• 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!

3e Audio A5 Stereo Amplifier Review

Rate this amplifier:

  • 1. Poor (headless panther)

    Votes: 5 1.7%
  • 2. Not terrible (postman panther)

    Votes: 11 3.8%
  • 3. Fine (happy panther)

    Votes: 49 17.0%
  • 4. Great (golfing panther)

    Votes: 224 77.5%

  • Total voters
    289
Where are you pulling these figures from, precisely? Looking at the reviews I'm not seeing anything over 90-ish dB SINAD at any power level for the 3800H and 4800H.

Also keep in mind that Amir has given the Denons a bit of a handicap here by using "Pure Direct" mode for the measurements. In practice, almost nobody will be using them that way since that disables all processing. When not in that mode, SINAD will decrease a fair amount.
I realized I had confused SINAD measurements with pure signal-to-noise ratio. SINAD is generally about 5 dB lower across the board for all AVRs. Interestingly, Denon AVRs seem to have a gain of around 26–29 dB, which is slightly lower than I expected.

That said, my main point still stands. Achieving 85 dB SINAD at 5 watts (worst case of all Denons) of the integrated amps in AVRs is just really good. This level of performance is great —essentially completely transparent for any listening purposes.

Also, the "pure direct" argument applies to DACs as well. Once you introduce preprocessing like EQ, loudness curves, or room correction, SINAD will likely drop by an amount proportional to the corrections applied (in dB). This is worth considering for anyone prioritizing transparency over processing features.
 
That said, my main point still stands. Achieving 85 dB SINAD at 5 watts (worst case of all Denons) of the integrated amps in AVRs is just really good. This level of performance is great —essentially completely transparent for any listening purposes.
No disagreement from me here.
Also, the "pure direct" argument applies to DACs as well. Once you introduce preprocessing like EQ, loudness curves, or room correction, SINAD will likely drop by an amount proportional to the corrections applied (in dB). This is worth considering for anyone prioritizing transparency over processing features.
I'm having trouble recalling/finding an example, but I do recall seeing reviews of devices with no (or marginal) negative impact to their SINAD when doing DSP functions versus without.
 
Manufacturer impedance rating are almost always completely wrong. Vast majority of speakers I test have a minimum impedance between 3.5 and 4.5 ohm. None have ever been 8 ohm.
Yes, I have no statistics to back my opinion but I feel like this misbehavior only became worse over the last 40 years. Probably at the same rate that speaker efficiency has decreased. ;) Back then many speakers were tagged as being "4 to 8 ohm". While this is nonsense for a nominal value, ironically it might have been easier for the casual buyer to pick a matching amp, since people always try to match numbers in some way (whether it makes sense or not) and chances were that amps were specified "4 to 8 ohm" as well. :p

The answer is crest factor. The average power of music is much lower than during its peaks. The wattage of the transformer in a traditional AVR will also be much lower than its 9x 180 W or whatever advertised power.
@amirm measures power output using a 1 kHz sine wave, both channels driven. So, 2 x 122 Watts output still means that > 244 Watts have to be fed by the power supply for as long as the measurement lasts. One single frequency, so the crest factor should not play a role here.
 
Back then many speakers were tagged as being "4 to 8 ohm".
Not IIRC -- and I was selling audio "Back then". Some receivers and amps "were tagged" like that to indicate they could be used with speakers in that range of nominal impedances, but speakers from legit companies were always "tagged" with a single number.
 
Not IIRC -- and I was selling audio "Back then". Some receivers and amps "were tagged" like that to indicate they could be used with speakers in that range of nominal impedances, but speakers from legit companies were always "tagged" with a single number.
Tricky question ... what's a legit company? This mainly just proofs that our memory differs. How many examples would I have to provide to make you reconsider your statement? ;)

From the top of my head I could remember the Marantz P 1230, P 1030 and P 830 series from the early 1980s. Here's a scan from a (German) brochure I still have:
Marantz P X30 series.png


Don't get me wrong, I'm not saying that this has been the norm nor am I advocating its use. But it did happen. More often towards the lower price region.

How do I avoid this post getting totally off topic? I probably can't. :p But there's no question that the A5 can drive pretty much any dynamic loudspeaker, irrespective of its nominal impedance.

PS:
Just found this picture of the actual terminal and it states "4 ohm". Maybe just a matter of badly written marketing material but I have seen other examples back then.
Marantz P 1230.jpg
 
Last edited:
@amirm measures power output using a 1 kHz sine wave, both channels driven. So, 2 x 122 Watts output still means that > 244 Watts have to be fed by the power supply for as long as the measurement lasts. One single frequency, so the crest factor should not play a role here.
True, but that signal is run only for 1 second or so, my understanding is.
 
While this is nonsense for a nominal value
Yes, but a loudspeaker is not a resistor with a fixed value, but a complex, varying resistance depending on the frequency.
It has never been the right way to stick a fixed value on a changing complex structure.
Even with built-in impedance linearization, a loudspeaker is still a dynamically changing resistance for the connected amplifier to deal with.
The better it manages this, the better the measurements under variable conditions and the better the acoustic result.
 
True, but that signal is run only for 1 second or so, my understanding is.
That would hardly qualify "continuous" power. I don't remember @amirm talking about this test in more detail but I think @Archimago mentioned that his power test lasts like 30 seconds or so. That's too long to be fed from the buffet capacitors.

Yes, but a loudspeaker is not a resistor with a fixed value, but a complex, varying resistance depending on the frequency.
It has never been the right way to stick a fixed value on a changing complex structure.
Even with built-in impedance linearization, a loudspeaker is still a dynamically changing resistance for the connected amplifier to deal with.
The better it manages this, the better the measurements under variable conditions and the better the acoustic result.
That's all true, of course, but totally not the point here. :) And it's not just the change with frequency. We could go on and discuss at length how many amplifiers struggle to keep their output voltage into reactive loads (not the 3e Audio A5, which does remarkably well in this area). And we could talk about how and if EPDR is able to provide a better metric for compatibility.

But in this case the question was simply if the A5 could drive a certain speaker (Revel Performa3 F208) marketed as being nominally 8 ohm with a sensitivity of 88.5 dB. Truth is, that this is clearly not an 8 ohm speaker so its sensitivity should be stated at 2.0 V, Not at 2.83 V, which would make it look less attractive at 85.5 dB.


The next question is, if the A5 (PBTL, stable down to 2 ohm loads) or the A7s (BTL, stable.down to 4 ohm) was the better suited amp.

Personally, I'd.probably.still.pick the A5 or the A7, not the A7se. But we don't have any independent measurements if the A7se (or A5se for that matter) yet.
 
No disagreement from me here.

I'm having trouble recalling/finding an example, but I do recall seeing reviews of devices with no (or marginal) negative impact to their SINAD when doing DSP functions versus without.
Oh not like that, I meant with EQ you always have to lower the total volume to give the desired frequencies more space. Thus if you add for example a 6dB shelf bass boost, you will have to crank up the DAC or the audio from the source by 6 dB to keep the overall volume similar as compared to before the shelf bass boost. So for any DSP related boosting/cutting you usually trade off total volume, lowering SINAD.
 
A bit late on this one, but a quick plot for the A5:

THD+N Comparison_A5.png


The usual disclaimer applies: I used different sources for the data, so measurement conditions might not be identical. Gains are likely different for the different amps.

Overall, this looks like a competent implementation of the TI TPA3251 chip(s). The performance of the PA5 II shown here is very similar to the original PA5, which is why I didn't plot that one. The A07 uses a TPA3255 and is of course much cheaper and without PFFB.


Sources:
  • Aiyima A07
  • Hypex NC252MP
  • 3e Audio A5
    This thread
  • Topping PA5 II
 
Last edited:
A bit late on this one, but a quick plot for the A5:

View attachment 420054

The usual disclaimer applies: I used different sources for the data, so measurement conditions might not be identical. Gains are likely different for the different amps.

Overall, this looks like a competent implementation of the TI TPA3251 chip(s). The performance of the PA5 II shown here is very similar to the original PA5, which is why I didn't plot that one. The A07 uses a TPA3255 and is of course much cheaper and without PFFB.


Sources:
  • Aiyima A07
  • Hypex NC252MP
  • 3e Audio A5
    This thread
  • Topping PA5 II
The hypex link is not working.

Post edit: interesting having it (and A07) in the mix.
 
Last edited:
Does anyone have a feel for the additional cost to the 60 watt 3e Audio A5 Stereo into a 100w power amplifier into 8 ohms with the same SINAD etc.

Is it only a matter of a more powerful chip and/or power supply?

If it only say another US$100 then you will have a very versatile high performing bargain at only $US400, especially when compared with the hifi jewlery we see in the magazines.
 
Does anyone have a feel for the additional cost to the 60 watt 3e Audio A5 Stereo into a 100w power amplifier into 8 ohms with the same SINAD etc.

Is it only a matter of a more powerful chip and/or power supply?

If it only say another US$100 then you will have a very versatile high performing bargain at only $US400, especially when compared with the hifi jewlery we see in the magazines.
???
This already exists, please take a look at 3E Audio A7 and A7se.
 
What is the “improvement” of A7-mono over A7, is it just a power supply per channel vs. one for both or is there more to it?
 
What is the “improvement” of A7-mono over A7, is it just a power supply per channel vs. one for both or is there more to it?
Simply put, an A7 Mono is exactly half an A7 with its own power supply.
A7 has 2 pieces of TPA3255, one per channel as PBTL, the A7 Mono has a TPA3255 as PBTL.
The advantage of completely mono is no crosstalk, no mutual interference and its own power supply.
And of course you can place it directly next to your speakers.
 
  • Like
Reactions: gvl
It's simply ridiculous. If a hobbyist like @gamerpaddy can depot and reverse engineer the vaunted Topping PA5 module, with just the motivation of fixing his amp, imagine what any competitor with a motivation to make money can do.
 
Is this amp then better than getting two Fosi Audio Mono V3s?
 
Back
Top Bottom