Review and Measurements of Paradigm PW Amp

[audioBliss]

Imo room correction is only useful under the schröder frequency. A good amp is useful for many different variables in the sound quality. That said I use Dirac Live myself <260Hz which makes a huge impact to sound quality - probably more than any amp could but that said a good amp is important too. Quite hard to compare in the end. In this day and age it's not super difficult to find transparent electronics at a reasonable price thanks to sites like this. Basically just get both

 

[estuardo4]

It's such a bummer this amp, that the thread almost has no discussion. Usually when a new amp is tested, I use to expect minimum 100 responses in the first two days. With this PW Amp, not even the usual suspects has responded.

Amir, thank you for your hard work. Do you have in your pipeline a review or two of sub $500 amps?

 

[Vnowinski]

Yes, that is also here from the same owner.

Very much looking forward to these measurements. I have a PW Link, using it (via its optical in and optical out, with a couple of SPDIF/toslink converters) as an ARC processor, in combination with a networked streamer (Pi + Allo digione) and the Khadas tone board. Would love to see its performance (digital and A>D for kicks) using the digital output with processing active...!

 

[amirm]

Do you have in your pipeline a review or two of sub $500 amps?

Yes, I have a few more to test although I am not sure they all fall under $500 bucket.

 

[deafenears]

I’ve been designing and building audio components for a few years now to help develop my analog electronics skill and I can tell you that you’ve been utterly misinformed.

Building a 120 dB SINAD headphone amplifier is not trivial, but is not that hard either—even for a relative beginner like me. Designing a 120 dB SINAD power amplifier on the other hand requires true mastery of the skill, at least, in my humble opinion.

Maybe @tomchr can comment...

 

[BillG]

There is a rule among software engineers that says to never trust a hardware company with software. DTS exemplifies this. Who thought they should trust this company with building a streaming solution?

One would think that given the company's experience in building multichannel audio technologies, which would require some serious software engineering, that they'd understand how to create good user-facing software. However, it's readily apparent that they don't.

The Android and Windows versions of their applications don't even conform to the UI design guidelines of their respective platforms. Hence average users, who generally aren't that technologically inclined to begin with, are left feeling dissatisfied with the experience...

 

[Tks]

I’ve been designing and building audio components for a few years now to help develop my analog electronics skill and I can tell you that you’ve been utterly misinformed.

Building a 120 dB SINAD headphone amplifier is not trivial, but is not that hard either—even for a relative beginner like me. Designing a 120 dB SINAD power amplifier on the other hand requires true mastery of the skill, at least, in my humble opinion.

Oh so power amps definitively are more difficult to build to the same spec with respect to distortion levels (when general use-case is translated properly between each)?

The reason the person said this to me, and his reason was because headphones amps have to support far more resistance ranges. Does that hold any truth at least with respect to that persons reasoning why headphone amps are more complex?

Oh and, one other thing I am confused about, is the concept of pre-amps, and why they aren't required with headphone amps, or why they aren't a part of every power-amp (like why is the paradigm: No need to provide a pre-amp for power amps, go buy your own... but with headphone amps, no need to have any to begin with?).

Sorry, I'm new to speakers in general, this is perplexing from a standards perspective.

 

[snip3r77]

And the search for a $200 amp you want continues...

You beat me to it....may I add a digital amp

 

[RayDunzl]

Oh and, one other thing I am confused about, is the concept of pre-amps

Traditionally, a preamp has:

Inputs for several devices, and a way to switch between them
Volume control, with amplification for weak sources

May have:

Tone control
Phono input
Remote Control
Output for a second system
Input trims (adjust volume for different inputs)
Headphone Output
Tape Out (recording) / Input (Monitor)

When these functions are provided in the same box with a power amplifier, it becomes an "Integrated Amplifier".

Add an AM/FM Tuner, it becomes a "Receiver".

 

[folzag]

Everyone I've heard speak tells me designing a headphone amp is far more of a complex thing than designing a power amp.

Why are all power amps aside from a handful seemingly pure bin-tier ?

My guess is it's sampling bias. High quality power amps are heavy so that makes them considerably more expensive to ship in for testing.

 

[TimW]

Oh and, one other thing I am confused about, is the concept of pre-amps, and why they aren't required with headphone amps

Headphone amps usually have the necessary preamp features built in. As in volume control and maybe even input switching. That's why you don't typically need a preamp with a headphone amplifier.

In the speaker world people often prefer to have separate components often referred to as "separates". Traditionally this term refers to having a separate preamp and power amp(s). @RayDunzl's explanation of integrated amps and receivers is spot on. Imho a headphone amplifier is basically a simple low power integrated amplifier.

 

[Tks]

Headphone amps usually have the necessary preamp features built in. As in volume control and maybe even input switching. That's why you don't typically need a preamp with a headphone amplifier.

In the speaker world people often prefer to have separate components often referred to as "separates". Traditionally this term refers to having a separate preamp and power amp(s). @RayDunzl's explanation of integrated amps and receivers is spot on. Imho a headphone amplifier is basically a simple low power integrated amplifier.

Ah, now I understand. But the last question if you would aid me:

Why? Is the market filled with more people that demand separates, than not in the speaker world? Also, why aren't there headphone power amps, or is this yet again a market demand sort of ordeal?

Or am I misreading the general speaker market, and this "separates" group is relinquished only to the high-end where a sort of mysticism exists about having separates as much as possible in the chains?

 

[deafenears]

This should all go in a "separate" thread. I'm sure it'll be useful to future travelers

 

[HamNRye]

First, thanks to Amir for measuring this. As a happy owner of the PW/Amp I have conflicted feelings about the results.

Measured as a power amp, I am pretty sure that Amir's testing was fair and accurate. In fact, the few published mfg specs of the PW/Amp (and Identical Martin Logan Forte) are largely in alignment.

...

Compared to recently tested power amps...the results are mediocre...significantly better than a few, significantly worse than others...but they seem in alignment with Paradigms published specs and assumed design goals. That is, Paradigm must have felt that this kind of performance would not result in noticeable distortion for the intended audience. Mine sounds just fine to me. I have not done objective blind listening tests with ARC turned off.

So why am I satisfied with mediocrity? Lets focus on the PW/Amp rather than my worldview

The PW/Amp (and PW/Link I also own) have had a greater positive impact on my two audio systems sound than any components I have changed, besides upgrading speakers. My experience corroborates Amir's statement that "No doubt room EQ will make a bigger difference than any of these measurements do in the sound of the amplifier". In one of my setups the impact of ARC is large, in the other, a very difficult desktop setup, it is phenomenal.

...

The monitors have dip switches to adjust HF/MF/LF and Low Pass...and the red lines are the best I could do with careful listening. The Green is the Corrected ARC response. The change is not subtle. In fact, my wife, who was vaccumming upstairs suddenly...never mind
So while I am pleased that Amir took the time to measure the PW/amp and believe his testing was accurate; I am also convinced that what you measure is equally important, and the real benefit of this kind of device on sound quality was referenced, but not measured.

Thank you for this. Sometimes I have to remind myself that it's really about the music, what fits my needs and the point of diminishing returns. Yes, the PW Amp isn't perfect, but room correction in a small form factor is more important to me. And I'm not sure if there's anything else worthwhile that's much better at the moment. Amir's own reviews of the MiniDSP products aren't glowing either. I'm hoping that NAD makes a version of their D 3045 digital amp with Dirac built in. Or Paradigm updates the PW Amp with digital inputs and a cleaner amp.

 

[tomchr]

Building a 120 dB SINAD headphone amplifier is not trivial, but is not that hard either—even for a relative beginner like me. Designing a 120 dB SINAD power amplifier on the other hand requires true mastery of the skill, at least, in my humble opinion.

Having built both, I'd say getting to 120 dB SNR is harder in a power amp as they tend to be noisier. At line level (which is basically what a headphone operates at) impedances can be lower, idle currents lower, etc. which all leads to lower noise. Both require attention to detail and good design and layout skills, though.

Tom

 

[Hephaestus]

Thanks Amir for the review. What a bummer measurement wise, had high hopes for this unit...

 

[Tks]

Having built both, I'd say getting to 120 dB SNR is harder in a power amp as they tend to be noisier. At line level (which is basically what a headphone operates at) impedances can be lower, idle currents lower, etc. which all leads to lower noise. Both require attention to detail and good design and layout skills, though.

Tom

So simply because of the power output and lower impedance of speakers makes power amps more complex to make?

 

[GGroch]

Thank you for this. Sometimes I have to remind myself that it's really about the music, what fits my needs and the point of diminishing returns......Amir's own reviews of the MiniDSP products aren't glowing either.....

That's true. The MiniDSP 2x4 HD measurements showed a 15-18dB worse SNAD than a Topping D50. The MiniDSP SHD measurements were much better, but both models measured significantly better than the PW/Amp. As was pointed out in the MiniDSP 2x4 HD review, some level of degradation is to be expected due to the DSP hardware/software. If Amir measures the PW/Link that would tell us more about whether most of the distortion & noise are caused by the room EQ DSP or the power amp section in the PW/Amp.

In both Paradigm PW models there is no way to by-pass the DSP circuitry (which I understand involves at least three AKM A to D and D to A chips plus a Marvell processor for the DSP manipulations). Any analog source runs through them. The ARC software allows you to remove computed equalization curves to return to a flat response, but you are still going through the AKM/Marvell/ARC stuff. So, one should expect increased distortion at least at these lower price points. The question becomes whether the room EQ benefits are a worthwhile trade off, or, as HamNRye asks, whether a better performing RoomEQ/Amp is available at at reasonable cost.

My google searches find almost no measurements of these MiniDSP or ARC type products. They are clearly much more complicated to test than traditional DACs/Amps because so much software is involved. BTW, on the PW/Amp, in addition to the Play-Fi firmware updates, there is also an ARC firmware update that downloads the 1st time you run ARC software. I do not know whether this impacts just ARC or overall performance as well.

 

[tomchr]

So simply because of the power output and lower impedance of speakers makes power amps more complex to make?

I should perhaps have been more specific. The lower impedances in the amplifier circuit of a headphone amp makes it easier to make a low-noise headphone amp.

If you took a headphone amp and increased the supply voltage to support the ±40 V swing needed for a 100 W (8 Ω) power amp, many of the resistors in the circuit would need to be power resistors. Those are not associated with precision (higher excess noise, higher temperature coefficient, unknown (probably high) voltage coefficient, etc.). You'd probably also find that the power dissipated in the various active devices would exceed the ratings of those devices. The currents in the active devices would be higher, which leads to higher noise.
Solution: Increase the resistances to lower the currents and lower the power dissipation. Now you can use precision resistors, but because the resistances are higher, you'll get higher thermal noise from the resistors. It's a tradeoff between noise caused by current flow (shot noise) and noise caused by random movements of electrons in resistors (thermal noise) all while keeping the various devices operating within their limits.

Basically, noise optimization is like squeezing a balloon. You try to minimize the volume of the balloon by squeezing it, but if you push too hard in one place, the balloon just pops out in another place. It just so happens that the "noise balloon" in a headphone amp tends to be smaller than that of a power amp.

Tom

 

[HamNRye]

That's true. The MiniDSP 2x4 HD measurements showed a 15-18dB worse SNAD than a Topping D50. The MiniDSP SHD measurements were much better, but both models measured significantly better than the PW/Amp. As was pointed out in the MiniDSP 2x4 HD review, some level of degradation is to be expected due to the DSP hardware/software. If Amir measures the PW/Link that would tell us more about whether most of the distortion & noise are caused by the room EQ DSP or the power amp section in the PW/Amp.

In both Paradigm PW models there is no way to by-pass the DSP circuitry (which I understand involves at least three AKM A to D and D to A chips plus a Marvell processor for the DSP manipulations). Any analog source runs through them. The ARC software allows you to remove computed equalization curves to return to a flat response, but you are still going through the AKM/Marvell/ARC stuff. So, one should expect increased distortion at least at these lower price points. The question becomes whether the room EQ benefits are a worthwhile trade off, or, as HamNRye asks, whether a better performing RoomEQ/Amp is available at at reasonable cost.

My google searches find almost no measurements of these MiniDSP or ARC type products. They are clearly much more complicated to test than traditional DACs/Amps because so much software is involved. BTW, on the PW/Amp, in addition to the Play-Fi firmware updates, there is also an ARC firmware update that downloads the 1st time you run ARC software. I do not know whether this impacts just ARC or overall performance as well.

I believe holding down the mute button for a few seconds turns off ARC on the PW Amp. Perhaps this bypasses ARC and the A/D conversion? The sound cuts out when I hold down the button and when the sound comes back ARC should no longer be active. Or perhaps this simply flattens the ARC curve and the DSP is still engaged.

Thank you GGroch for pointing out that both MiniDSP units are still much better than the PW Amp. I believe you're referring to the high noise floor on the PW Amp? I am still very much new to this site (which is amazing!) and learning to read these reviews/measurements. I may just pick up the MiniDSP 2x4 HD now to give it a try.