ASR. I just turned on rew with my Umik1 and the dB meter. Average over a minute was about 68dB, Max (over about a minute) was 75dB

I'm listening near field about 1m (or less) from each speaker. Speakers are 87dB and 6Ohm, so 1.5W for 88dB and 1W gives just over 86dB

So that is about -18dBW average, and -13dBW max: so about 16mW and 50mW.

If I turn it up to a "critical" listening level you can add less than 10dB to that - so 78dB average and maybe 85 Max. That is pretty loud for me, and I'd probably have to be in the house on my own. Still less than a watt at loudest parts (ignoring transient peaks), and around 158mW average.

Even if I was listening at 3m my "loud" listening will be less than 1W average.
Why are you ignoring transient peaks in your analysis of needed power?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
The graph tells us that it's clipped harder than the spectrum, but the power value at the knee don't give this information.
That measurement is more of an exception than rule in that regard. Here is another example that shows it more clearly:

index.php


You can clearly see the knee is earlier at 20 Hz than others.

This graph has limited resolution on purpose. The amp is being stressed repeatedly with the full sweeps. If I increase resolution around clipping point, that highly exaggerates the problem which could result in destruction of the amplifier. The coarse resolution still makes the point that in most cases 1 kHz power rating is optimistic. Last thing we want to do then is to pick the peak clipping point there as well.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
I'm listening right now at the level I listen when doing other stuff - in this case browsing ASR. I just turned on rew with my Umik1 and the dB meter. Average over a minute was about 68dB, Max (over about a minute) was 75dB
Averaged over a minute??? Peaks last milliseconds in music, not minutes. Averaging silence/low levels with the peaks generates meaningless numbers.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
Adding on, SPL meters are useless in determining peak power. Research done in this regard uses custom peak detection meters. See Fielder AES papers.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Why are you ignoring transient peaks in your analysis of needed power?
Because we don't choose a listening volume based on the transient peaks. The discussion earlier was about what level of THD+N is needed at lower power levels. If you want THD+N to be inaudible, it must be inaudible at the lowest average power you listen at. Instantaneous transients EG drum strikes - are not so relevant to this discussion - they just dicate what max power your amp needs to avoid clipping them.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Averaged over a minute??? Peaks last milliseconds in music, not minutes. Averaging silence/low levels with the peaks generates meaningless numbers.
Sorry - I was late to the party replying to a discussion from this morning about what power level we measure THD+N at (normally 5W) - IE AMP rated THD+N. Pointing out that we often listen at very low power levels of 250mW or less, and that can still be a reasonably loud volume.

This is not trying to muddy the waters further in the current bunfight about how to determine amp max power rating.


It relates to this subthread:

https://audiosciencereview.com/foru...ock-power-amplifier-review.39678/post-1398878
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,280
Location
Gold Coast, Queensland, Australia
A very long time ago, Yamaha proposed what they termed NDCR (Noise Distortion Clearance Range). Obviously, we have much lower noise and distortion these days, but the basic tenet is very valid. It aligns with my suggestion earlier of drawing a line from the intersection at 250mW across to the intersection with the rising THD at high power.

1670105999538.png


1670106037920.png


The above graphic(1kHz) was for an amplifier specified 65+65wpc@8R 1kHz 0.05%THD from 20Hz-20kHz.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
2,018
Likes
1,241
Location
Australia
A lot of criticisms of how I measure amps is already addressed. Let me summarize:

1. …


3. I do run power vs frequency tests. Here it is:
index.php


This is superior to picking one power output and running the sweep vs frequency. You see full data for complete power sweep, albeit at discrete frequencies. Notice how picking the "knee" helps here to show the limited power at 20 Hz.

To my ear, the flaw ^here^ Is assuming that a high distortion at 20 Hz is something I will hear.
I am usually much better at hearing distortion in the 500-3kHz range than I am at 20 Hz.

But using the knee of the hockey stick seems completely reasonable.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
To my ear, the flaw ^here^ Is assuming that a high distortion at 20 Hz is something I will hear.
You will absolutely hear 20 Hz full clipping of the amplifier, assuming the speaker can reproduce it. We are not talking about small amounts of distortion here. I have tested a lot of underpowered amps and distortion in bass frequencies is quite audible when amp heavily clips.
 

Sokel

Master Contributor
Joined
Sep 8, 2021
Messages
5,840
Likes
5,775
Because we don't choose a listening volume based on the transient peaks. The discussion earlier was about what level of THD+N is needed at lower power levels. If you want THD+N to be inaudible, it must be inaudible at the lowest average power you listen at. Instantaneous transients EG drum strikes - are not so relevant to this discussion - they just dicate what max power your amp needs to avoid clipping them.
I usually agree,but depends.
I'm listening to this right now (the blue highlighted one):




level.PNG


If you try to put an average at 70db you will surely grab the remote to lower it at some points.
And it's a single piano.
 

xaxxon

Active Member
Joined
Apr 3, 2022
Messages
244
Likes
277
That measurement is more of an exception than rule in that regard. Here is another example that shows it more clearly:

index.php


You can clearly see the knee is earlier at 20 Hz than others.

This graph has limited resolution on purpose. The amp is being stressed repeatedly with the full sweeps. If I increase resolution around clipping point, that highly exaggerates the problem which could result in destruction of the amplifier. The coarse resolution still makes the point that in most cases 1 kHz power rating is optimistic. Last thing we want to do then is to pick the peak clipping point there as well.
Could result in destruction of the amplifier??

Wouldn’t that be exceptionally good to know if it does if that’s a worry??
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,135
Likes
4,779
Location
Portland, OR, USA
You will absolutely hear 20 Hz full clipping of the amplifier, assuming the speaker can reproduce it. We are not talking about small amounts of distortion here. I have tested a lot of underpowered amps and distortion in bass frequencies is quite audible when amp heavily clips.
Hi Amir. Is the audibility due to the 20Hz fundamental distorting, or is it the higher order harmonics I'm hearing? I always wonder, since I have such a hard time sorting out what is going on under these conditions, and I haven't been bold enough to measure a system while it is subjected to large bass distortion.
 

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
I usually agree,but depends.
I'm listening to this right now (the blue highlighted one):




View attachment 247599

If you try to put an average at 70db you will surely grab the remote to lower it at some points.
And it's a single piano.
Sometimes there is a confusion between crescendo, and transient. The transients I'm talking about are sub 1/10th of a second - like drum strikes. Of course there is always a line to be drawn - if you get 20 high level drum strikes a second over a few seconds you may well be reaching for the volume. But that I would call a crescendo.
 

Sokel

Master Contributor
Joined
Sep 8, 2021
Messages
5,840
Likes
5,775
Sometimes there is a confusion between crescendo, and transient. The transients I'm talking about are sub 1/10th of a second - like drum strikes. Of course there is always a line to be drawn - if you get 20 high level drum strikes a second over a few seconds you may well be reaching for the volume. But that I would call a crescendo.
To comfortably and cleanly listen to this from at my MLP I have to put it in the 85db region (as it has really low parts),so it's not only about transients (which it also has).
At my 84-85db sensitivity speakers at 3 meters I think I need a fair anount of power,not extreme but fair.And it's a single piano,I have far more "difficult" plays.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,280
Location
Gold Coast, Queensland, Australia
Could result in destruction of the amplifier??

Wouldn’t that be exceptionally good to know if it does if that’s a worry??

Absolutely. But Amir is between a rock and a hard place. If he pushes the amplifiers to test their limits, some will fail/explode and then what does he do? Many are on loan from members and some are vintage, expensive or unrepairable.

But in my experience, well designed amplifiers are very hard to destroy. These Purifi and Hypex based amplifiers are virtually bulletproof and that means pushing them to their limits is not a problem. I'd be shorting them out, running them at full power for an hour etc. But that's me- I have a full repair lab...
 

xaxxon

Active Member
Joined
Apr 3, 2022
Messages
244
Likes
277
Sometimes there is a confusion between crescendo, and transient. The transients I'm talking about are sub 1/10th of a second - like drum strikes. Of course there is always a line to be drawn - if you get 20 high level drum strikes a second over a few seconds you may well be reaching for the volume. But that I would call a crescendo.
crescendo doesn't mean loud, it just means a smooth transition to louder than before.

You just mean "consistently (edit: or suddenly) loud"
 
Last edited:
Status
Not open for further replies.
Top Bottom