3e audio A7/A7 Mono Amplifier Review

[ziggurcat]

I had provided the temperature data for my A5 earlier. I have since decided to upgrade to the A7 since it appears that the TPA3251 chip used in the A5 does not have DC speaker protection enabled in PBTL mode, while the TPA3255 used in the A7 does. I wish I had discovered this before I got the A5.

In any case, I am looking forward to testing the A7 and comparing it against the A5. I will be posting temperature readings on the A7 within the next few days. I am expecting it to run a bit warmer.

What do you mean by "PBTL mode"?

 

[bboris77]

What do you mean by "PBTL mode"?

Parallel Bridge-Tied Load - without going into technicalities, it is the mode in which the A5 and the A7 use their respective TI chips, TPA3251 and TPA3255. The downside of using the TPA3251 in this mode is that it disables the DC Speaker Protection feature of the chip. The TPA3255 does not have this limitation and has DC speaker protection even when used in PBTL mode.

Perhaps 3e Audio engineers added certain additional protections to the A5 to compensate for this, but I could not find any evidence of that so I jumped ship to the A7.

 

[ziggurcat]

Parallel Bridge-Tied Load - without going into technicalities, it is the mode in which the A5 and the A7 use their respective TI chips, TPA3251 and TPA3255. The downside of using the TPA3251 in this mode is that it disables the DC Speaker Protection feature of the chip. The TPA3255 does not have this limitation and has DC speaker protection even when used in PBTL mode.

Perhaps 3e Audio engineers added certain additional protections to the A5 to compensate for this, but I could not find any evidence of that so I jumped ship to the A7.

I apologize, I'm still not following (I want to be clear, it's not you, it's my idiot brain). What puts the amp into that mode?

 

[NTK]

I apologize, I'm still not following (I want to be clear, it's not you, it's my idiot brain). What puts the amp into that mode?

Short answer: Which mode is used is determined by the 3e Audio design of the amplifier. It is not something the user can change. The A5, A7 and A7 mono operate in PBTL mode. The A5se and A7se operate in BTL mode.

Longer answer:
The TI TPA 3251 and 3255 chips are actually 4 channel amplifiers. It can be configured to operate in 3 different modes:

In the SE (single ended) mode, you can use it to amplify 4 separate (per chip) and independent channels. This mode is rated for speaker impedance down to 2 Ω.

In the BTL (bridge tied load) mode, it pairs up the channels together to double the output voltage capacity (but not current), giving you a 2-channel amplifier (per chip) with double voltage output capacity but same current output capacity of each of the channels as in the SE mode. This mode is rated for speaker impedance down to 4 Ω. Max output power can be as high as 4X the SE mode with high impedance speakers (only when the amplifier is not output current limited).

In the PBTL (paralleled bridge tied) mode, it again pairs up the BTL channels to double to output current capacity. This gives a 1-channel amplifier (per chip) with double voltage and current output capacities as in the SE mode. This mode is rated for speaker impedance down to 2 Ω. The amplifier in the PBTL mode can output twice the current as in the BTL mode, which means it can potentially output twice as much power into low impedance speakers as in the BTL mode.

 

[ziggurcat]

Short answer: Which mode is used is determined by the 3e Audio design of the amplifier. It is not something the user can change. The A5, A7 and A7 mono operate in PBTL mode. The A5se and A7se operate in BTL mode.

Longer answer:
The TI TPA 3251 and 3255 chips are actually 4 channel amplifiers. It can be configured to operate in 3 different modes:

In the SE (single ended) mode, you can use it to amplify 4 separate (per chip) and independent channels. This mode is rated for speaker impedance down to 2 Ω.

In the BTL (bridge tied load) mode, it pairs up the channels together to double the output voltage capacity (but not current), giving you a 2-channel amplifier (per chip) with double voltage output capacity but same current output capacity of each of the channels as in the SE mode. This mode is rated for speaker impedance down to 4 Ω. Max output power can be as high as 4X the SE mode with high impedance speakers (only when the amplifier is not output current limited).

In the PBTL (paralleled bridge tied) mode, it again pairs up the BTL channels to double to output current capacity. This gives a 1-channel amplifier (per chip) with double voltage and current output capacities as in the SE mode. This mode is rated for speaker impedance down to 2 Ω. The amplifier in the PBTL mode can output twice the current as in the BTL mode, which means it can potentially output twice as much power into low impedance speakers as in the BTL mode.

Correct me if I'm wrong, but the gist I am getting is that the TP3251 chip in the A5 (not the A5se) has to sacrifice DC protection in order to take advantage of its PBTL functionality, and the TP3255 chip in the A7, being more capable, doesn't have that compromise.

The "mode" phrasing was confusing me because it makes it seem as though it's something you select or go past a certain threshold to trigger internally -- for example, setting ECO mode on an AVR to Auto where it conserves energy up to a certain volume level, and full power mode kicks in once you go beyond that threshold.

 

[somebodyelse]

Correct me if I'm wrong, but the gist I am getting is that the TP3251 chip in the A5 (not the A5se) has to sacrifice DC protection in order to take advantage of its PBTL functionality, and the TP3255 chip in the A7, being more capable, doesn't have that compromise.

That's what the datasheets say.
TPA3251:

9.4.2 DC Speaker Protection
The output DC protection scheme protects a connected speaker from excess DC current caused by a speaker
wire accidentally shorted to chassis ground. Such a short circuit results in a DC voltage of PVDD/2 across the
speaker, which potentially can result in destructive current levels. The output DC protection detects any
unbalance of the output and input current of a BTL output, and in the event of the unbalance exceeding a
programmed threshold, the overload counter increments until its maximum value and the affected output channel
is shut down. DC Speaker Protection is disabled in PBTL and SE mode operation.

TPA3255:

9.4.1.3 DC Speaker Protection
The output DC protection scheme protects a speaker from excess DC current in case one terminal of the
speaker is connected to the amplifier while the other is accidentally shorted to the chassis ground. Such a short
circuit results in a DC voltage of PVDD/2 across the speaker, which potentially can result in destructive current
levels. The output DC protection detects any unbalance of the output and input current of a BTL output, and in
the event of the unbalance exceeding a programmed threshold, the overload counter increments until its
maximum value and the affected output channel is shut down. DC Speaker Protection is disabled in SE mode
operation.

Note that it's about protection when one of the terminals is accidentally grounded, not when the amp is fed DC input, or in case of an amp fault.

Most people don't need PBTL as few speakers dip much below 4R, so should probably get the cheaper SE models.

 

[NTK]

Echoing what somebodyelse posted — the intended purpose for the DC speaker protection is very narrow. The protection threshold for both the TPA3251 and TPA3255 is 1.5 A, which for a 4 Ω load means 6 V before protection will kick in.

These TPA amplifier chips operate with "single supply" (only positive supply PVDD is needed) and are not DC coupled (DC blocking capacitors for all inputs). If there is significant amount of DC in the output (spec is ≤ 60 mV for BTL mode), something is seriously malfunctioning. At that point, (and if I am paranoid about this failure mode) I would not rely on the built-in protection circuitry to protect against the failure but to use external protection, for example: this.

 

[bboris77]

For reference and comparison, my A5 runs at 34C and my ambient temperature is 23C. The 38V5A adapter runs around at 27C.

Got my A7. It runs at 35C and the power adapter is at 30C with ambient temp of 23C. So, very marginally warmer compared to its little brother. Still more than acceptable.

The potentiometer on the A7 feels more substantial and has a decent amount of resistance compared to my sample of A5 which turned a bit too lightly. Not sure if they use the same potentiometer, but it feels quite different. To be honest, I use the A7 as a power amp only, bypassing the potentiometer anyway, so I never bothered to test channel matching on lower volume settings.

DC offset on the left speaker out is only 3.95mV and 3.63mV on the right. These values are excellent and a bit better than the A5 sample I tested (L 14.16 mV R 8.43 mV).

Additional gain is also quite welcome since it allows me to use more aggressive room EQ which requires applying up to -8dB on the preamp setting of my DAC (The Element IV).

The pop/tick sound on the power up and power down is exactly the same like on the A5 and is very mild - a non issue. Tweeter hiss at low gain is at the exactly same level as high gain on the A5 - basically inaudible unless you put your ear right next to the tweeter. It is much better than my Yamaha A-S501 which was already inaudible from about 10cm distance from the tweeter. If you want the absolute least amount of self-noise, the A5 in low gain is the absolute champion, but this is purely academic.

 

[Maniek]

Temp 25-40°c normal

 

[harkpabst]

I'm not an English speaker and I used a translator, so I think something is wrong

"getting hotter" and "warm" might be right

In conclusion, it's really hot

I surely wasn't nitpicking on your use of language. I'm also not a native speaker.

But that's precisely the problem: a human is completely unsuitable as a thermometer, and the perceived difference between warm and hot is often only 3-4°C.

The feel of the surface also has a huge impact. With smooth aluminum, you could burn your fingers slightly at 55-60°C, while you could easily hold slightly rough plastic at 110°C for a few seconds without any harm. That's how different the heat transfer to our fingers and the perception of heat are.

At 40°C, there's absolutely no need to worry about amplifiers, but we do perceive them as hot.

Accurate infrared thermometers can be found on Aliexpress and Amazon for as little as €6-18.

This is it. You really have to take measurements to judge if the temperature is abnormally high. A case temperature of 40° C (or more, if the ambient temperature is above 25° C) is absolutely no problem at all.

The real thing to worry about with the TI TPA325x chip amps is not the case temperature. The critical point is the heat transfer between the chip(s) and the case as a heat sink. If there are no general design flaws a hot case mostly indicates that the heat transfer is working well.

 

[Pulkass]

This is a review and detailed measurements of the A7 / "A7 Mono" stereo power amplifier with balanced inputs. It was sent to me by the company. I have no idea what it costs in US. Audiophonics lists it for 379,00 €tax incl.
View attachment 444144
While the A7 may look like its competitors, it has a heft to it that I don't recall others not having. It really feels like it is filled with lead or something! A nice volume control in the middle adjusts the gain but it can also be bypassed using a switch in the back. Pushing the volume control quickly switches between XLR balanced and RCA. Two thoughtful clipping indicators are provided, making me wish it would be a standard feature on any amplifier. A minor nit: LED colors indicating which input is active is red and clipping in green. I think based on convention, the latter should be red.

Back panel shows nice set of features:
View attachment 444145
Trigger input is provided as is the aforementioned combo TRS/XLR balanced inputs. Low and high gain switch is provided but I focused on low gain as that is fully compatible with current audio products outputting 2/4 volts. Yes, there is that monster GaN power supply! It can of course be purchased with other options. The input connector says max is 48 volt but per designer, the mono configuration, which is just using the left channel, can go up to 52 volts. He said I could test in stereo just the same and that is what I did. I found no issues.

FYI, I was impressed with the secure way the power supply input accepts the plug. There is a lot of current and power involved here so good to see a tight fit.

One nice feature, beside trigger, is auto-shut off in case of no input signal. I left it on by accident and was pleasantly surprised when I came back and it had gone to sleep.

3e audio A7 Amplifier Measurements
Let's start with low gain, volume control set to (volume bypass didn't make a difference), and XLR as inputs:

View attachment 444146
Very nice. Well above average in performance with distortion below threshold of hearing:
View attachment 444147
View attachment 444148

RCA input as usual loses a bit, but not much, scoring 100 dB with again, inaudible distortion:
View attachment 444149

SNR is the limit here, which is quite good actually:
View attachment 444150
At full power, it provides transparency for practically all music produced!

The A7 uses post filter feedback (PFFB) so it is essentially load (speaker) independent:
View attachment 444151

Crosstalk while much better than average, is a bit worse than 3e audio A5:
View attachment 444152

As is typical here, above 1 kHz, distortion does climb:
View attachment 444153
View attachment 444154

This is one powerful little amplifier, easily besting its nicely mentioned specs:
View attachment 444155
View attachment 444156

View attachment 444157

Edit: forgot to include the power vs frequency sweep:

View attachment 444184
Power is more or less sustained across the full audible band:
View attachment 444158
View attachment 444159

Its distortion goes from 0.05% to huge amount as it clips so I could not test it at 1% THD at higher frequency. So I searched for power point at just 0.05% which disadvantages this amp a bit.

My reactive loadbox still doesn't like these bridged amplifiers so I could only test the left channel (although per above, stereo and mono performance were the same):
View attachment 444160

View attachment 444161

You have to keep reminding yourself that this much power is coming out of this little box, no bigger than the old CD cases!

Finally, the amplifier is rock solid on power up:
View attachment 444162

Oh, too much on my mind so I forgot to run the power up/down glitch test.

Conclusions
If my memory serves me right, 3e audio was the first company to bring us high performance TI TPA3255 class D amplifiers with PFFB. While they have more competition now, it is clear they intend to stay on top of this game, squeezing every bit of performance out of this architecture, while providing nice feature set. We are talking high performance, high power amplifiers that are tiny yet deliver what people need in most cases.

I am happy to recommend 3e audio A7 / A7 Mono stereo audio amplifiers.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

I bought one. Very good indeed as it sounds. But.....
Got it — you want to turn the 3E Audio A7 completely off, not just disable the auto-sense.

Unfortunately, the A7 does not have a dedicated on/off power button — it’s designed to stay in standby when idle.
The only way to truly switch it off is to:

Unplug the power cord from the mains, or

Use an external power strip or inline switch on the AC side so you can cut power without unplugging every time.


Some users also plug it into a smart plug so they can power it off fully from a button or app.

If you want, I can tell you how to set it so it never turns on by itself after you’ve switched it off at the strip. That way it won’t unexpectedly power up when there’s signal.

 

[harkpabst]

Unfortunately, the A7 does not have a dedicated on/off power button — it’s designed to stay in standby when idle.

Standby and idle are two completely different states. And pressing the button for two seconds switches the amp to standby. You don't have to wait for it entering standby automatically.

Did you measure the power consumption in standby? My power meter couldn't measure it.

If you want, I can tell you how to set it so it never turns on by itself after you’ve switched it off at the strip. That way it won’t unexpectedly power up when there’s signal.

I'm not sure if @amirm is too keen to learn about this but why don't you just demonstrate this trick to all of us?

 

[Pulkass]

Standby and idle are two completely different states. And pressing the button for two seconds switches the amp to standby. You don't have to wait for it entering standby automatically.

Did you measure the power consumption in standby? My power meter couldn't measure it.


I'm not sure if @amirm is too keen to learn about this but why don't you just demonstrate this trick to all of us?

I just want to turn it off ! And it seems impossible. The other suggestions came from my enquiring AI. Tricks l have not yet tried..

 

[ziggurcat]

I just want to turn it off ! And it seems impossible. The other suggestions came from my enquiring AI. Tricks l have not yet tried..

You can. You hold the knob on the front down for a few seconds to put it into standby mode. If you want to disable the auto-sensing, hold the knob down for 5 seconds while it is in standby mode, and either the XLR or RCA light will blink for 3 seconds (RCA blinking = auto sense disabled, XLR blinking = auto sense enabled, if I recall correctly). With auto sense disabled, it will not turn on unless you short press the knob on the front to turn it on, and it won't turn off unless you hold the knob down for a few seconds. It's really not that difficult.

Power draw while it is in standby mode shouldn't be a concern if @harkpabst says it doesn't even register.

What exactly are you trying to prevent/avoid?

 

[Maniek]

Chcę to po prostu wyłączyć! A wydaje się to niemożliwe. Pozostałe sugestie pochodziły od mojej dociekliwej sztucznej inteligencji. Sztuczki, których jeszcze nie wypróbowałem.

 

[Maniek]

Normal

 

[harkpabst]

I just want to turn it off ! And it seems impossible.

Even if you could switch the amp completely "off" (so none of its circuitry was powered and it could not wake up from standby), which you can't (indeed), the power supply was still not "off". The only way to make sure that the entire apparatus is off is by means of a power strip, anyway.

@ziggurcat asked the right question: What exactly is it that you don't want to happen?

The other suggestions came from my enquiring AI. Tricks l have not yet tried..

So, you offered help to our host, based on untested AI results but won't share them with us simple forum visitors? Come on, now we're curious.

Well, at least I am, since the 3e Audio amps are designed to start up in "on" mode as soon as power is applied. It might be able to change that through a firmware update. I see no easy way to do that.

 

[gekkekoe]

For this amp, they seem to bundle it with the 48v/5a power supply. You need to get the 10A power supply to get the max output of 2 x 240W (48*10/2) at 4 ohm ?
They don't seem to specify what the output is with the 48v/5a?

 

[Maniek]

3eAudio,Fosi,WiiM etc Klass D TI 325…

 

[harkpabst]

3eAudio,Fosi,WiiM etc Klass D TI 325…

What exactly is your point?

For this amp, they seem to bundle it with the 48v/5a power supply. You need to get the 10A power supply to get the max output of 2 x 240W (48*10/2) at 4 ohm ?
They don't seem to specify what the output is with the 48v/5a?

You did the math correctly. If you want a continuous output of 2 x 240 W for an arbitrary amount of time you will need a PSU capable of outputting >= 480 VA for the same period of time.

With a lower speced PSU you can still have 240 W per channel for a limited time, covering the peaks. 3e Audio did mention before that the current rating for the power bricks they supply is very much on the conservative side. If you don't want to rely on out-of-spec tolerances you have to bite the bullet and go for the 48 V 10 A PSU.