- Dieses Thema hat 8 Antworten sowie 2 Teilnehmer und wurde zuletzt vor vor 4 Tage, 15 Stunden von
aktualisiert.
-
Beiträge
-
Hello éric and all,
It’s been a month or two that i discovered this website and I come now to share the result of my readings.
First, as i have no knowledge of German and a basic English, this is a Google translation that is worth what it is worth.
For information, my Hifi system start this way : Internet box connected in RJ 45 to an low latency Ethernet switch D-link 106XT connected in RJ45 to two audio PCs (Roon Server one and HQPlayer DSD 1024, the other) Everything is powered by a linear diy PSU
So, i begin with the Intel I350 T2 Net card.
First i install it in the server pc but with no audible difference 🙁 But in the player pc, it’s not the same, big surprise :-)) It reminded me, my feeling after installing the Sotm card, a nice sound improvement and my SoTm card is with his « Superclock »
Then i received the TP-link 902AC V4, i powerbanked it, pluged it in the PC player and it’s sound better, very nice again Pluged in PC server, sound improved but a little less than with PC Player. As i have an LVDOVR LT3045 3,3V 1A card in reserve, i make in stride the hardware modification, and as expected the sound improve again.
Then, i buy another Intel card for the PC server (an used one is worth in France around 30/50 €) and plan to brigde the two RJ45 of the card and link the two PC together with an « choke » cable.
Like that, the sound keep these previous improvments ( dynamic, very clear, stronger bass, accurate sonore space, it’s like to see with a magnifying glass).
So i withdrawn the Ethernet switch of my system and i don’t plan to connect the PC player with a second TP-Link („less is more“ is the goal).
For the choke cable, i begin with a V1 and sound improve again with it, and i make after a more sophisticated one (see picture) but i am not sure to heard more improvements with it.
And to finish, as i have a TCXO board in reserve, i changed the quartz clock of the PC player Intel Net card with it. And it’s worth it ! Sound improved again 🙂
I just desoldering the original quartz and weld a pine (in picture) to plug the new clock board. ( i add welding before)
I plan now to try this kind of OCXO clock board to synchronized the two Intel Net board with it : https://fr.aliexpress.com/item/1005008554739836.html?spm=a2g0o.detail.0.0.1476t2GUt2GUPE&mp=1&pdp_npi=5%40dis%21EUR%21EUR%20107.10%21EUR%2096.39%21%21%21%21%21%40211b655217444887224124000e51f8%2112000045690547826%21ct%21FR%216205309863%21%211%210&gatewayAdapt=glo2fra
To conclude, it is clear that the Ethernet links were the weak point of my system which made a bound forward.
Many thanks éric
<span style=“vertical-align: inherit;“><span class=““ style=“vertical-align: inherit;“>un de plus</span></span>
Hallo Jean Pierre,
vielen Dank für den wirklich interessanten Beitrag!
Ich bin bzgl. Clock Updates etwas vorsichtig. Ich habe aktuell an dem Adum4166 Board die Stellarz clock angeschlossen. Allerdings sieht man sehr starke Störungen im Gleichtakt durch das Rechtecksignal der Clock in der Frequenz der Clock und den harmonischen Oberwellen.
Beste Grüße,
Eric
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>Bonjour Eric et tous.</span></span>
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>Eric, j’avais lu votre article sur les horloges. Vos mesures mettaient clairement en évidence le bruit de leurs harmoniques dû à leur signal de sortie carré. Je ne suis donc pas surpris de vos réserves concernant le changement d’horloge :-)</span></span>
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>Mais comme j’avais déjà un TCXO „low jitter“ de Paul PANG Audio à disposition et que les avis sur le remplacement d’un XO à cristal passif sont généralement positifs, j’ai voulu l’essayer et c’était bénéfique à l’écoute.</span></span>
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>Peut-être que les meilleures spécifications (gigue, bruit de phase, stabilité de fréquence et sa dérive annuelle) du TCXO / OCXO produisant un signal carré l’emportent sur l’augmentation du bruit en mode commun ?</span></span>
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>Sinon, je corrige mon premier post après avoir pris le temps de comparer le rendu sonore des connexions Ethernet entre mes deux PC et la box internet :</span></span>
<span style=“vertical-align: inherit;“><span style=“vertical-align: inherit;“>La meilleure configuration consiste à connecter le PC HQplayer au Wi-Fi de la box via le D-Link 902AC, puis le PC serveur à la box via un câble Chock. Finalement, une carte Intel sur le PC player seul peut me suffire, car je n’étais pas sûr d’entendre une différence sonore avec une carte Intel sur le PC serveur.</span></span>
<span style=“vertical-align: inherit;“><span class=““ style=“vertical-align: inherit;“>J’aimerais maintenant alimenter la carte Intel I350 directement via la batterie et des modules LDO comme le D-Link 902AC. En fait, je note sur la carte : 4,4 V en L3, 3,28 V en L2 et 0,99 V en L1, et 5 W sont nécessaires. </span></span><span style=“vertical-align: inherit;“><span class=“VIpgJd-yAWNEb-VIpgJd-fmcmS-sn54Q“ style=“vertical-align: inherit;“>Quelqu’un a-t-il approfondi ? Eric ? </span></span>
<span style=“vertical-align: inherit;“><span class=“VIpgJd-yAWNEb-VIpgJd-fmcmS-sn54Q“ style=“vertical-align: inherit;“>Cordialement </span></span><span style=“vertical-align: inherit;“><span class=“VIpgJd-yAWNEb-VIpgJd-fmcmS-sn54Q“ style=“vertical-align: inherit;“>Jean-Pierre</span></span>
Hello Eric and all.
I had read your article on clocks, your measurements clearly highlighted the noise of their harmonics due to their square output signal. So I’m not surprised by your reservation about changing the clock 🙂
But since I already had a Paul PANG Audio „low jitter“ TCXO available and the reviews on replacing a passive crystal XO are generally positive, I wanted to try it and it was beneficial to listening but i don’t no if the output signal of these clock is square or sin.
Perhaps the better specifications (jitter, phase noise, frequency stability and annual drift) of TCXO / OCXO outputting a square signal outweigh the increase in common mode noise ?
Otherwise, I’m correcting my first post after taking the time to compare the sound rendering of the Ethernet connections between my two PCs and the internet box :
Best configuration is to connect the HQplayer PC to the box’s wifi with the D-Link 902AC and connect the server PC to the box with a „chock“ cable. Finally, a single Intel card on the player PC can essentially be enough, I wasn’t sure if I would hear a difference with an Intel card on the server PC.
What I would like now is to power the Intel I350 board directly through LDO modules.
All I have been able to do so far is to note on the board: 4.4V at L3, 3.28V at L2 and 0.99V at L1.
Eric, have you thought about it?
Has anyone gone further?
Best regards
Jean-Pierre
Hallo Jean-Pierre,
ich bin nicht grundsätzlich der Meinung, das ein Clock-upgrade nichts bringt – ich habe selbst die Mutec an dem Etherregen positiv wahrgenommen.
Bzgl. der Stromversorgung bei der i350 habe ich noch keine Anstrengungen unternommen, da sie aktuell bei mir ein untergeordnete Rolle spielt – aber ich wäre auf Ergebnisse hiervon gespannt.
Beste Grüsse,
Eric
- Du musst angemeldet sein, um auf dieses Thema antworten zu können.