3e audio A7/A7 Mono Amplifier Review

[Bernard23]

On paper maybe, but can you point to an actual amp showing these benefits?

not in the world of audio no.

 

[Bernard23]

That just goes to show how easily people fall for marketing hype.
GaN FETs really shine in high-power—and, above all, constant-power—applications. And what is the one thing an audio amplifier *doesn't* have? Exactly: constant power output.
Furthermore, the advantages are usually cited in comparison to standard applications; yet amplifiers like the 325X series—as well as those from Purifi and Hypex—are already highly efficient switching amplifiers, operating at 90–93% efficiency.
On top of that, the operating parameters required for GaN FETs impose various design constraints.
Given these circumstances, a 1% efficiency gain over current switching amplifiers is actually substantial, while a 2% gain is pure fantasy.

There is a reason why none of the current manufacturers of GaN FET amplifiers show genuine efficiency measurements taken under real-world conditions in comparison to other switching amplifiers. They might not even be any better.

Which would you prefer: an amplifier optimized for audio reproduction, or one optimized for a specific technology?
Current GaN FET amplifiers don't convince me—partly due to reliability issues and their measured performance.
There will certainly be further, improved developments in the future, though the technology might well be overtaken by something else entirely.

Agreed. There are theoretical benefits of a GaN FET over a MOSFET, that's just the physics. How it is implemented is critical, it's an entirely new technology. Is it worth it for consumer audio? Who knows, probably if you're a manufacturer because placebo sells. There would likely be measurable differences, but are they audible? Personally I like the idea of 500+W in a box that isn't the size of a small fridge and weighs 100kg.

 

[Roland68]

not in the world of audio no.

Agreed. There are theoretical benefits of a GaN FET over a MOSFET, that's just the physics. How it is implemented is critical, it's an entirely new technology. Is it worth it for consumer audio? Who knows, probably if you're a manufacturer because placebo sells. There would likely be measurable differences, but are they audible? Personally I like the idea of 500+W in a box that isn't the size of a small fridge and weighs 100kg.

And there are reasons for this.

Even a major manufacturer like Infineon—which, having invested heavily to acquire GaN FET technology, ought to have a strong vested interest in its success—has little to show for it beyond marketing promises. I have yet to see any tangible operational advantages compared to their own conventional products or those from other manufacturers.

Why not take a look at offerings from Purifi and Hypex?
There are plenty of amplifiers in the 500- or 700-watt range that weigh around 10 kg (or less); any extra weight is simply due to higher-quality chassis construction.
Or consider the 3E Audio PAM258, for example: over 1,000 watts of continuous power and 2,000 watts max power for €1,000, with a weight of 6.8 kg according to the manufacturer.

GaN FET amplifiers don't get any smaller or lighter than that, either.

 

[Mnyb]

IS not the consensus that GaN makes sense in power supplies ( they seems to pop up everywhere for portable stuff and laptops etc ) or extremely powerful amps in the several kW range ? but not much else ?

 

[Roland68]

IS not the consensus that GaN makes sense in power supplies ( they seems to pop up everywhere for portable stuff and laptops etc ) or extremely powerful amps in the several kW range ? but not much else ?

I would never dispute that GaN power supplies and chargers are recognized and offer advantages.

GaN-FET power supplies definitely offer advantages, even if they aren't as significant when used for amplifiers compared to other applications.

But what advantages do GaN-FET amplifiers offer in audio applications?
1% better efficiency, perhaps?

However, the moment GaN power supplies are used for audio amplifiers, their efficiency drops compared to charging applications or constant power supply scenarios.
This is the aspect that marketing campaigns invariably gloss over.

This becomes even more apparent when they are used in power output stages. To date, neither amplifier manufacturers nor semiconductor makers have been able to demonstrate any advantages—even though doing so would be quite simple.
It is also important to realize that even under optimal conditions—and setting aside the issue of highly fluctuating loads—the current 90–93% efficiency of switching amplifiers could only be improved by about 1–3%. Even under the best circumstances, achieving an efficiency significantly above 96% with a GaN switching power supply is currently unrealistic; furthermore, there are several limiting factors when used in a power output stage.

Claims of 20–30% improvements are another marketing gimmick. Sure, that might hold true if you compare them to conventional power supplies or chargers from the 80s and 90s. But with modern devices, the advantage is more like 2–10%—or just 2–6% if you compare high-quality conventional units with GaN ones.

The audio sector—specifically regarding GaN amplifiers—really highlights how easily people fall for marketing hype when claims are heavily exaggerated and crucial information is withheld.

That said, this doesn't mean future generations of GaN amplifiers won't offer advantages; it’s just that they don't yet.

Let’s not forget that it took Class D amplifiers 70 years to reach their current state, yet they can still be outperformed in measurements by Class AB, or B amplifiers.
Digital amplifiers have been around for nearly 30 years, yet they are still in their infancy.
The first GaN FETs were sold only 22 years ago; let's give the technology some more time to develop.

 

[Aimarketingbot]

And hopefully GaN FET class D amps will be the new de facto standard, though they might not see the full benefit of the low cost of the FET devices pull through to the amp unit prices. That's after the new tech premium price wave has passed.

Will be nice to see in future indeed! I get a feeling some EU regulation will be the straw though.

 

[Bernard23]

GaN is already in use for critical industrial applications, as a wide band gap semiconductor material it has significant advantages over silicon which definitely isn't marketing hype, but then consumer audio isn't energy critical or a harsh environment. I spend time in wide band gap materials in the day job, but I'm not an electronics engineer, I'm more interested in how we process the materials into useable devices. The UK is investing significantly into wide band gap and ultra wide band gap materials for power applications in all their forms. Interesting times. We're likely to see a broader spread of semicon materials that are more specific to their application, it's a great time to be a materials scientist in electronics.

 

[Aimarketingbot]

But with modern devices, the advantage is more like 2–10%—or just 2–6% if you compare high-quality conventional units with GaN ones.

Checks out with my experience. In 2008 I was running an array of Meanwell HLGs which do up to 93~94% efficiency at peak load, still use em, GaN beats them at peak efficiency by 2%, absolutely not worth it without near-price. Nothing under the sun is new it seems you'd barely see a return with audio applications.

GaN is already in use for critical industrial applications, as a wide band gap semiconductor material it has significant advantages over silicon which definitely isn't marketing hype, but then consumer audio isn't energy critical or a harsh environment. I spend time in wide band gap materials in the day job, but I'm not an electronics engineer, I'm more interested in how we process the materials into useable devices. The UK is investing significantly into wide band gap and ultra wide band gap materials for power applications in all their forms. Interesting times. We're likely to see a broader spread of semicon materials that are more specific to their application, it's a great time to be a materials scientist in electronics.

Oh yes, there are already applications where it 100% makes sense and pays for itself.
Fascinating, so you do lot of crystallography work?
Makes sense UK would go there, especially for EV applications, every % counts on a mass scale.

 

[Roland68]

GaN is already in use for critical industrial applications, as a wide band gap semiconductor material it has significant advantages over silicon which definitely isn't marketing hype, but then consumer audio isn't energy critical or a harsh environment. I spend time in wide band gap materials in the day job, but I'm not an electronics engineer, I'm more interested in how we process the materials into useable devices. The UK is investing significantly into wide band gap and ultra wide band gap materials for power applications in all their forms. Interesting times. We're likely to see a broader spread of semicon materials that are more specific to their application, it's a great time to be a materials scientist in electronics.

Our high-energy products are arguably one of these critical industrial applications. It can work extremely well, though it really depends heavily on the specific use case; GaN FETs and GaN technology aren't inherently superior in every respect.
We do use the technology in the control circuitry for our high-energy products, but there we are dealing with power ranges involving kilovolts and kiloamperes. It generally only makes sense at high switching frequencies—in the high hundreds of kilohertz or even the megahertz range. However, we have encountered instances involving wildly fluctuating power demands and variations in switching frequency and pulse duration where GaN FETs performed poorly—incidentally, these are the same parameters relevant to audio amplifiers.

With the 3E Audio A7 discussed here, we get the best of both worlds: an energy-efficient TPA3255—featuring a power stage with very low idle losses (under 2.5 W) and optimized MOSFETs with optimized gate drivers that achieve far lower idle losses than conventional discrete implementations—combined with highly efficient GaN power supplies.

 

[metal4ever]

OK, right now I have my Klipsch Heresy IV, Turntable, Schiit Mani 2 Phono Preamp to the 3e A7. I love it. I may just keep it that way, but I am considering adding the WIIM Ultra Preamp because of room correction feature. Would it be a good recommendation to use the room correction with my Klipsch Heresy's? Are there any settings on the WIIM and 3e7 I should use?

 

[Bernard23]

OK, right now I have my Klipsch Heresy IV, Turntable, Schiit Mani 2 Phono Preamp to the 3e A7. I love it. I may just keep it that way, but I am considering adding the WIIM Ultra Preamp because of room correction feature. Would it be a good recommendation to use the room correction with my Klipsch Heresy's? Are there any settings on the WIIM and 3e7 I should use?

Room correction is potentially good with any speakers, as it's the room that's the biggest influence on how your speakers sound.

 

[Bernard23]

Our high-energy products are arguably one of these critical industrial applications. It can work extremely well, though it really depends heavily on the specific use case; GaN FETs and GaN technology aren't inherently superior in every respect.
We do use the technology in the control circuitry for our high-energy products, but there we are dealing with power ranges involving kilovolts and kiloamperes. It generally only makes sense at high switching frequencies—in the high hundreds of kilohertz or even the megahertz range. However, we have encountered instances involving wildly fluctuating power demands and variations in switching frequency and pulse duration where GaN FETs performed poorly—incidentally, these are the same parameters relevant to audio amplifiers.

With the 3E Audio A7 discussed here, we get the best of both worlds: an energy-efficient TPA3255—featuring a power stage with very low idle losses (under 2.5 W) and optimized MOSFETs with optimized gate drivers that achieve far lower idle losses than conventional discrete implementations—combined with highly efficient GaN power supplies.

Interesting. Why are GaN fets less efficient under fluctuating voltage / current conditions? Would other wide band gap semiconducting materials have the same issue?

 

[Ruediger]

OK, right now I have my Klipsch Heresy IV, Turntable, Schiit Mani 2 Phono Preamp to the 3e A7. I love it. I may just keep it that way, but I am considering adding the WIIM Ultra Preamp because of room correction feature. Would it be a good recommendation to use the room correction with my Klipsch Heresy's? Are there any settings on the WIIM and 3e7 I should use?

I recommend getting a Umik-1 with that. Works with a mobile and WiiM app. RoomFit is pretty good.

 

[CptKaos]

I recommend getting a Umik-1 with that. Works with a mobile and WiiM app. RoomFit is pretty good.

the Dayton Audio iMM-6C Calibrated Measurement USB-C Microphone is 1/3 the cost of the Umik-1 and works great with RoomFit

 

[Bleib]

I recommend getting a Umik-1 with that. Works with a mobile and WiiM app. RoomFit is pretty good.

I suppose no need for an Apple phone in this case?

"Using a miniDSP UMIK-1 with the WiiM Home app provides a major upgrade over using your phone's built-in microphone for RoomFit™ room correction. The app natively recognizes the UMIK-1 on most Android devices (and supported iOS setups) and supports importing the mic's custom calibration file"

Checked sources. Yes, should work with Android.

 

[Ruediger]

I suppose no need for an Apple phone in this case?

"Using a miniDSP UMIK-1 with the WiiM Home app provides a major upgrade over using your phone's built-in microphone for RoomFit™ room correction. The app natively recognizes the UMIK-1 on most Android devices (and supported iOS setups) and supports importing the mic's custom calibration file"

Checked sources. Yes, should work with Android.

I don't know if more mics are recognized by the app.

But the Umik-1 works with my tablet and phone (Samsung) incl. calibration files. Compared the results of RoomFit with Umik-1 to REW measurements and what it does is very solid in my opinion.

 

[john61ct]

What is a "good enough" not too pricey way to get overall SPL accurately say x.x dB resolution?

Ideally on an Android phone and storing DPs.

I have UMIK-2 already, rather not fire REW up for just that

 

[Aimarketingbot]

What is a "good enough" not too pricey way to get overall SPL accurately say x.x dB resolution?

Ideally on an Android phone and storing DPs.

I have UMIK-2 already, rather not fire REW up for just that

PEQ'd flat IEMs (e.g. a $10 USD pair of EO-IA500s measured on a clone IEC 711 which are as good or better than 1k+ IEMs in most cases), running via BT off a qudelix or similar, so you avoid room treatment? If you can PEQ off a PC it's basically free if you want to tweak existing measurements.

Aside from that, lots and lots and lots of room treatment and EQ work from speakers that are already very flat (e.g. +/- 3dB or better).

 

[john61ct]

PEQ'd flat IEMs (e.g. a $10 USD pair of EO-IA500s measured on a clone IEC 711 which are as good or better than 1k+ IEMs in most cases), running via BT off a qudelix or similar, so you avoid room treatment? If you can PEQ off a PC it's basically free if you want to tweak existing measurements.

Aside from that, lots and lots and lots of room treatment and EQ work from speakers that are already very flat (e.g. +/- 3dB or better).

Sorry no idea how that relates?

Not talking about EQ at all, nor rooms

just overall SPL measurements, A/B comparisons swapping out speakers / amp / PSU pairings.

If there is good Android software that will work with UMIK-2 then that's all I need?

 

[Roland68]

Sorry no idea how that relates?

Not talking about EQ at all, nor rooms

just overall SPL measurements, A/B comparisons swapping out speakers / amp / PSU pairings.

If there is good Android software that will work with UMIK-2 then that's all I need?

I can tell you from our project experience that microphone-based measurements are simply too imprecise for making comparisons.
Early on, we had a specialist engineer with truly expensive measuring equipment—capable of precision down to three decimal places—working with us. Unfortunately, we later had to admit that it was far too imprecise for comparing devices, such as amplifiers.
A multimeter is much more accurate, faster, and more practical for that purpose.

However, when it comes to loudspeakers, there is unfortunately no alternative to using a measurement microphone positioned precisely at ear level and at a consistent distance.