Fosi vs NCore/Purifi: When would there be an audible difference?

[charlielaub]

So, am I hearing that I'm unlikely to notice an audible difference unless I am using speakers that present a difficult load?

For the amplifiers like the Fosi ones and other TPA3255 types without PFFB, the frequency response deviations with a load that varies strongly from a resistive one only happen above 1k or 2kHz. This implies that a 3-way with a large impedance peak above this frequency range will experience a small frequency response hump in concert with the speaker's impedance magnitude. It's not really that the load is "difficult" it is that the load and amplifier interact to some degree. At lower frequencies there is no dependence on load impedance.

 

[boXem]

Got a reference or data to support this claim?

@pma made measurements some time ago. They are lying somewhere in the forum.

 

[charlielaub]

@pma made measurements some time ago. They are lying somewhere in the forum.

PMA has published a lot of measurements on amplifiers and IMD. What in particular leads you to make this claim? It doesn't really make sense to me, what you are saying... I am more interested to find out more about the topic than I am trying to say you are wrong, if that is how I am coming across.

Maybe @pma can comment?

 

[welwynnick]

Got a reference or data to support this claim?

I think he's referring to this:

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

NC252MP (class D) vs. A250W4R (classAB) burst measurements into 4ohm//2.2uF load As I have been doing, quite intensively, wide range of power amplifier measurements, I have decided to compare 2 different power amplifier design philosophies, class AB and class D. The amplifiers chosen have equal...

www.audiosciencereview.com

 

[boXem]

PMA has published a lot of measurements on amplifiers and IMD. What in particular leads you to make this claim? It doesn't really make sense to me, what you are saying... I am more interested to find out more about the topic than I am trying to say you are wrong, if that is how I am coming across.

Maybe @pma can comment?

I need some time to find it. He measured an Aiyima 07 into his speakers. Unfortunately I think that he doesn't post anymore here.

I think he's referring to this:

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

NC252MP (class D) vs. A250W4R (classAB) burst measurements into 4ohm//2.2uF load As I have been doing, quite intensively, wide range of power amplifier measurements, I have decided to compare 2 different power amplifier design philosophies, class AB and class D. The amplifiers chosen have equal...

www.audiosciencereview.com

Nope, this thread is quite misleading to my eyes.

 

[MaxwellsEq]

So, am I hearing that I'm unlikely to notice an audible difference unless I am using speakers that present a difficult load?

Given your question is "Fosi vs NCore/Purifi: When would there be an audible difference?", I think most of us can't be certain that you would hear a difference. Certainly I would expect the NCoreX / Purifi amplifiers to measure more consistently than chip-amps across a very wide range of different speaker types generating large amounts of relatively undistorted sound. Whether you would reliable and unsighted be able to hear a difference if all the amps are operating inside their safe design window is a different matter. I'm sorry - this may not help!

If my approach helps, it's to buy the best measuring electronics that you can afford, but focus on a flat frequency/amplitude regardless of load and prioritise noise behaviour over distortion. There are many tests you can try and most people who take them struggle to hear harmonic distortion. But noise is more detectable than distortion (and not just hiss) and because it acts as a limit to information retrieval and can interact with the audio. So, if you base your decision on SINAD, pay more attention to the noise than distortion.

 

[mfaughn]

Given your question is "Fosi vs NCore/Purifi: When would there be an audible difference?", I think most of us can't be certain that you would hear a difference.

Thanks for the rather helpful reply. I think my question might have been better stated. There is certainly a region where some people are going to hear a difference between two amplifiers and others won't. There is also a region in which you wouldn't reasonably expect anyone* to hear differences, where the differences are simply expected to be beyond the bounds of the human ear and brain to discern differences (and I'm sure some people would argue here but let's assume we're basing this on published science). (*Yes, all humans are different. Assume then that we're talking about humans that fall within two or three standard deviations from normal.)
And of course this will be use case dependent. How loud do you want to listen, how sensitive are your speakers, do your speakers present a challenging load, what is your room like, how much ambient noise is there in your listening space, etc., etc.. There are some situations in which even a golden-eared engineer might be hard pressed to tell the difference between a Lepai 2020 and a Purifi while in other circumstances the differences would be stark. So my question might have been better stated (but probably still imperfectly) as: Under what circumstances would differences between these Fosi amps and an NCore/Purifi amplifier be expected to be above or below the threshold of what is generally accepted in the scientific literature as being below the threshold of human perception (again, humans within 2-3 SD). The answer could even be as simple as, "There are a lot of normal situations where there'd be a difference," or "there aren't many situations where there'd be a difference." I know this is still not an easy question to answer. We can take as many objective measurements as we like but in the end we as listeners always have a subjective experience. Making a little progress, where possible, towards correlating measurements with likely subjective experience is helpful to some of us

 

[MaxwellsEq]

Most engineers would probably answer "there aren't many situations where there'd be a difference.". So you have two probabilities - the distribution of engineers and the distribution of people's hearing. Regarding the latter, most of the differences due to load-dependant variance are at the top end where there is very little content and most people over 40 are unlikely to be able to hear the impacted frequencies.

 

[MaxwellsEq]

However, there are some engineers that believe that some power amplifier tests should be more challenging. I'm partially in that camp. I'd like to see pulse-recovery tests undertaken.

Also, some engineers feel confident that you can learn everything you need to know about load dependency from 2, 4, 8 resistive load comparisons, but I feel reactive loads which have largeish phase angles may show power supply designs that struggle to handle current and voltage delivery out of phase.

 

[mfaughn]

Most engineers would probably answer "there aren't many situations where there'd be a difference.". So you have two probabilities - the distribution of engineers and the distribution of people's hearing. Regarding the latter, most of the differences due to load-dependant variance are at the top end where there is very little content and most people over 40 are unlikely to be able to hear the impacted frequencies.

So, this brings me to "I'd probably be fine with Fosi" while at the same time "just avoid further / future questions by going to NCoreX/Purifi now". I'm not sure I'm a whole lot closer to my decision...but that is entirely my fault . Thanks for help all the same (everybody!).

 

[MaxwellsEq]

I have a Purifi amplifier and am pleased with it. I have no experience of Fosi devices, but I've been impressed with their objective results given their cost and size.

 

[Chazz6]

If you can imagine that you might want different speakers in a few years, better to get a fine amplifier like a Boxem.

 

[sarieri]

I don’t remember which thread exactly it is but pma tested Ncore and TPA3255 before. To the best of my memory, ncore is pretty bad in terms of handling capacitive loads not until they upgraded their feedback in their NcoreX and purifi designs. Tpa 3255 is much better than offerings like nc252/502mp unless I’m interpreting something wrong. Whether it’s audible is totally another question though.

 

[boXem]

interpreting something wrong

Indeed. The capacitive load @pma used was completely surrealist and the NC252MP was submitted to a load below it's minimum specified load.

 

[Sokel]

This thread?

AIYIMA A07 TPA3255 Amplifier Measurements and Review - LM4562 (and OPA2134) option

AIYIMA A07 TPA3255 Amplifier Measurements LM4562 option I received my A07 yesterday, on January 6, 2021, shipped from Italy. Purchased at Amazon.de and as stated on their website, it arrived without a power supply. I am testing the A07 with 2 different power supplies. One of them is a linear...

www.audiosciencereview.com

Yep,dependency was serious.

 

[charlielaub]

I don’t remember which thread exactly it is but pma tested Ncore and TPA3255 before. To the best of my memory, ncore is pretty bad in terms of handling capacitive loads not until they upgraded their feedback in their NcoreX and purifi designs. Tpa 3255 is much better than offerings like nc252/502mp unless I’m interpreting something wrong. Whether it’s audible is totally another question though.

Ah yes, I remember that rather "stupid" test by PMA. He was not just using a reactive load with some capacitive region in it, the load was a capacitor! This has impedance with a magnitude that falls to zero at high frequencies, e.g. a short circuit No consumer amplifier is designed for a load like that. In all of speaker history there has been one model of loudspeaker that had such a "stupid" impedance, and honestly it was the fault of the designer/manufacturer to put out such a ridiculous product regardless of how it performed.

I would ignore PMAs testing on this sort of thing. It was not testing the amps within their intended performance envelope. He only wanted to gloat that his own amplifier was better under these conditions, but these conditions are never found so it was pointless and only serve to mislead readers such as yourself.

 

[charlielaub]

This thread?

AIYIMA A07 TPA3255 Amplifier Measurements and Review - LM4562 (and OPA2134) option

AIYIMA A07 TPA3255 Amplifier Measurements LM4562 option I received my A07 yesterday, on January 6, 2021, shipped from Italy. Purchased at Amazon.de and as stated on their website, it arrived without a power supply. I am testing the A07 with 2 different power supplies. One of them is a linear...

www.audiosciencereview.com

Yep,dependency was serious.

I don't think this is the thread that we are looking for. But it is a good example of how the output network of a cheap TPA3255 amp interacts with the load impedance, causing the frequency response to vary above 2kHz or so. But this is happening because of the particular output network (external thru-hole components) and not because of the TPA3255 itself. A different TPA3255 based amp with an output filter network that is better designed and that uses PFFB does not do this. The Ayima amp is a very cheap implementation, and you get what you pay for (in this case).

 

[Skelethor]

I Don't know if Buckeye is any better, but I found the speaker wire binding posts on the Fosi V3 to be useless unless you have banana plugs.

So they're not useless, as you said just get banana plugs.
Problem solved!

 

[Ron Texas]

So they're not useless, as you said just get banana plugs.
Problem solved!

That violates one of my principles which is anything new out of the box should not be a DIY project unless it was intended to be from the start. Just because it can be fixed doesn't change the fact that is broken.

 

[Kimbrough Xu]

If your listening distance is just 3 meters and then it depends on your speakers whether you need a more power amp.