JCAT Net Card Femto

[Summit]

On 2017-12-16 at 2:52 AM, simonklp said:

I see why it helps to improve the SQ now. Thanks!

 

But I would be more interested to know its effect in a dual PC setup of HQPlayer with NAA.

 

I got better sound by adding the JCAT Net Card in bridge mode in my system with an ultraRendu.

[Summit]

On 2017-12-20 at 7:32 AM, simonklp said:

Thank you very much for your information.

 

But is there also any information about how the effect of this JCAT NET Card on SQ when compared to an optical network via a optical switch between PC with HQPlayer and the NAA?

 

The reason why I ask this is because I think that the effect of the NET Card is to reduce both jitter and electrical noise. But for the optical network, it can also reduce electrical noise by providing galvanic isolation. For jitter, some of the CA members claim that optical network induced additional jitter.

 

Therefore I am interested to know if this NET Card can improve the SQ when compared to the optical network. Thanks.

 

 

 

The JCAT NET Card did reduce electrical noise in my audio chain that already has galvanic isolation both in the DAC, DDC and in the ultraRendu. With the JCAT NET Card the sound got notable clearer than while using the input and output on the mobo. With clearer I do not mean the sharper and more edgy kind of sound that you sometime get with some gear. It can sometime be a bit difficult to distinguish real reduction in electrical noise VS a more analytic and sharper sound, but the difference is really big in the long run.

 

With the JCAT NET Card the biggest difference was the clearer sound, other notable differences was enhanced ease and a bigger and more effortless presentation. The sound is not soft, it’s just flows with more ease and like @elan120 said the sound is more alive.

[Summit]

On 2017-12-20 at 3:31 PM, tboooe said:

 

 

I do not have a NET card (yet) but something happened over the weekend that got me to seriously consider getting one.  I currently have a dual pc setup with HQP and NAA.  I also use a dual fiber nic operating in bridge mode in my HQP pc and 2 FMC, 1 between my wireless bridge and HQP pc, and 1 between the HQP pc and NAA pc.  Here is the network chain:

 

Wireless bridge >> CAT 6 cable >> FMC >> fiber cable >> Port #1 of dual fiber NIC >> Port #2 of dual fiber NIC >> fiber cable >> FMC >> CAT 6 cable .> NAA pc

 

This weekend I had a little electrical mishap that blew out the downstream FMC so in order to keep playing music, I had to remove the FMC and bridge the mobo ethernet port of the HQP pc.  So my network chain became:

 

Wireless bridge >> CAT 6 cable >> FMC >> fiber cable >> Port #1 of dual fiber NIC >> mobo ethernet port >> CAT 6 cable .> NAA pc

 

In this configuration, I was amazed at how much more dynamic and energetic my system sounded (as Marcin described above).  This change has got me seriously considering removing the dual port fiber nic and replace it with another standard gigabit nic like the NET card.  I would then bridge it with the mobo ethernet port and use it between my HQP pc and NAA pc.   It would be ideal though if the NET card was able in a dual port configuration.  @Marcin_gps any plans for that?

 

@Summit I assume you used the NET card between your PC and Ultrarendu in bridge mode and the mobo ethernet port to connect to the rest of your network?

 

 

I use the NET card in bridge mode. I have tried other ways of connecting it and liked the bridge mode best with my uR.

 

I have never had any FMCs in my audio system and believe that the gain you my get from the fiber insulation (in theory) is small compared to the loss of using more cheap clocks and switching regulator etc that is often used in FMC.     

[Summit]

On 2017-12-22 at 3:19 AM, simonklp said:

 

I am using fibre NIC instead of FMC on both the PC with HQPlayer and the Audio PC for NAA that is powered by a LPS. In other words, the fibre NIC on NAA draws the power indirectly from LPS.

 

For the clock on fibre NIC, I am not sure it is cheap or not. But it is another discussion about how much this clock affects the SQ, given the fact that asynchronous FIFO of NAA paired with HQPlayer provides maximum isolation between processing and audio reproduction which also achieves via asynchronous USB playback by DAC.

 

This is something that I am interested to know. Thanks.

 

I can only tell you what I have experienced and it is that reclocking and FIFO buffering both in the DAC, DDC and in the ultraRendu take you a long way, but what’s upstream still has a notable impact. Your use of fiber NICs is more or less the same approach as going with a NET card and shilded LAN cables, I think. Ultimately I would speculate that which design that sound best depend on implementation and components used in those cards (PSU, voltage regulator, clock etc).  

 

Btw by placing ferrets on my SPDIF cable the sound cleans up even more and make it more analogue.

[Summit]

2 hours ago, mourip said:

I have the new card installed in my Roon Core server. My Roon endpoint is a NUC running AudioLinux(AL).

 

Out of the box the card offered an improvement. I spent some time tweaking the NIC settings in my network bridge configuration and then listened for a while. At some point I had a Oops! moment. I had forgotten to hook up the 5v connection to the card from my HDPlex LPS so I had been running on PCI bus power. Once connected, there was a very clear change. "Clear" is the best word to describe it.

 

I will not go farther into sound quality changes now since the card is too new but I can already say that the card is worth it.

 

One caveat or rather a pointer. I have an ethernet cable going directly from my NUC to the JCAT NIC in my server. The JCAT is bridged with my onboard NIC, which is connected to my LAN. When I started making the suggested changes to my bridge network settings, each time I made a change(s), and then saved it, the connection to my NUC would drop briefly. It would return quickly but Roon would stop on the endpoint and the NUC would stop showing as available in Roon. The fix was easy but it took a while to find. With a connection established to the NUC using Putty, all that I needed to do was to stop audio services in AL and the restart Roon Bridge. I had to do this multiple times while making and testing the network config changes.

 

Yes clearer is how I would describe the effect of using the JCAT NET Card too.

 

https://audiophilestyle.com/forums/topic/37151-jcat-net-card-femto/page/2/?tab=comments#comment-756669

 

[Summit]

22 hours ago, mourip said:

 

Based upon Marcin's recommendation of only using one of the JCAT ports when bridging I am doing that and using my onboard NIC for the LAN connection. I was using FMCs to go from my onboard NIC to my LAN switch and powering the clean-side FMC from an LPS. That worked fine until recently but for some reason I cannot get my FMC's to link up.

 

Are you isolating your network bridge on the LAN side?

 

 

Sorry am not sure I understand your question. If you’re asking if I use fibre optic, the answer is no. I have LPSU on both the LAN card and switch though.

[Summit]

1 hour ago, mourip said:

 

I was wondering if you were using any method of isolating your audio system grounds and power from the rest of the LAN and AC wiring.

 

I use an large isolation transformer with all audio equipment plugged into the same AC power strip. I use fiber to isolate from the connection out to the LAN. 

 

I did note that you were not using fiber and I agree with or at least understand your reasoning.

 

Thanks!

 

Yes indeed, I believe it’s paramount to separate the audio system from the noise coming from the computer, router and mains power. I therefore have my computer with JCAT Card + LPS in another room and on another power line than there my uR and stereo is. I have two power conditioner, one Gigawatt PC3 EVO for my DAC, uR, amp and DDC and an Oehlbach XXL Powerstation 909 for my router and switch. Both power conditioner has separated/isolated outputs for digital and analogue devices.

 

I like to joke a bit with Jesus at Sonore, but the truth is that if they would manage to make a good opticalModule without the cheap clocks, switching regulator, low budget PSU etc that is normally used in FMC, it could probably be a great device. Stopping the noise as early as possible is normally best.