3e Audio A5 Stereo Amplifier Review

[jlx]

let me wipe my black makeup and test this
update: blasting some Sisters of Mercy (my neighbors thank you). no pop

apologies in advance, I had to do it

 

[ziggurcat]

Very unlikely with a post feedback control. The amp surveilles its own output. If you hear cracks or pops you should check your DAC first and make shure it is isolated from the mains.

I can actually confirm the popping. Two pops after pausing/stopping any audio. It would occur when my Mac was connected to my DAC via USB and then out from the DAC to the amp (Balanced -- was virtually gone with an RCA connection), but is not present when streaming from the Mac to the DAC via Bluetooth or TOSLINK out from a TV to DAC and then out from there to the amp.

What I have not done is try connecting everything to a different computer.

It doesn't mean there's something wrong with the amp, there's just something about it that brings it to light.

 

[harkpabst]

let me wipe my black makeup and test this
update: blasting some Sisters of Mercy (my neighbors thank you). no pop

I'd really like to respond with any kind of laughing smiley, but the forum software won't let me. Please, imagine me rolling on the floor, laughing.

Pops and clicks can be caused by a number of devices under certain circumstances. DACs (not necessarily the chip, but the device sold as a product) are not unlikely to be the culprit, but there can be other issues. Some devices might be more prone to pick that up than others. It's not an easy subject. But all in all a power am is just not so likely to be the root cause of such problems. It's not impossible with modern amps, having their own firmware and stuff, but just not the first candidate on the list.

 

[Punchy_Power_Cat]

TPA3251 is typically rated for 36V. There is support for 38V in the data sheets but it is considered the limit. Do you guys think this is an issue long term usage wise? Maybe one would be advised to swap the PSU for a decent 36V one. Output power decrease should be minimal. Sound quality should -if anything- improve ever so slightly.

edit: same problem decribed in this thread we have here

On the GAN power brick is says:
input: 230V 3.5A / output: 38V 5A

yet, the power cable that ships with the amp says:
230V 2.5A max. / it is a 2x0,75mm only

Also the PSU is not grounded. Untypical solution for a PBTL amplifier that in theory reaches 355W. There is a third ground pin from the PCB to the DC-IN roundport. But since there is no third grounding pin to the wall from the PSU, the ground from the PCB leads only to the PSU. Isn't this (aside from the unmatching cables) a safety hazard?

 

[harkpabst]

TPA3251 is typically rated for 36V. There is support for 38V in the data sheets but it is considered the limit. Do you guys think this is an issue long term usage wise? Maybe one would be advised to swap the PSU for a decent 36V one. Output power decrease should be minimal. Sound quality should -if anything- improve ever so slightly.

True, the data sheet says typ. 36 V and max. 38 V DC supply voltage. What we don't know (at least I don't) is if the output of the external PSU is really directly fed into the PVVD pins without any further regulation.

However, given the already excellent measurements I cannot imagine anything could be gained here, either way.

On the GAN power brick is says:
input: 230V 3.5A / output: 38V 5A

yet, the power cable that ships with the amp says:
230V 2.5A max. / it is a 2x0,75mm only

Good idea to keep an eye on this, but I don't see a problem here. The writing on the power brick (at least on the ones that I received) is actually:

Input: 100-240 V AC 50/60 Hz 3.5 A
Output: 38.0 V DC 5.0 A 190 W

So, the specified maximum power output is 190 W which (randomly assuming a 90% efficiency) translates to around 230 V 0.92 A on the input side. No problem at all with the supplied power cord.

3e Audio once claimed that in their internal testings these power bricks could output much higher than rated current, something around 8 A IIRC. Let's stretch this and do the math for double the rated output, so 38 V 10 A. Not surprisingly, under identical preconditions, 380 W output would mean 230 V 1.84 A. This is still far away from the specified max. of 2.5 A. And in such a scenario I really wouldn't worry about the power cable, but the power brick itself!

The power brick is correctly labelled as a Class 2 device, so doesn't need a safety ground. At least in Europe Class 2 devices are limited to a current intake of 2.5 A, so providing a "better" two prong cable wouldn't even be allowed.

There is a third ground pin from the PCB to the DC-IN roundport. But since there is no third grounding pin to the wall from the PSU, the ground from the PCB leads only to the PSU. Isn't this (aside from the unmatching cables) a safety hazard?

Did you examine your own amp or did you study the pictures posted here, mostly showing pre-production models?

There shouldn't be a third pin on the DC input socket. There is a third wire to be seen connecting to the trigger input, which the production units do not have. Is it possible that you've been confusing this from the pictures?

Again, I don't see a safety hazard.

 

[Punchy_Power_Cat]

Everyone please be careful when you power it the first time.

The button for putting the amp into bypass mode is somehow inverted (from common logic).
If you look directly at the button and at the backside of the amp one would assume "to the right = bypass on" but it's the other way.
Thankfully I was clever enough to play the first sound with 5% vol otherwise the combination of DACmode+bypassmode would have been really nightmarish.

 

[ziggurcat]

Everyone please be careful when you power it the first time.

The button for putting the amp into bypass mode is somehow inverted (from common logic).
If you look directly at the button and at the backside of the amp one would assume "to the right = bypass on" but it's the other way.
Thankfully I was clever enough to play the first sound with 5% vol otherwise the combination of DACmode+bypassmode would have been really nightmarish.

I remarked how they should have the volume bypass disabled by default because when I set my A7 up for the first time and hit play, it was excruciatingly loud. Since my DAC is in DAC mode (which outputs at full volume), both devices were outputting at max volume, and probably did some damage to my speakers. I spent 20 minutes comforting my cat because it scared the living crap out of him.

 

[Punchy_Power_Cat]

I remarked how they should have the volume bypass disabled by default

With my A5 it was disabled by default. However I got confused due to the bypass trigger being "on" at the left position.

I really try to fall in love with the A5, there is no pop or crack, it is very powerful, probably won't even reach 50% even with low gain. The auto on/off works beautifully with the 20min timer. Everything seems perfect. But I am struggeling with the sound, especially with voices. While being superbly detailed I also find it lacking something not easy to pin point and probably misleading to readers. Don't get me wrong, this is a highly detailed and very wide soundstage but it is not as open as it was with my PBTL 12070 Merus amp. My LS50 Meta speakers went from "open baffle" to "normal speakers". I really need to talk to Daniboun about this. It probably comes down to the difference between a filtered IC and a non filtered IC. It is very interesting. There is no way the Merus amp measures better, but it sounds so much more vivid.

 

[Panelhead]

I almost bought one. Seemed “better than a PA5 II” in frequency response with 8 ohm load above 20Kz and the has the auto off feature.
They were a little hard to find at that time. Amazon has them now.
Got a PA5 II. The lack of turn off pop was final deciding factor. The limited availability and higher price of the A5 made justifying the Topping a little easier. Percentage wise there is a difference in price. In absolute dollars the difference is minimal. Both are what I consider as low cost.
Still wonder if an A5 is not the “better” amplifier.

 

[Punchy_Power_Cat]

The lack of turn off pop was final deciding factor.

There is no turn off pop! Amir got a pre-production unit that probably got fixed in that regard. His unit also had 12V Trigger which they abandoned for the auto-off feature which is awesome. There is a "click" from a relay within the enclosure when it shuts itself off and two clicks when it shuts itself on (soft clicks to the ear). I like that click. It's reassuring the amp is doing what it is supposed to do. The amp is as silent as a grave to the speakers. However, I have also wrapped the AC-IN cable many times around a big donut shaped N30 Toroid to clear the PSU from high frequencies.

Using pure-path/bypass the amp got a lot more tamed in terms of output but even more clarity was achieved. I suppose the RCA gain feature works only in bypass mode. The soundstage is huge and instrument seperation is fantastic, better than the 2x12070 Merus design. I can pin point cues and symbols within the mix with such accuracy that I am confident you can power a professional monitor and really work with this amp. The problem with the vocals tho is bugging me. There is different approach in how the ICs handle vocals within the mix. With the TPA3251-A5 you will get a "dense wall of sound" with vocals being as present as the symbols/cues but not much more. With the 2x12070 Merus I have a little more compression in the mix, but the singer is always "in your face" and I mean like a private show or something, you close your eyes and he or she stand right next to you. With the A5 the vocals are not nearly as up front, the vocal image does not fill the entire space you are looking at, they're much more laid back in the mix. The 2xTPA3251-A5 is based upon multiple filters. The 2xMA12070 designs are non filtering ICs, so they represent a very different approach. Both seem to have advantages and disadvantages. The former being more powerful and detailed, the latter being more "linear/analogue" sounding.

2xTPA3251-A5 : James Brown sounds awesome, everything else slightly muffled
2xMA12070-A20a : Everything sounds awesome, James Brown give you a headache.

 

[harkpabst]

Everyone please be careful when you power it the first time.

The button for putting the amp into bypass mode is somehow inverted (from common logic).
If you look directly at the button and at the backside of the amp one would assume "to the right = bypass on" but it's the other way.
Thankfully I was clever enough to play the first sound with 5% vol otherwise the combination of DACmode+bypassmode would have been really nightmarish.

As said elsewhere before ...

3e audio A7/A7 Mono Amplifier Review

Just to close the loop completely, and remove 100% doubt, I did try the A7 with my E50 DAC on my bedroom TV setup. The connection chain was: Cable Box -- HDMI --> TV -- TOSLINK --> E50 DAC -- Balanced Out (TRS to TRS) --> A7 No issues whatsoever. As was expected based on information from the...

audiosciencereview.com

... there is no bypass mode switch to be enabled or disabled.

There is a switch, whitch is either in the BYPASS position (left) or in the VOL position (right).

 

[avdoc]

Received my A7 stereo today from Amazon. I can confirm it outperforms the Fosi V3 monos in my setup. However, it runs hotter than I would like. At idle, the top hits 108 degrees Fahrenheit vs 116 for the Fosi monoblocks. I'm a little disappointed, as I was hoping not to have to use cooling fans with it. However, it sounds better than the Fosi and the rca inputs sound just as good as the xlr. I had planned to get another, but it's no longer listed on Amazon. Hopefully, it will come back.

 

[Punchy_Power_Cat]

There is a switch, whitch is either in the BYPASS position (left) or in the VOL position (right).

That's the point. You can read " | bypass vol" or you can read "bypass | vol".

However, it runs hotter than I would like.

We're in summertime but my A5 also runs a little hotter than I would prefer. You have to keep in mind that the amp board is hanging from the top down, hence the PCB is directly heating up the upper side (which is the side you would most probably touch). Actually this is a better solution than a laying PCB that most amps have since the heat wants to go up anyway.

We could add some cooling fins to the top. If 3E Audio was cool too, they could be making sure we get some perfectly matching fins with some preinstalled thermal pad to glue it right onto the top. I'd pay up to 20 Euros.

 

[Alannn]

That's the point. You can read " | bypass vol" or you can read "bypass | vol"...

I set the bypass volume switch to off so that it bypasses the volume pot. Once set I never have to read anything because my needs are not so fickle. Set and forget. Who cares what it says, as long as it provides switching ability.

If the first time you power the new amp on, you have a loud source inputting, you are reckless. I am never that trusting when my speakers are at risk.

 

[Punchy_Power_Cat]

I am never that trusting when my speakers are at risk.

Me neither. As long as the DAC does not for some weird bug go into DAC mode all over sudden, I am fine.

The problem I had with the rendering of vocals disappeared to great extend. Maybe there is a burn-in effect either in the GaN PSU or in the amp (coils) or in the speaker drivers or in my brain or in all of those but it sounds more open now. Very impressed with the sound stage and the locality of instruments.

However, there seems to be a problem with the PSU. The amp seems limited by it. I have direct comparison to a 2x12070 Merus Design amp which is also PBTL 2x90W@4ohm running off a 400W medical PSU and it wipes the floor with the A5 in terms of base. You'd have to consider that the 12070 in PBTL can reach down to 1.6ohm, whereas the A5 can go only down to 2ohm, but it shouldn't matter practically because the LS50 Meta I am driving are I belive around 3.2ohm at their lowest.

Playing "praise the lord -asap" bass heavy track with full range signal ~85db (very loud)
2x12070 PBTL = strong visible woofer excursion
2xTPA3251 PBTL = no woofer excursion at all

That doesn't mean that there is no base with the A5 it is just not nearly as full and deep as I am used to. I need much higher volume to get that woofer to bounce. So high actually that it starts to get painful. I will try swapping the PSU for a more powerful one and report here if there is a benefit. Could alsobe gain related since I am currently running @low gain RCA.

 

[Roland68]

We're in summertime but my A5 also runs a little hotter than I would prefer. You have to keep in mind that the amp board is hanging from the top down, hence the PCB is directly heating up the upper side (which is the side you would most probably touch). Actually this is a better solution than a laying PCB that most amps have since the heat wants to go up anyway.

We could add some cooling fins to the top. If 3E Audio was cool too, they could be making sure we get some perfectly matching fins with some preinstalled thermal pad to glue it right onto the top. I'd pay up to 20 Euros.

"Hotter" is a very subjective term that can mean 32°C or 60°C.
In addition, the ambient temperature is very important. At an ambient temperature of 20°C, 45°C would be quite high. But at an ambient temperature of 38°C in the summer, 45°C would be a good value.

If you can leave your hand on the amplifier for 10 seconds without burning or hurting yourself, there's nothing to worry about. Above that, you need a reliable measurement.
Simple and reliable thermometers are available for around €2. Reliable infrared thermometers cost €15-20; I've seen some on Aliexpress for €6.

If you really want to use a heatsink, it's sufficient to coat it thinly with thermal paste and press it onto the amplifier case. It adheres more than adequately.
Used heatsinks and leftover sections can often be found on eBay. There are also many (passive) CPU fans you can use. Hard drive coolers (approx. 70x100mm) or two or four SSD/NVME coolers can also significantly reduce the temperature. With the latter, you'll need to ensure the bottom surface is flat.

 

[Punchy_Power_Cat]

If you can leave your hand on the amplifier for 10 seconds without burning or hurting yourself, there's nothing to worry about.

For sure this is the case. Anyway there are two things I don't like electronics running on for years to come. a) at the peak of their allowed voltage b) at temperatur
Both can be mitigated, the latter even with considerably low cost. I don't want an amplifier to die ever. Not in two years, not in twenty.

 

[Roland68]

For sure this is the case. Anyway there are two things I don't like electronics running on for years to come. a) at the peak of their allowed voltage b) at temperatur
Both can be mitigated, the latter even with considerably low cost. I don't want an amplifier to die ever. Not in two years, not in twenty.

You're overlooking a few points.
- In older full-size amplifiers, the temperatures of the heat sinks and components were often 50°C and above, while the case temperatures were closer to 30°C. These amplifiers lasted 20-30 years without any problems.
- Many people ignore the temperature-dependent operating point of components, which requires both a minimum temperature and a constant temperature.
- Unlike analog amplifiers and transformer-based power supplies, heat is no longer the problem for components, especially capacitors, in Class D and switching power supplies. What causes capacitors and other components to fail is the high switching frequencies. This experience in the audio sector also coincides with experience with IT hardware, many industrial sectors, as well as long-term and aging tests.

So you're worrying a lot about a heat problem that doesn't even exist in most cases.

 

[pherax]

I just received my e3 audio A5 amplifier. I intended to use it to upgrade the amplifier bit of my sr5011. Unfortunately I do not hear any audible differences on my MA silver 500 7g speakers. Does this mean the DAC in my sr5011 is the bottleneck? I could also go for another dac but I like the convenience of my sr5011... I stream music via rpi3b (24bit flacs)

 

[Beave]

I just received my e3 audio A5 amplifier. I intended to use it to upgrade the amplifier bit of my sr5011. Unfortunately I do not hear any audible differences on my MA silver 500 7g speakers. Does this mean the DAC in my sr5011 is the bottleneck? I could also go for another dac but I like the convenience of my sr5011... I stream music via rpi3b (24bit flacs)

In audio reproduction, the "bottleneck" is almost always the speakers and the room acoustics and hardly ever the DAC. The amp can be a bottleneck in certain situations where it can't provide sufficient power (inefficient speakers played loud in large room, for example). But in most cases, you should not hear much if any difference between properly designed and functioning amplifiers.