Ethernet Cables and Sound

Here’s a picture of the standard Jcat femto fed by a Farad. It’s hard to say, but I think the horizontal lines are a bit less good as with the modified Jcat and maybe a bit less steady

Here’s a picture behind an 100mbit optical adapter fed by a Farad. I think it’s dancing a bit around the baseline and it’s less stable. The optical network is behind an AQvox SE switch. The optical network is connected with my streamer, but this with a Cat6 copper cable


This is a picture behind the AQvox SE switch. It’s feeding the optical network.What I see isn’t much better than after the optical adapter. I think it’s less good, but the optical network already has a better start then a direct connection from the router

Here the same position as the picture’s above with the AQvox SE switch, but now with a netgear switch fed by a high grade linear powersupply.

My conclusion is that even the tiniest improvements of the digital signal has to our ears major impact on the sound. The differences in the digital signal seems to be small. The effect of a lineair powersupply is measurable. The effect of a better clock is measurable, but it’s very hard to point out the mearurable effect of a better ethernet cable. Still I think the differences of cables are easy to hear, but they are far less important to the sound than the good powersupply’s, clocks, opticalconverters, and switches.

Hope these measurements will help us understand the importance of a good digital signal to the sound!

1 Like

In my experience this is indeed the case, which led to an almost endless series of upgrades…

PF LAN-isolator, EtherRegen (UpTone UltraCap 1.2), AQ Diamond RJ/E into Melco, AQ Diamond USB to Matrix x-spdif 2 fed by upgraded Farad Super3 and AQ Diamond HDMI into the DS.

No regrets though, each and every upgrade was noticeable and the total result is more than the sum of its parts.

1 Like

Thanks for the measurements. They surely show some “digital” differences. By chance, did you measure the analog output of the dac, to see if/how such differences manifest themselves?

1 Like

Hello Wijber, nice to hear. It seems you have an ultimate digital solution. I hope to get an etherregen as well. I’m curious how it will measure with and without an external clock.

Hi Jvvita, yes it does! Especially the Jcat with the ultra clock shows a very nice signal. I’m happy it is clearly measurable I hope to do the analog measurements soon. Last year I didn’t succeed to point out incontrovertible evidence of what we hear in a measurement. Again I have high expectations.

Hoi Wijnand,

Not (yet…?) ultimate but very happy with it. Did consider an external clock but hey, than comes LPS, decent cabling…How many clocks does one need? (ER, Melco, Matrix, DS).

However still very curious about possible added value using an external clock for the ER, so maybe next time…:sweat_smile:

Eagerly waiting for those curves…

1 Like

Well that actually is a good question :slight_smile: There’s no need for extra clocks when a system already works. Besides this I even think it’s not the precise timing alone of the better clocks, but the lower phasenoise on low frequencies as well. Normally a simple crystal oxylator is used in many occasions for switches and routers and in premium audio equipment you will find TCXO clocks.

The ultra OCXO clock runs with a preciseness of 60 femtoseconds and the phasenoise is already at 10Hz around 130DBc/Hz and at 100hz already at 145DBc/Hz. This is realy great!

I replaced 5 clocks for Pink Faun ultra OCXO clock’s 3 in my streamer and 2 in my dac. You can see in the picture what it does to the signal of the Jcat. After the Jcat the stream will find one extra clock on the networkcard of my dac (this was a normal crystal oxilator). On the mainbord of my dac I replaced the dac TCXO clock for the Ultra OCXO. So the digital signal is as clean as possible and it took my system to the next level!

1 Like

here they are! The analog pictures. Above is conditioned Below is without. you can doubleclick the pictures

At 50us

At 50mV and 500nS. Sorry I should have lined them to the same position

I think we should look to pictures with a bit more time so 500uS
Sorry the pictures aint levelled out on the baseline. I only captured two curves. The differences are to small to see. We can’t hear differences in microseconds and we can’t see them either :grin:

Nice effort. But I fail to see any difference. Are you able to null them? The differences (if any) should stand out.

1 Like

Thanks! Yes I can do that. What we can learn about these pict’s is that it’s not noise on the waveform. So it must be the jitter. I think it’s hard to detect jitter in microseconds, but when you look into the first picture (at 2mS) I can detect some time differences and tiny differences on “spikes”. What i’m going to do is. Picture 3 waveforms conditioned and 3 waveforms unconditioned. When I see those differences in all pictures then we have something.

1 Like

I looked hard! There’s nothing to find… not even on 50ms… if you think you can find something than it’s probably the pixel difference