Cheap and cheerful class D comparison (CCCDC)

[RandomEar]

This is what the title says: A quick comparison of cheap-ish class D amps. Rough inclusion criteria:

• Is a class D speaker amplifier
• Does not include a DAC
• Is below 350 $ for a stereo setup

The contestants

Manufacturer

Model

Price range

Power

SINAD

PFFB

Load depend.

Balanced

Vol. control

Sub out

Low pass

High pass

Remarks

-

-

USD, 01/26

4Ω, 1% THD

dB @ 5W, 4Ω, 1kHz

-

dB @ 20kHz, 4Ω vs 8Ω

-

-

-

-

-

-

3e Audio

A5

200-250

133

105

✔

<0.2

✔

✔

✘

✘

✘

Exceptional channel sep.

3e Audio

A7

300-375

243

102

✔

<0.2

✔

✔

✘

✘

✘

Aiyima

A07

~75

77

83

✘

~1.5

✘

✔

✘

✘

✘

Aiyima

A07 Max

~100

96

86

✘

1.5

✘

✔

✘

✘

✘

Aiyima

A20

200-250

175

94

✔

1.5

✔

✔

✔

?

✔

FR up ≥0.8 dB @ 20 kHz, bad PFFB

Aiyima

A70

~200

178

95

✔

0.25

✔

✔

✔

✔

✘

Aiyima

A70 Mono

100-150

220

94

✔

1.0

✔

✔

✔

✔

✘

„Grassy“ multitone noise < 1 kHz

Douk Audio

A5

~100

75

88

✔

0.8

✘

✔

✘

✘

✔

Distortion overtakes noise around 1 W

Fosi Audio

V3 mono

300-350

244

100

✔

0.7

✔

✘

✘

✘

✘

About 50 W down at 20 Hz

Fosi Audio

ZA3

150-200

146

88

✘

1.5

✔

✔

✔

✔

✘

Mono mode with 210 W & lower SINAD

Topping

Mini 300

100-150

120

105

✔

<0.2

✔

✔

✘

✘

✘

About 40 W down at 20 Hz, exceptionally quiet power on/off

Topping

PA5II

~200

120

105

✔

<0.2

✔

✔

✘

✘

✘

About 40 W down at 20 Hz, exceptionally quiet power on/off

The plots

I've split them into two groups for the THD+N vs. power plots. The data for all plots comes from @amirm's reviews. Group 1, "the average guys" (A07 and PA5II serve as references):

Group 2, "the high performers" (A07 and PA5II repeated as references):

It looks like the 3e Audio A5, Topping PA5II and Mini 300 might have reached the lower limit of what can be achieved noise-wise with their TI TPA32XX amplifier chips. Compared to what you could get a couple of years ago around the 100 $ price point with the Aiyima A07, developers have pushed the noise at low power levels down an order of magnitude, which is pretty impressive. Power-wise, not much has moved at that price point. However, you can now get cose to 250 W of clean power if you're willing to spend a bit over 300 $, which is also pretty nice.


The conclusion

I'm currently just keeping track of what comes to the market. At some point, I might upgrade to one of these from my current AO200 setup, mostly to get slightly more peak power. Stand out performers in my opinion:

• Overall
  3e Audio A7 - Low noise floor and good power, simple case. Offers clipping indicators, 12 V trigger and volume bypass.
• For mono block lovers
  Fosi V3 Mono - Low noise floor, good power, clean design, but slightly weaker bass power levels. No volume controls.
• Top performance
  3e Audio A5 - Extremely low noise floor, decent power and state of the art channel separation in a simple case. Offers clipping indicators, 12 V trigger and volume bypass. Not cheap.
• Bang for buck
  Topping Mini 300 - Extremely low noise floor and decent power for just over 100$. 12 V trigger but slightly weaker bass power levels. Probably ideal for desktop and near field setups when combined with a sub. The Mini 300 seems to be PA5II hardware in a cheaper case for ~80$ less.

 

[dguillor]

The Wiim Vibelink $299 competes well in this group. It is closest to the Aiyima A70, and it costs a little more, but it has an internal power supply, which is preferable to me.

 

[Jarrett]

Thanks! I think the 120 W spec is for the PA5 II Plus model, btw

 

[RandomEar]

Thanks! I think the 120 W spec is for the PA5 II Plus model, btw

The value in the table isn't the spec, but the actual performance in the ASR review. The PA5II was tested with a 38 V 3.2 A supply, which was the non-plus model. According to the Topping website, they might now ship with a 3.0 A supply (or their website is wrong), which would reduce theoretical peak power to 114 W.

 

[harkpabst]

There's no 12 V trigger input on the production versions of the 3E AUDIO A7 and A5. Amir reviewed early samples, which still had the trigger input.

I agree with @dguillor that the WiiM Vibelink Amp would have been an interesting addition, even though it violates the "no DAC" criterion. It still fits into the accepted price range.

 

[Joe Smith]

It would be good to add in the Fosi V3 stereo model (already reviewed) and the BT20a MAX (when it is reviewed).

That 3e A5 measurement is pretty great.

Nice summary, I think this will be helpful to folks, as there have been so many new models of late.

 

[mhardy6647]

Just a sidebar: The title can be condensed to C^3DC and/or simplified mathematically to C^4D.


PS I am, personally and professionally, a big fan of metanalysis of data. I guess AI will make/has made this much easier for those of us who are into such things.

 

[john61ct]

I realize the Wiim Amp Ultra is a different beast and costlier

But the inclusion of that DSP functionality for room correction, bass management, loudness contour and PEQ in general

even as just a power amp, I believe close to A7 specs (PFFB?)

makes for good value (to me) even setting aside the streamer functions. Certainly a lot of flexibility when repurposed later on...

However apparently limited to a "downstream node" for multi-channel in the same room, not great as a "master upstream"

2xWAU used as crossovers

I am planning a stereo system with "full-range" OG LS50s main pair, HPF'd somewhere between 170-300Hz stereo MBM pair, bandpassed between LPF blending with the above, and an HPF between 70-90Hz then mono sub(s) below that. My idea is to use two WAU for handling the crossovers, the "top one"...

forum.wiimhome.com

 

[DrTom]

Daydreaming slightly... Would any of these be suitable for difficult loads such as (potentially stacked) ESL 57 going down to 0.5ohm at high frequencies?

 

[Roland68]

I realize the Wiim Amp Ultra is a different beast and costlier

But the inclusion of that DSP functionality for room correction, bass management, loudness contour and PEQ in general

even as just a power amp, I believe close to A7 specs (PFFB?)

makes for good value (to me) even setting aside the streamer functions. Certainly a lot of flexibility when repurposed later on...

However apparently limited to a "downstream node" for multi-channel in the same room, not great as a "master upstream"

2xWAU used as crossovers

I am planning a stereo system with "full-range" OG LS50s main pair, HPF'd somewhere between 170-300Hz stereo MBM pair, bandpassed between LPF blending with the above, and an HPF between 70-90Hz then mono sub(s) below that. My idea is to use two WAU for handling the crossovers, the "top one"...

forum.wiimhome.com

Before attempting something like this, you should look into professional solutions like Acourate or similar software. Using two WiiMs as DSPs, on the other hand, is really just a toy, quite apart from the timing problems.

 

[peng]

Daydreaming slightly... Would any of these be suitable for difficult loads such as (potentially stacked) ESL 57 going down to 0.5ohm at high frequencies?

You would have to try it to find out, but based on specs/measurements I bet the 3e A7, being PFFB and has more current capability (relatively) should have better luck with the ESL 57. Obviously it depends on how high you need to crank the volume up, capacitive reactive or not, the amp and the load still obey Ohm's law, so current = Voltage/Impedance, if the voltage is low enough, say if below 3 V, then current would be below 6 A even right at the 0.5 ohm minimum point. Music signal won't stay at one frequency so that 0.5 ohm point won't result in high current for long, so again, it all depends. In terms of stability concern, I would think PFFB should help, to a point..

 

[DrTom]

You would have to try it to find out, but based on specs/measurements I bet the 3e A7, being PFFB and has more current capability (relatively) should have better luck with the ESL 57. Obviously it depends on how high you need to crank the volume up, capacitive reactive or not, the amp and the load still obey Ohm's law, so current = Voltage/Impedance, if the voltage is low enough, say if below 3 V, then current would be below 6 A even right at the 0.5 ohm minimum point. Music signal won't stay at one frequency so that 0.5 ohm point won't result in high current for long, so again, it all depends. In terms of stability concern, I would think PFFB should help, to a point..

Thanks. My electronics knowledge is pretty minimal but I think I understood some of that!

I'm quite tempted to try an A7 or A5. Given that the ESL 57 actually don't like too much power I think the A5 might suffice... as far as I can see it is similarly stable at 2ohm to the A7? Any idea if there are likely to be safety limiters below 2ohm?

I don't need huge volumes, at the moment I'm feeding 2vrms into a 6wpc SET amp and probably only using 2wpc.... looking at the gain of 26db on the A5, am I right in thinking that I could feed it 1vrms and not exceed 20vrms out (I've seen 23.4vrms quoted as a safe limit for ESL 57 before arcing)? I wouldn't expect to use anywhere near that much power, but 'just in case'.

 

[mhardy6647]

Given that the ESL 57 actually don't like too much power I think the A5 might suffice...

Yes. Do be careful. They are easy to arc and tedious to rebuild.

FWIW, empirically, a McIntosh MC-225 (push-pull 7591A) mates exceptionally well with the ESL-57.
(nope I cannot afford, i.e., justify the cost of, one either )

I, also, usually use either a SET amp (2A3, in my case, which is a bit too anemic for best results) or pp EL84 (usually an EICO HF81) with mine (again, strictly FWIW).

 

[peng]

I don't need huge volumes, at the moment I'm feeding 2vrms into a 6wpc SET amp and probably only using 2wpc....

what's the gain od that SET amp?

looking at the gain of 26db on the A5, am I right in thinking that I could feed it 1vrms and not exceed 20vrms out

That is about right, about 20 v, I hope you don't need more than 0.5 V peak.

 

[DrTom]

what's the gain od that SET amp?


That is about right, about 20 v, I hope you don't need more than 0.5 V peak.

I have no idea on the gain on the current amp, it's a Reisong A12 with 'Skunkie' mods.

Reisong A12 Modifications

Modifying the Boyuurange / Reisong A10/A12 amp.

www.skunkiedesigns.com

 

[peng]

I have no idea on the gain on the current amp, it's a Reisong A12 with 'Skunkie' mods.

Reisong A12 Modifications

Modifying the Boyuurange / Reisong A10/A12 amp.

www.skunkiedesigns.com

Thank you, I am just trying to understand what you meant by:

I don't need huge volumes, at the moment I'm feeding 2vrms into a 6wpc SET amp and probably only using 2wpc....

Is the 2 V just the max output., from your input device (example: CD player) connected to the Reisong A12's input ?

 

[DrTom]

Thank you, I am just trying to understand what you meant by:

Is the 2 V just the max output., from your input device (example: CD player) connected to the Reisong A12's input ?

Yes I think so; A12 being fed from a Wiim Ultra, line out level set to 2vrms. The A12 volume pot is usually around 1/4-1/3rd of the way around at my usual listening levels.

 

[peng]

Yes I think so; A12 being fed from a Wiim Ultra, line out level set to 2vrms. The A12 volume pot is usually around 1/4-1/3rd of the way around at my usual listening levels.

That would make sense, so the A12 is attenuating the Wiim output signal to less than, much less than 0.5 Vrms for sure. That means any of those class D amp should work for you, but still, should stick with those with PFFB.

 

[harkpabst]

The Quad ESL not only has a very low impedance in the highs, this low impedance is also almost purely capacitive. Personally I don't think that any TPA3255 based class-D amplifier is a perfect match for a single ESL, let alone stacked ones. If at all, I would only consider designs with PFFB and PBTL (one chip per channel). But even then the output impedance rises sharply somewhere above 20 kHz due to the LC output filter becoming dominant. Even if we don't directly hear those high frequencies, but we may absolutely hear intermodulation products and instability artefacts.

According to the very informative website Https://quad-hifi.de the ESL was equipped with a simple protection circuit from ~ S/N 16800 on. It protects the HF module, only.

The obvious (solid state) amplifier choice for the Quad ESL was the original Quad 303. The new 303 shout work well, too. Purifi or Hypex based class-D amps are certainly more suited than TPA based amps.

 

[mhardy6647]

The Quad ESL not only has a very low impedance in the highs, this low impedance is also almost purely capacitive.

They essentially are capacitors.