Noted. But it might be the bandwidth of ~50Mbps required for DSD512 over the powerline adapters, not necessarily your DAC itself. Of course it could be your DAC input at DSD512 too.
A 50m Cat 5e patch cable is around USD30 (give or take, depending on which country you’re in). Obviously a shorter length cheaper.
That test should reveal a lot more info (hopefully).
I had issues with the gigabit TP-Link powerline adapters:
https://www.tp-link.com/au/products/details/cat-18_TL-PA9020P-KIT.html
You might ask why was DSD512’s ~50Mbps bandwidth an issue for a gigabit connection and it’s a fair question.
Real world connection speeds between these powerline adapters varied throughout the house, from ~200Mbps for a nearby adapter, down to ~30Mbps for a very far unit… This furthest unit even struggled with PCM192kHz and that’s not a bandwidth issue but just reliability/performance issues of the powerlines and maybe other stuff happening on the circuit.
Performance depends on the length of the path between adapters, the condition of the wiring and other stuff happening on the circuit too.
It might not be the issue in your case though - a really long ethernet cable at <USD30 might help provide more info.
Yeah, I was convinced it would be attributable to the Solwise Powerline Adaptors, but the use of a fully-wired 8-metre fibre link (using FMCs) from Router to switch then switch to STi7 made, to my great frustration, no improvement to the problem - hence this SOS being posted on the forum.
Incidentally, although the noise is worse and actually pulses when up-sampling to DSD512, it is still present (in diminishing extents) at lower up-sampling steps and also slightly present when sample rate conversion has been deactivated.
The only way to eradicate it completely is to totally disable DSP - which kinda defeats the object of having an SU-1 and Spring in the chain both at a config (firmware) ready to process DSD512.
I’m based in the UK, so I can’t trial your very welcome ideas immediately as I could be at work or asleep! 
The fiber links will solve any RF issues with the ethernet cable coming out of the powerline adapter - but not coming out of the AC outlet at your HiFi end… Is it one of those powerline adapters like I linked, that has an AC pass-through? i.e. your entire HiFi setup is plugged into it (via a powerboard)?
They have some RF filtering but maybe not enough in this case? Just another idea. I can’t think of an affordable way to solve that… apart from the WiFi mesh solution or getting ethernet hard wired but neither of those options is cheap. Power conditioning starts to get expensive and may not even fix it. I would still test the long ethernet (at under 30USD) idea before thinking about those more expensive solutions.
Oh of course, not suggesting you drop everything and try. I’m just pinging ideas as I think of them.
Hello again.
The fibre link was trialled with the Powerline Adaptors completely out of circuit. The Router was connected to FMC1 (local to the Router) via a short run of Supra STP cable, FMC1 was connected to FMC2 (local to the Hifi) via 8 metres of optical cable (‘dual run type’, correct name escapes me!) then FMC2 was connected to the switch via a short run of UTP cable. FMC2 was powered by an iFi iPower 9V wallwart and the switch was (and is) powered at 12V by an HD Plex LPSU.
The 50m of UTP…do you mean I should run this from the Router or should it be between the STi7 and MicroRendu? I tried the config with both items connected to the switch and this made no difference. I would suspect that the ‘fibre trial’ would negate the need to do the wired test from Router to ‘Hifi switch’, but I could see if fibre helps between the STi7 and MicroRendu?
The STi7 is located two shelves below the DAC and MicroRendu in my listening room. Maybe I should try it located by (and directly connected to) the Router to see if physical isolation makes any difference. Note: The noise seems to appear when performing functions or browsing in Roon as well as during quiet passages during playback.
Oh, that’s why I asked about the distance initially. My old ST i7 (which I gave to my old man) and router are on the complete opposite site of his house to his 2 x HiFi setups. Now I realise you don’t have that issue.
It sounds like you’re running out of cheap/er solutions to try and I’m running out of ideas 
Thanks for at least coming up with suggestions 
One more thing I could try that springs to mind is I could fish out the 19V SMPS that came with the STi7 and try it powered by that instead of the HD Plex LPSU. The LPSU is rated to 100W and I use it to power the 12V switch and the 19V STi7. I know that the later draws around 4.5A max, so there should be just about enough headroom to cater for the switch too (I need to check it’s mA demand), but maybe just maybe the LPSU is struggling for power whenever the STi7 is in a state of higher workload. Clutching at straws here, I feel, but we’ll see later.
You might try dropping the Audeze Preset and the Headroom Adjustment. That will have Roon only doing the one sample rate conversion to DSD and not the intermediate upsampling to 352 PCM. Like in my Sig Path:
Well, even with a Naim Uniti 2 as end point (via the Sonore UPnP Bridge feature in the STi7), the crackling is present, so that kind of narrows it down to the STi7, network or RFI due to the proximity of my various boxes.
I can’t track down the 19V SMPS that came with the STi7, so I’ve ordered one (they aren’t expensive) to eliminate the HDPlex from the investigation.
If anything, the SMPS made matters worse.
Downloaded Roon onto a late 2012 Mac Mini and tried it with the same tracks and settings as those most noisy on the STi7 and the problem disappeared. Mac loads tracks slower but has that elusive silent background I want.
So, Andrew Gillis from SGC has been trying to help diagnose the issue (including a remote session via Teamviewer), but the issue want go away when DSP is engaged. Frustrating.
Hey Dave, so was Roon acting as your ‘Core’ on the Mac Mini, in this test?
i.e. the sonicT i7 was completely out of the chain for this test?
Hey @Dave_Woollard,
Was there any difference in the way that the Mac Mini was connected to the network, or the way the devices were connected to the Mac Mini, compared to the Sonic Transporter?
Thanks,
Dylan
I tried to keep as much common to both servers as I could (although the Mac is a.c. powered and not using the LPSU feed). Yes, both run with Roon Core, both wired ethernet from the local Powerline Adaptor, both trialled with the same external SSD music storage.
I even connected the MicroRendu to the Netgear GS105 switch instead of the bridged output on the STi7, as that way, both the Mac and the STi7 would feed it via the switch.
The noise (a bit like an intermittent Geiger Muller tube sound) is noticeable when browsing in Roon Remote with no music playing and during quiet passages in playback - typically at the start or end of a track. It is exacerbated by upsampling and is almost like a repetitive background beat when upsampling to DSD512.
The following summarises the testing I’ve done (this was sent to Andrew Gillis before the Teamviewer session went ahead):
All of the following are with DSP enabled with upsampling to DSD512.
Mac Mini with 12V thunderbolt-connected external SSD music storage, with direct USB connection to SU-1:
NOISE FREE (other than a brief pop when track has finished and Signal Path light extinguishes).
Mac Mini with 12V thunderbolt-connected external SSD music storage, with MicroRendu endpoint into SU-1:
NOISE FREE (other than a brief pop when track has finished and Signal Path light extinguishes).
Mac Mini with USB3.0 connected external SSD music storage (with SBooster USB VBus blocker and external 5V from LPSU) with MicroRendu endpoint into SU-1:
NOISE RETURNS, albeit fainter than with ST.
As per 3, but with VBus blocker removed and 5V disconnected so that USB enclosure uses 5V from Mac:
NOISE FREE.
Sonictransporter with USB3.0 connected external SSD music storage (but without SBooster USB VBus blocker or external 5V from LPSU):
NOISE AS PER USUAL.
Sonictransporter with USB3.0 connected external SSD music storage (but without SBooster USB VBus blocker or external 5V from LPSU), 19V SMPS in place of the HDPLEX 19V Supply:
NOISE AS PER USUAL, maybe a little worse than normal.
Sonictransporter with 12V USB2.0 connected external Newertech Minimax HDD music storage (without SBooster VBus blocker), 19V SMPS in place of the HDPLEX 19V Supply:
NOISE CONTINUOUS AND WORSE THAN USUAL, continues beyond Library scan completion and with no playback. Sounds like a Geiger Counter detecting radiation!
I tried a different USB output on the back of the ST but this made no difference.
Note: The Mac and the ST (SMPS or HDPLEX) were powered from the same 8-way mains block.
Just reverting to the Mac with the USB3.0 SSD powered by the Mac’s 5V bus again outturns a perfect result, no static or crackling in quiet passages or between tracks. It’s just a little slow to load tracks that’s all.
Andrew was wondering if filters are handled differently by Roon in the Mac OSX vs a Linux OS. Any thoughts?
Hey @Dave_Woollard,
I’d like to propose a test that should give us a little more information about what you’re experiencing.
Try disabling all DSP settings that you’re using. Then, one by one, enable each DSP settings. What settings cause you to hear this noise? With this test we should be able to narrow down exactly what settings are causing the behavior you’re experiencing.
Thanks!
Dylan
Hi Dylan,
I will get a chance to try ideas tomorrow afternoon, but not before unfortunately.
The problem is there even with DSP off, but it is subtle.
With Upsampling to DSD512, it is worst (constantly there).
With Upsampling to PCM192, say, it is somewhere in between the two.
Andrew Gillis has suggested it’s a ground loop issue (this could hold some weight as the noise-free Mac is of course supplied via a 2-wire AC mains feed).
Is there likely to be a notable change in current demanded by a Roon Server as DSP gets more and more process intensive?
In my simple head, higher current equals a higher potential difference between grounds within the system.
Thanks for the info, @Dave_Woollard.
No rush at all on the tests, let me know what you find after you’ve done that.
Thanks for this clarification. Since it’s constant even without DSP, I’d like to recommend some additional tests as well.
Let’s try to see if we can completely remove the noise, even without DSP. First, let’s try making sure that the ST has the simplest setup possible (i.e. ST connected directly to the router via Ethernet). Once you’ve done this, let’s slowly start adding in variables until we notice the noise.
With the above method we should be able to hopefully pinpoint the root of the noise.
Thanks!
Dylan
The direct-to-router would have to be an FMC-based fibre connection as I do not have a long run of cable to hand.
I have compared fibre vs Powerline Adaptors with both the STi7 and the Mac and I couldn’t tell any difference (no noise attenuation noticed in the STi7 - much to my surprise). But if you think it’s a good idea, I’ll try it for the experiment tomorrow pm. Thanks.
I don’t subscribe to TIDAL, unfortunately. I’m quite ‘old school’ in that I own a sizeable amount of music. Given the intermittent opportunity I get to properly listen to music, I think this is fine.
Well, that’s weird.
Moved the STi7 across the room by the TV, away from the Hifi and used the local powerline adaptor I use for the Blu-ray player and found that the noise issue was cured! So, I decided to reinstall it back in the Hifi - fully expecting the noise problem to return. Nope, the STi7 is behaving itself now and any ground loop I had seems to no longer be present. I then ensured that the arrangement was exactly as it had been for months, i.e. PS Audio Humbuster feeding the Naim Amp and Hicap DR power supply, with everything else off the same 8-way power block, and still, all was well. The only slight ‘pop’ I get is when I first hit play or after a track when the signal path light extinguishes.
Just hoping it stays like this, but wish I knew what I’d done to fix this issue.
When Andrew Gillis did the remote session on Teamviewer, he deleted my out-of-date version of HQ Player. I’ve reinstalled the latest version but, even with the .xml ‘Key file’ successfully uploaded, I cannot get any playback with HQ Player selected. Please take a look at the following and let me know what’s wrong in settings - cheers!..
