• 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!

[Gallium Nitride amplifier] Impressions of the Mini Gan 5 by Premium Audio, pictures / video inside

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,479
Likes
25,222
Location
Alfred, NY
Agree. But question for you : Do you think the Purifi is transparent ? The same for the TPA3255 and for the MA12070 ?
Do they have flat frequency response and low distortion?
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire
Agree. But question for you : Do you think the Purifi is transparent ? The same for the TPA3255 and for the MA12070 ?
Purify measured at as close to transparent as I've seen. FR dead flat, and distortion inaudible to most (certainly me), especially with pregain off. I've no knowledge of the others you mention.

The point is though - if people are "praising its tonal qualities" (And they are not just hearing stuff that isn't there) then they're not praising a transparent amp.
 

Phorize

Major Contributor
Forum Donor
Joined
Apr 26, 2019
Messages
1,539
Likes
2,071
Location
U.K
Agree. But question for you : Do you think the Purifi is transparent ? The same for the TPA3255 and for the MA12070 ?

Amirs measurements of the purifi show it to be transparent. Not sure about those op amps. Have you got the manufacturers specs?
 

Bob from Florida

Major Contributor
Joined
Aug 20, 2020
Messages
1,286
Likes
1,180
[QUOTE="tonycollinet, post: 915165, member: 36536"

The point is though - if people are "praising its tonal qualities" (And they are not just hearing stuff that isn't there) then they're not praising a transparent amp.[/QUOTE]

Are you sure about the intent of the OP's description. If I were to say "I appreciate the tonal response of XYZ" that would be a way of saying it sounds "right" without added emphasis anywhere. Now that could be because all the components are transparent and the speakers plus room don't screw things up too much. It also could be something complimentary is happening. Either way measurements could shed light on why. Sighted listening differences is even less accurate if one knows the measurements before hand. Just another form of expectation bias - knowing how a device measures can predispose the conclusion. If you were to audition this gear, would reading this thread or a review by Amir affect your perceptions?
 

ta240

Major Contributor
Joined
Nov 7, 2019
Messages
1,417
Likes
2,840
Getting a load of opinions won't do anything to get rid of subjectivity. All it will do is bring the subjectivity of a load of random strangers into play.

This is not what ASR is about. This whole thread is the exact opposite.

Measurements, and Blind listening tests at your showroom - or you might as well go to a different forum. You are only achieiving turning people off the product here.

A quick trip to the review section and I find nothing of blind listening tests. It seems this person is being held to a different standard.
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire
A quick trip to the review section and I find nothing of blind listening tests. It seems this person is being held to a different standard.

Measurements are carried out here. Any listening comments are clearly labeled as subjective and are pretty much downplayed.

Every poster that promotes subjective hearing characteristics (airy, open, nuanced, dynamic, forward, warm, PRaT etc etc ad infinitum) is told to come back with blind test results if they want to be taken seriously, as you will see if you review a few posts from this search:
https://www.audiosciencereview.com/forum/index.php?search/26847067/&q=blind+abx&o=relevance

Human hearing is variable and always influence by what the listener knows about what they are listening to (expectations).
 

Phorize

Major Contributor
Forum Donor
Joined
Apr 26, 2019
Messages
1,539
Likes
2,071
Location
U.K
A quick trip to the review section and I find nothing of blind listening tests. It seems this person is being held to a different standard.

I can see that you are meaning to be fair, which is laudable (I mean it, not patronising you) but I don’t think the OP is being treated badly or unfairly. He provided some fairly useless information on what he thinks the amplifiers may sound like, with good intentions no doubt. The responses simply pointed this out. OP then doubles down, very politely, so he was told again that measurements and controlled testing are informative, and subjective descriptions based on sighted listen generally are not. I don’t see a problem.
 
OP
daniboun

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,874
Likes
2,181
Location
France (Lyon)
Hi amigos

Got the answer about the heatsinks... Here what has been said by Premium Audio :

"The inductors were changed because as most know, there is a world-wide chip and parts shortage. The original coils we used were back ordered for months. These toroids have the exact same performance as the others we used. We did just got a few thousand pieces in (that were ordered for months). The toroids actually run a little cooler and performance is exactly the same. As for the as you called it flying heatsinks, we originally had them mounted on the FETs with thermal pads and other, but found the amp actually runs a little cooler like this. The heatsink raised like this draws some heat from the PCB and the FETs can run at rated power without heatsink. "
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire
I am sorry - but the heat sinks raised like that draw NO heat from the PCB. Without thermal contact, the only way heat can be drawn away is convection directly from the power devices, then with the heat sinks in the way from air to the flat surface of the underside of the heat sink - then to be given back to air on the top side.

That is total nonsense: Heat transfer from air to metal is very inefficient which is why fins are used to increase surface area. Far more effective for convection to be allowed to continue without the barrier of the heat sink. Even as it is, disrupted convection around the heat sink will carry more heat away than conduction from air to heat sink to air.

I also don't believe the heat sinks were anything other than a visual con. As they've said, the fets don't need them. This would have been clear during development even if they were thermally connected.

Premium are obfuscating, and the glue between the chokes and the heat sink tell a different story.

I don't have any issue from a product quality point of view with the glue as long as they have good manufacturing controls for its application. But then all they need is a mechanical bracket to support the chokes which could be made of plastic to eliminate any magnetic interaction, and insulation risks. Further, a plastic bracket could incorporate clips around the chokes enabling them to eliminate the glueing process.

I do have an issue with the apparent dishonesty of putting a heat sink shaped object in the product for no reason other than - what?
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire
I could make an alternative and really generous analysis that would go like this.

  • Heat sinks designed in as expected to be needed.
  • During development its found out they are not needed, but there is insufficient time in the project for a design change to remove them, and the amp works fine with them - so they are left in place.
  • Mod needed for chokes due to shortage reasons (or cost - nothing wrong with design to cost)
  • Vibration problems due to new chokes only supported by wire legs, which fracture. If a quick fix is needed (eg original choke supply running out) then that can be achieved as they have done - simply with longer posts to support the existing known not to be needed heat sink, and glue applied.
Sometimes engineering to a time constraint results in compromise. Very often supply chain problems - as the whole industry is experiencing - result in compromise.

Hopefully if that is the case, they are also planning a medium term fix with a better suited (lower cost and process improved) support bracket to replace the sticking plaster.

But also - if that is the case - come clean and state it. Don't treat us like idiots telling us a lump of metal placed like that in the convection airflow can improve cooling - it can't.

@daniboun - not aimed at you here, I'm not interested in shooting the messenger. And thanks for your efforts to get info from the manufacturer. :)
 
Last edited:

Bob from Florida

Major Contributor
Joined
Aug 20, 2020
Messages
1,286
Likes
1,180
I have seen a lot of chokes like these hot glued together or glued to something nearby - almost seems to be standard practice. My interpretation of the heat sinks is based on the assumption that given explanation is accurate and as follows. OEM measured temperatures on transistors and circuit board under 3 conditions - no heat sink, heat sink in contact via thermal pads, and heat sink raised. According to the explanation lowest temps for transistors and the board were achieved with heat sink raised. The heat sinks are coupled to the board by 2 standoffs and screws - from the picture both have appearances of aluminum construction which would satisfy no magnetic coupling. A direct mechanical connection from heat sink to board does provide a thermal path for cooling via the heat sink. How much cooling without the comparative measurements is unknown but some cooling of the circuit board would be reasonable as a result.
 

sq225917

Major Contributor
Joined
Jun 23, 2019
Messages
1,369
Likes
1,640
Actually, the placement of the heatsinks could act as a chimney to force greater heat exchange via convection by entrancing the airflow.

Note the use of the word, could...

Much more likely they are just too lazy to remove them. and they remain as a visual conceit only
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,670
Likes
38,765
Location
Gold Coast, Queensland, Australia
A transparent amplifier doesn't have tonal properties. That is what transparent means. Flat frequency response, no distortion.

No. Absolutely incorrect. Transparent means you can see through it. Light. Not sound. Not electricity and certainly not audio equipment. The term is not remotely applicable to HiFi gear and never will be.
 
OP
daniboun

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,874
Likes
2,181
Location
France (Lyon)
About Heatsinks > that seems to say a lot about the honesty of the manufacturer ... what a pity !
However, the amp sounds wonderful and I get used to it) But if I were in the place of ClassDAudio, I will review my specifications or I would have announced it from the start at least ... (I mean about the PCB upgrade)

In any case, even after listening at sustained volume for a good 4 hours, it heats very little, which reassures me. ouffffff
 

EB1000

Senior Member
Joined
Jan 8, 2020
Messages
484
Likes
579
Location
Israel
If you don't take the full advantage of high switching frequencies (>1MHz), then using GaN FETS will give you no advantage besides a slight efficiency boost, due to lower Rds-on. It will not add more "details" or improve any audio parameter for a switching type of an amp... BTW, are those actually gold plated heat sinks??? I just hope you didn't use audiophool-grade contact cleaner instead if a thermal paste to attach those floating heat sinks...
 
OP
daniboun

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,874
Likes
2,181
Location
France (Lyon)
@daniboun - not aimed at you here, I'm not interested in shooting the messenger. And thanks for your efforts to get info from the manufacturer. :)

What you have to understand is that I have two interlocutors on the manufacturer side) we will see if the arguments of Premium Audio converge with those of ClassDAudio (which owns the brand)... I asked the same question to ClassDaudio, let's wait now !)
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,589
Likes
12,760
Location
UK/Cheshire
If you don't take the full advantage of high switching frequencies (>1MHz), then using GaN FETS will give you no advantage besides a slight efficiency boost, due to lower Rds-on. It will not add more "details" or improve any audio parameter for a switching type of an amp... BTW, are those actually gold plated heat sinks??? I just hope you didn't use audiophool-grade contact cleaner instead if a thermal paste to attach those floating heat sinks...

I think they are gold/yellow coloured anodised. No gold contained:
https://alumet.nl/en/colours/alugold/
 
OP
daniboun

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,874
Likes
2,181
Location
France (Lyon)
Top Bottom