Modded router + modded switch + Modded access point = OMG

[Kingrex]

Switch X looks very nice. Audiophile are lucky people like you make products for audio use.

 

[luca.pelliccioli]

How do you control your digital player without wifi ?

My iPad is hardwired connected to the router, using also a waversa filter to isolate it. My room is a sort of wifi free zone.

Thank you for your answer, that I’m partially able to understand due to my technical ignorance, for instance I wouldn’t even know how to set my router in “bridge” mode.

I recently invested my money in the NA Tempus since I had the chance to try it at home. I had several conversations with Rob, the NA co-owner, that helped me a lot understanding the product, I have been allowed also to test their Muon Pro filter in my system. The improvements I experienced were easy audible to my ears, this pushed me to make the purchases.
Can you offer a 30 day trial for the Switch X?

 

[Network Acoustics]

The NA Tempus is a old technology in several areas. It also uses internal power regulators based on switching regulation - very typical with these type devices. They do not address the noise AFTER the switched mode PWM regulators on the board feeding the chips. I do. This is a huge difference in performance.

The linear i use has noise canceling using remote sensing. It cancels noise inside the Switch X. There is no other linear doing this. This method results in 20uV RMS of noise on the DC supply and DC regulation at 0.00001Vdc max under load. These numbers cannot be reached without remote sensing. The noise canceling extends to 10Mhz. Its based on a R-Core transformer which is a next gen transformer technology. The Sean Jacobs PSU is old tech and does not use sensing / noise cancellation and because of this it cannot reach the low noise levels under load that a sensing supply can. Its typical of old school 12V supplies of years back. It also uses a toroid VS the new technology R-Core transformer which has much higher isolation.

The combo of the linear I make with a R-Core transformer, sensing 4 wire connection and impedance matching to the exact device makes for a vastly better solution.

You have a switch. The Switch X uses NAT to also work as a router. So in your case you could "bridge" your home ISP router so it does minimal processing and hook the Switch X up directly where it will act as a router.

Features and engineering of the Switch X are covered here. https://dejitterit.com/SwitchXFeatures.htm

The switch X has a 10Gbps backplane and processing power. So it is 10X more powerful and faster then the NA Tempus.

I do not see that the NA Tempus even supports AES67/Dante/Ravenna. Which is really important for the future for super high end audio. The Switch X supports 16 channels of DSD256 via Ravenna using Merging gear in studio work.

I don't see any measurements on the site of performance.

The NA Tempus does not do NAT and so cannot create a clean isolated network, not that you need one in your situation as long as the ONLY devices on this are the your audio devices and a ipad on the wireless. How do you control your digital player without wifi ?. The best way to do this is to have a special wifi access point MADE to work with / around audio gear plugged into the clean side, or at least into your switch.

The NA Tempus, and others, were good tech some years back. There are new generation devices now based on 10Gbps and better power supply solutions. ALso the future involves layer 2 protocols not just a "switch".

I realize there are bold statements. I want to allow the reviews in progress to discuss the Switch X more and then there can be more discussion. The Switch X + dedicated linear is a next gen device on many levels. I have applied all this same engineering to a house router and a DOCSIS ( cable ) modem and a ONT BGW320-505 + SFP+ fiber WAN connection. I have both fiber and DCOSIS here at my lab. So I have addressed the entire path all the way to the ISP fiber or ISP coax.

Once formal reviews come out, I will discuss more.

The Switch X

@Xymox its really interesting to read about what you are doing with Switch X and to get your view of what the future might hold for highend streaming, but you have made a few points regarding our tempus switch hat aren't quite right, so i hope you dont mind us jumping in here with our response:

”The switch X has a 10Gbps backplane and processing power. So it is 10X more powerful and faster then the NA Tempus” -

tempus has a 40Gbs backplane, so based on these on terms tempus is 4x more powerful and faster than switch X.

"The NA Tempus is a old technology in several areas. It also uses internal power regulators based on switching regulation - very typical with these type devices. They do not address the noise AFTER the switched mode PWM regulators on the board feeding the chips"

The onboard voltage regulation in tempus is not based on switched mode PWM regulation. It uses linear regulators that are much lower noise. That’s one of the reasons tempus sounds as good as it does.

Finally just to clarify, tempus is a managed switch that can do NAT and which can also support VLANs as well as other L2 functions. We chose not to implement many of these as currently as this not something we've found any demand for, but also becuase we want keep things fast and powerful, and sounding great, but not technically/mechanically taxing.

thank you and wish you all the best with Switch X.

NA

 

[Xymox]

@Xymox its really interesting to read about what you are doing with Switch X and to get your view of what the future might hold for highend streaming, but you have made a few points regarding our tempus switch hat aren't quite right, so i hope you dont mind us jumping in here with our response:

”The switch X has a 10Gbps backplane and processing power. So it is 10X more powerful and faster then the NA Tempus” -

tempus has a 40Gbs backplane, so based on these on terms tempus is 4x more powerful and faster than switch X.

"The NA Tempus is a old technology in several areas. It also uses internal power regulators based on switching regulation - very typical with these type devices. They do not address the noise AFTER the switched mode PWM regulators on the board feeding the chips"

The onboard voltage regulation in tempus is not based on switched mode PWM regulation. It uses linear regulators that are much lower noise. That’s one of the reasons tempus sounds as good as it does.

Finally just to clarify, tempus is a managed switch that can do NAT and which can also support VLANs as well as other L2 functions. We chose not to implement many of these as currently as this not something we've found any demand for, but also becuase we want keep things fast and powerful, and sounding great, but not technically/mechanically taxing.

thank you and wish you all the best with Switch X.

NA

Interesting.... I love to hear directly from you.

First up.. Oops.. Sorry, i was incorrect.. The Switch X is a 3rd party measured 80 Gbps thruput

Switching

Non blocking Layer 2 throughput

79,946.7 Mbps

77,954.9 Mbps

61,714.3 Mbps

There are full published specs..

CRS309-1G-8S+IN | MikroTik

Desktop switch with one Gigabit Ethernet port and eight SFP+ 10Gbps ports

mikrotik.com

"total switching capacity of 162 Gbps and total non-blocking throughput of 81 Gbps"

So the Switch X, used for pure switching, is twice as powerful as the tempus. And these numbers are not backplane. These are tesed using industry standards thru the ports. "

So the backplane is 162Gbps while this spec is kinda useless, VS real thruput, the backplane is 4 times faster then a tempus. But this kind of spec is kinda meaningless.

So the speeds indicate that all 8 ports can do 10Gbps at the same time. Wireline speeds at 10Gbps. So the switch, even at 10Gbps acts like a wire without delay or limits. So a 1 Gigabit bitstream is nothing for it. This vast overhead is important, as you mention.

What OS does it run ? What chipset ? You have published thruput ? Is it a linux box ? like a NUC ? Just checking.

Speed test should be done to standardised methods. "Xena Networks specialized test equipment (XenaBay),and done according to RFC2544 (Xena2544)"

Do you have a internal pic of the board ? Your using linear regulators, not PWM regulators ? Now lets be sure we are talking about the same thing, the voltage rails to the chips.

Do you have spectrum analysis studies of your power rails ? I dont see those ?

Did you study the impedance of the switch chip supplies ? IE the current spectra. They are of course not DC continuous draw. You have tailored the regulator for each chip rail for the current spectra using the appropriate caps and such ?

So you dont support AES67/Dante/Ravenna then ? The Switch X has landed in some recording studios that do DSD work. So these protocols are required. Of course Merging is present in high end audio and Ravenna is needed there.

Toroid ?

I would love to have more conversation. I would like pics of the internal board and know what chipset your using.

Im not sure how you do NAT with a L2 device ? So you can offer a NATed clean side ? I dont see that in the lit.

Do NOT get me wrong. I have respect for your switch, but, times are changing and you need to be able to do L3 and of course have a very clean parts count and good layout. I assume your not using a CPU and running linux as that would be a very jittery mess VS a dedicated switch chip. The Switch X uses one chip total + ram and flash. Its very clean, I assume yours is also clean as its just a switch.

Yes I would love to see a pic of the board and know what chip is in use.