• 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!

SylphAudio's High-Performance Amplifier Modules (2024 and Beyond)

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Nous le saurons une fois que j'aurai envoyé le FB100 pour le test ASR. Mais c'est environ 90W à 100W.

Le TPA325x se clipse tôt s'il est limité en tension ou en courant. L'utilisation d'une alimentation régulée à haute puissance signifie uniquement que le bloc d'alimentation ne constituera pas un goulot d'étranglement et empêchera l'ampli d'être limité en tension ou en courant. Par exemple, si vous utilisez une alimentation non régulée sur le TPA325x et que vous la conduisez à des niveaux d'écrêtage avec des ondes sinusoïdales, la tension d'alimentation s'affaissera/diminuera, donc elle écrêtera plus tôt. Si vous utilisez une alimentation régulée à faible puissance, soit l'alimentation s'éteindra en raison de la protection contre les surintensités, soit la tension de l'alimentation chutera horriblement (cela dépend de la mise en œuvre). Ainsi, l'utilisation d'un bloc d'alimentation de 400 W sur le FB100 signifie uniquement que l'alimentation ne sera pas la cause d'un écrêtage précoce. Mais la puissance de sortie du FB100 devrait toujours être la même.

Pour les amplis TPA325x avec implémentation PFFB, le « genou » de l'écrêtage se comporte différemment, selon la conception du PFFB, car les comportements près du point d'écrêtage sont différents en fonction de la stabilité du PFFB, donc certains auront un THD+N plus élevé près de l'écrêtage, d'où le « genou » arrivera plus tôt.

Les chiffres 83W/4R et 48W/8R que vous avez fournis sont la puissance « non écrêtée ». Mais si vous autorisez les amplis à écrêter comme sur le CEA-2006/490A, les amplis TPA325x devraient avoir la même puissance de sortie en supposant que les alimentations utilisées sont égales. Et en supposant que l'ampli ne s'éteigne pas.

Pour les modules amplificateurs TPA325x avec LED CLIP/FAULT, une fois que les LED CLIP clignotent légèrement, vous pouvez savoir que le module amplificateur est au niveau du genou de l'écrêtage sans utiliser d'outils de mesure.

Les LED du clip FB100 commenceront à « pulser » à 20 V/4 R, ce qui signifie que le « genou » de l'écrêtage sera d'environ ~90 W à ~100 W. La seule façon de confirmer est d'effectuer des tests de puissance et de distorsion. Le genou d'écrêtage est le point où avant que le THD+N commence à grimper de façon exponentielle. De plus, toutes les charges 4R ne sont pas créées égales :D, certaines ont 4,1 ohms, et ainsi de suite haha. Ainsi, différents testeurs obtiendront des résultats légèrement différents. Même l'utilisation d'une charge de 3,8 ohms par rapport à une charge de 4R aura un impact sur le THD+N. C'est pour cela que nous souhaitons envoyer notre ampli chez ASR pour avoir une référence standard et ainsi tout le reste sera égal.;)
Nice explanation and useful ) Now let s wait for Amir with a concrete test. I cant wait for the review and many thanks for this detailed answer --))
 

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Hi for the followers )

I was able to free up some time to do a more in-depth listen to my new project. So I took advantage of my teleworking to listen to my HD Playlist on my two amps: Micro-Bamboo and ASH-400 (FB-100 Sylph Audio Module)

Me Setup is composed of an SMSL D6S DAC which was measured by Amir and has the advantage of having an XLR output adjustable up to 5V. Note that at this output level, this super little dac almost reaches a Sinad of 124db!

SMSL D6S Review

As for my speakers, I did the test on my excellent Elac FS247.4 measured here. These are very linear speakers with great measurements considering their price.

The ASH-400 + DAC SMSL D6S combo allows me to use the FB-400 module in the best conditions since they have the same output level (as reminder: 14dB (5V/V for the FB100). My old living room is around 30m2 and let’s say it is poorly arranged, the listening conditions are not optimal in my case…..

Quite honestly, my HD Playlist covers almost all styles, from Jazz to Rock and including Pop and Classical) I would be lying to you if I told you that my 45 year old ears are capable of telling the difference between the two amps lol. The fact is that the 3E Audio and Sylph Audio modules are so good at a more than reasonable listening level that they are very difficult to distinguish on the audible audio band . However, I still found that the articulation and dynamism of the Micro-Bamboo at the bottom of the spectrum remains a bit higher and I imagine that this is really linked to the unregulated Micro-Audio PSU. But in all sincerity, I don't really like to comment without tangible measurements because our ears are very approximate measuring instruments). Now regarding power and available reserve, no doubt here : the Micro-Bamboo is really more powerful !

I think it would be fairer to compare the Sylph Audio FB100 to the Topping PA5 / PA5 II which use the same TPA3251 chip. What would be really interesting to understand: is the advantage that the FB100 could have compared to the PA5 which comes with a 144W PSU (126W for the PA5 II). I am impatiently waiting for Amir's measurements to corroborate this point, we will then see on the Power VS Distortion test what the FB100 brings us with a power supply of 350W or more. I'm also waiting to see if the FB100 can exceed 18 bits / 107db on the Multitone ASR test.

My ASH-400 project costs around $280, my Micro-Bamboo project is close to $400. That's a little more than 40% difference for around 70% more power under 4R and almost double under 8R in favor of the most expensive project.
My opinion is the following: if you need power and reserve opt for a project similar to my Micro-Bamboo, if your speakers have a good efficiency / sensitivity and you have a rather small or medium sized listening room, I think the ASH-400 will do the trick. In any case, the latest generation 3E Audio and Sylph Audio modules have an exceptional quality/price ratio. You will be able to adapt the PSU according to your needs with those DIYs projects to play on the price.


1707488584988.png
 
OP
SylphAudio

SylphAudio

Active Member
Audio Company
Forum Donor
Joined
Jan 16, 2021
Messages
192
Likes
289
Location
Philippines
However, I still found that the articulation and dynamism of the Micro-Bamboo at the bottom of the spectrum remains a bit higher and I imagine that this is really linked to the unregulated Micro-Audio PSU.
For me, in general, TPA3255 should have a thicker perceived mid-bass (like bumping up the equalizer on 180hz to 400hz) compared to TPA3251 modules, I believe I mentioned it before in DIYAudio, I just dont remember the exact thread. TPA3251 should have more clarity (less hazy, less veil) compared to TPA3255. (assuming that TPA3251 and TPA3255 are on a similarly designed module).

As for the choice of power supply for TPA325x, my opinion is that the PSU affects the smoothness of the highs. Low registers, not that much unless the PSU is horribly substandard.

****
Update on the sample ASR amp for FB100 and FB360: chassis work is already done, only the wiring remains, then after that some quality checks and THD+N pre-measurements (ETA: 10days)

-Lester
 
Last edited:

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Update on the sample ASR amp for FB100 and FB360: chassis work is already done, only the wiring remains, then after that some quality checks and THD+N pre-measurements (ETA: 10days)
I thought you already shipped it to Amir....
You intend to make pre measurements before as far as I understood ?
So in my humble opinion we will have to wait two months before the ASR review.

Would you share pictures in the meanwhile ?
 
Last edited:

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
For me, in general, TPA3255 should have a thicker perceived mid-bass (like bumping up the equalizer on 180hz to 400hz) compared to TPA3251 modules, I believe I mentioned it before in DIYAudio, I just dont remember the exact thread. TPA3251 should have more clarity (less hazy, less veil) compared to TPA3255. (assuming that TPA3251 and TPA3255 are on a similarly designed module).
As per datasheet. Tpa3255 has a better THD figure under 8R. With 4R loads both are pretty close.

1707842062229.png
 
OP
SylphAudio

SylphAudio

Active Member
Audio Company
Forum Donor
Joined
Jan 16, 2021
Messages
192
Likes
289
Location
Philippines
As per datasheet. Tpa3255 has a better THD figure under 8R. With 4R loads both are pretty close.
Can you give the link of your screenshot? This graph lacks information like missing 8R graph of TPA3251, voltage rail used, etc..
Also, TPA3251 - 10W 4R THD is 0.0035% on that graph.

While On EVM graph below, 10W 4R of TPA3251 without PFFB is at 0.002%, and that is already THD+N (with N) measurement. If it's THD dominated, then THD is at 0.002%, if noise dominated, actual THD value may be lower. So there is some discrepancy on your graph.

Though we can't compare the 3251 vs 3255 distortion on the graph below since TPA3255 might be noise dominated due to a higher supply rail.

Nevertheless, I compared the sound of both TPA3251 and TPA3255 ICs on identical sylph-d200 board before and that's how I perceived the sound differences.

A better SINAD measurement only means that it has a high probability that it will have better measurements on other parameters as well like IMD, multitone, and other parameters that correlate to a better sound.

Also, SINAD / THD score or PvD graph alone cannot tell sound differences. Even amplifiers with identical THD score may have different harmonic distortion profiles, different IMDs, etc..

IMG_7354.jpeg
 
Last edited:

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Can you give the link of your screenshot? This graph lacks information like missing 8R graph of TPA3251, voltage rail used, etc..
Also, TPA3251 - 10W 4R THD is 0.0035% on that graph.

While On EVM graph below, 10W 4R of TPA3251 without PFFB is at 0.002%, and that is already THD+N (with N) measurement. If it's THD dominated, then THD is at 0.002%, if noise dominated, actual THD value may be lower. So there is some discrepancy on your graph.

Though we can't compare the 3251 vs 3255 distortion on the graph below since TPA3255 might be noise dominated due to a higher supply rail.

Nevertheless, I compared the sound of both TPA3251 and TPA3255 ICs on identical sylph-d200 board before and that's how I perceived the sound differences.

A better SINAD measurement only means that it has a high probability that it will have better measurements on other parameters as well like IMD, multitone, and other parameters that correlate to a better sound.

Also, SINAD / THD score or PvD graph alone cannot tell sound differences. Even amplifiers with identical THD score may have different harmonic distortion profiles, different IMDs, etc..
Thanks Lester. I share your arguments. Indeed the graph seems somewhat incomplete. On the other hand, it is taken from the 360 Customs site which has been a reference for a long time and even a precursor in the designs of TPA3255 and TPA3251 modules.

By the way, did you know 360 Customs?

In 2015 they already offered surgically designed TPA3255/51 modules.


You will also find below, and this could help you in your future projects, measurements carried out with a wide range of inductors.

inductor_cmp.png


1707858266422.png

1707858332457.png

1707858558608.png




And here the first TPA325X modules that I owned and which was inspired by the 360 Customs modules



1707859083927.png

1707859232639.png
 
Last edited:
OP
SylphAudio

SylphAudio

Active Member
Audio Company
Forum Donor
Joined
Jan 16, 2021
Messages
192
Likes
289
Location
Philippines
Thanks Lester. I share your arguments. Indeed the graph seems somewhat incomplete. On the other hand, it is taken from the 360 Customs site which has been a reference for a long time and even a precursor in the designs of TPA3255 and TPA3251 modules.

By the way, did you know 360 Customs?

In 2015 they already offered surgically designed TPA3255/51 modules.


You will also find below, and this could help you in your future projects, measurements carried out with a wide range of inductors.

inductor_cmp.png


View attachment 349449
View attachment 349450
View attachment 349452



And here the first TPA325X modules that I owned and which was inspired by the 360 Customs modules



View attachment 349460
View attachment 349463
He's Dr. Mord on DIYAudio right? Not Dr. Mordor (Russian)
PFFB scene on DIYAudio was primitive back then. I think he's inactive even before the pandemic.

I doubt if he really tested those inductors, looks like he just converted the TI's inductor testing SLAA701A into an excel sheet (see attached).
Inductors for SylphAudio's class-d modules were not only selected for the THD+N score, but also for harmonic profiles.

1707895385674.png

1707895354319.png

On the other hand, it is taken from the 360 Customs site
That's why I was asking where did you find the graph since it looks like the image is manipulated. (inconsistent fonts, and formatting, missing values etc..)
 

Attachments

  • slaa701a.pdf
    1.5 MB · Views: 19
Last edited:

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
He's Dr. Mord on DIYAudio right? Not Dr. Mordor (Russian)
PFFB scene on DIYAudio was primitive back then. I think he's inactive even before the pandemic.

I doubt if he really tested those inductors, looks like he just converted the TI's inductor testing SLAA701A into an excel sheet.


Inductors for SylphAudio's class-d modules were not only selected for the THD+N score, but also for harmonic profiles.
DrMordor (Russian alias Vitaly real name) is no more active indeed. I am talking about 360 customs from Germany (almost same name Dr. Mord). The latest has a solid reputation in TPA325X modules.

Anyway Lester ) that was just to share so you can get those inductors measurements from 360 customs--) BTW if you take a look on their site you will find substantial measurements about temperatures, ripple etc..

I like your modules and count on me to get your FB ultra when you re ready :):)
Still looking forward to get your FB100 pre measurements )
 
Last edited:

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Thanks. Keep us informed.
Read from XRK that keeping cables as short as possible improves measurements. BTW when you check the very first ASR review of the Benchmark and the new updated one > this seems to confirm that point.
 

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
@SylphAudio > any updates ?
Did you finally ship the amps to Amir ?
 

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Not yet, just bought Cosmos ADC iso to test it first before sending out. It'll arrive within 2 weeks.
Thanks )
Let us know the pre-measurements please.
 

daniboun

Major Contributor
Joined
May 2, 2020
Messages
1,867
Likes
2,176
Location
France (Lyon)
Not yet, just bought Cosmos ADC iso to test it first before sending out. It'll arrive within 2 weeks.

Any update ? Did you get your Cosmos ADC ?
 
OP
SylphAudio

SylphAudio

Active Member
Audio Company
Forum Donor
Joined
Jan 16, 2021
Messages
192
Likes
289
Location
Philippines
We will revive the production of D100 configured as non-PFFB, since there are some people who still prefers non-PFFB configuration for TPA325x.
Sample feedback below.

1710225064253.png
 
Last edited:
Top Bottom