Meridian Unplugged - Powered by MrTechGuy.co.uk Forums Discussion HiFi & HT Chat Cleaning the bits

Or, given that the OP has already got Muon in his system, then save the pennies and go for broke on buying the Tempus with integral hybrid power supply to replace his Netgear.

It kind of looks like there is a second case build around an other one. I wonder what's underneath it...
It's probably very common to improve an common switch and sell that as an audiophile version.

The Silent Angel Bonn N8 (400 euro) which has a small tcxo board with two crappy capacitors which should improve jitter performance. If you want to upgrade those crappy capacitors with better ones then kemp electronics can help, they also change 4 other capacitors and stick a piece of bitumen on the tcxo for the measly sum of 150 euro. You want some more resonant resistant feet under the device, 50 euro. Then there is the 670 euro English Electric 8Switch which is internaly the same as the Bonn 8. But with a better shielded housing. It seems to be a modded ZyXEL GS-108B V3 costing about 30 or 40 euro.

Needless to say they advise you to invest a few 100 more for a better suitable power supply....

And there is de Aqvox 800 euro audiophile switch its a kind of annoying video and they don't have a lot of knowledge but they take it apart to show what is done to the stock switch

Dear all, thanks for your helpful comments and ongoing suggestions.

Just to rehearse the general idea, according to my ears airborne time signals capture the amplitude and timbre of my sound world. Instantaneous amplitude is coded as bits in the digital signal. The more bits per sample, the more amplitude levels.
That amplitude information in the time signal is sufficient to capture sound timbre if it is sampled at a high enough rate. (Amazing). I think I now see that's where (electrical) noise added before the DAC can mess with sound quality by ghosting the timing of samples. Is this making sense?

So it's not about bit coding, but it is more about keeping noise out. I suppose this kind of noise can be in the air or in the power supply. Some time ago someone at Meridian told me the MS600 is slightly better than the 218 due to its slightly better power supply. Later some folks took the brave step of modifying their 218 with a much better power supply, (Farad 3 I think), and reported good results. So I was thinking a better switch might have a better power supply, better clock, better shielding, etc...

thanks again all,
Sprongle

Yes that's the Farad 3, The standard power supply of an 218 is a very small switching 12 volt supply. The one in the MS600 is a bigger switching PSU and is the same as used in the G68 among many others. They put out multiple voltages, +12, -12, +5 volts and ground. Meridian has sold linear power supplies in the past as an upgrade for the G68 and 800 series which gave a positive improvement.

The switching power supplies produced a lot of heat in the g68 giving al kind of problems and the PSU's died regularly.

"So I was thinking a better switch might have a better power supply, better clock, better shielding, etc..." It probably has. But the power supply is normally an outside the switch one, so you cab easily experiment with that on your cheap switch

Yes, OP has apparently drank the Digital Noise Cool Aid and no longer believes that Bit Perfect is all you can do. Now consider what David Snyder has to say over at the Roon forum:

The Roon Paradox: A Powerful Experience with a Sonic Caveat

Roon is celebrated, almost universally, as the most powerful and engaging music management system available. Its rich metadata and seamless user experience are second to none. However, this functional supremacy has long been dogged by a persistent critique from a vocal segment of the audiophile community: that Roon’s sound quality can be compromised, often described as “flat, dull, and lifeless” compared to other players.

This “Roon Sound” isn’t a myth, nor is it a flaw in Roon’s bit-perfect software. It is a potential symptom of Roon’s powerful and resource-intensive nature. Roon’s “heavyweight” Core requires significant processing power, which in turn generates electrical noise (RFI/EMI). When the computer running the Roon Core is in close proximity to your sensitive Digital-to-Analog Converter (DAC), this noise can contaminate the analog output stage, masking detail, shrinking the soundstage, and robbing the music of its vitality.

Moving Beyond “Band-Aids” to a Foundational Solution

Roon Labs itself advocates for a “two-box” architecture to solve this primary issue: separating the demanding Roon Core from a lightweight network Endpoint (also called a streaming transport). This is the correct first step, as it offloads the heavy processing to a remote machine, isolating its noise from your audio rack.

However, even in this superior two-tier design, a more subtle problem remains. Standard network protocols, including Roon’s own RAAT, deliver audio data in intermittent “bursts”. This forces the endpoint’s CPU to constantly spike its activity to process these bursts, causing rapid fluctuations in current draw. These fluctuations generate their own low-frequency electrical noise right at the endpoint—the component closest to your DAC.

High-end audio manufacturers attempt to combat the symptoms of this bursty traffic with various “Band-Aid” solutions: massive linear power supplies to better handle the current spikes, ultra-low-power CPUs to minimize the spikes’ intensity, and extra filtering stages to clean up the resulting noise. While these strategies can help, they don’t address the root cause of the noise: the bursty processing itself.

Audiolinux: This is a purpose-built, real-time operating system designed specifically for audiophile use. Unlike a general-purpose OS, it’s optimized to minimize processor latencies and system “jitter,” providing a stable, low-noise foundation for our endpoint.

Diretta: This groundbreaking protocol is the secret sauce that solves the root problem. It recognizes that fluctuations in the endpoint’s processing load generate low-frequency electrical noise that can evade a DAC’s internal filtering (as defined by its Power Supply Rejection Ratio, or PSRR) and subtly degrade its analog performance. To combat this, Diretta employs its “Host-Target” model, where the Host sends data in a continuous, synchronized stream of small, evenly spaced packets. This “averages” the processing load on the Target device, stabilizing its current draw and minimizing the generation of this pernicious electrical noise.

I happen to use the Farad Super3 for the last device (USB-to-AES converter) into my DSP9 speakers via Network Acoustics M-Link filters.

XE Happiness 😋😋

Dear oldDogCan,

what an interesting post! You have thought about this long and hard. Can I ask a couple of questions to clarify please?

Having drunk the Cool Aid am I safe to visit our local laundromat? Sorry couldn't resist. What I mean is: have I got it? Are we on the same page?
What's the "lightweight network Endpoint (also called a streaming transport)"? In my modest setup is that the MS600, or is it the Roon remote (client)?

thanks,
Springlike

Hello Springlike. Yes, we could be on the same page. I feel that my own DSP5200SE speakers crossed a line where they finally disappeared altogether into a big 3D soundstage once I put them on Gaia II feet, gave them decent Shunyata power, and implemented a couple of network changes, including a singlemode fiber segment. I didn’t add the Diretta endpoint for Roon until early 2024, which brought everything into a sharper focus

I never owned the MS600 or used the Sooloos software, but understand that you would need to replace it as a streamer with something that can incorporate the Diretta protocol to get SQ benefits. A Raspberry Pi 4 or 5 can be made to output SPDIF or AES which your DSP5200SE can use. MQA is on the menu, too. The problem for Meridian owners is that we like to use SpeakerLink or BNC Comms to control all aspects of music playback, whereas I now have only volume control and on/off using the remote unless I temporarily reconnect an M controller to adjust something like balance, for example.

The Diretta protocol was developed in Japan by otherwise smart people who have so far done terrible jobs with documentation and marketing for us English speakers. So the information I have is gleaned from a couple of Linux software partners: GentooPlayer and Audiolinux. You can buy standalone hardware devices from them or install licensed software on the RPi like I did.

Oops! Right now I have to run for a birthday dinner, so I’ll check back again here once I can.

FWIW, the SOtM sMS-200 also employs Diretta.

The device that receives network packets and outputs USB, SPDIF, AES etc. is a Diretta Target using their proprietary technology. Somewhere in the upstream network path you will also need a Host to send those packets, which can be a second Raspberry Pi as described by David Snyder, free software added to your server like I have or a machine sold by Diretta or SotM. Not your MS600.

For those with a healthy bank balance, the SOtM sMS-2000 offers everything (including Diretta) - but the kitchen sink.

So, this networking scheme produces less noise, revealing more fine details in the recording. And, if this segment is segregated from all of the non-audio activity buzzing around the wires/wifi where you live, you might be able to finally hear most of what the microphones captured.

I know, I know… how this rankles the bits-are-bits objectivists. For those who just can’t accept my blasphemy, I recommend that you take a few minutes to read this recent review in Positive Feedback by Dean Waters:

https://positive-feedback.com/reviews/hardware-reviews/dejitter-it-switch-x/