Floated Grnd on Digital gear
Discussion of AC mains isolation transformers
This is my current understanding, eager for correction, where appropriate... :-)
Whether starting out balanced (with the secondary grounded from a center tap) or unbalanced (with the secondary grounded at the neutral wire), if you then float the secondary, you will still get the same amount of common-mode noise reduction as you had before floating the secondary, with the only difference being that if you started out balanced, with a grounded center tap, you will have just converted it to an unbalanced transformer with a floating Secondary.
So... In your experience, you have essentially said you prefer the sound of an ungrounded (floating) secondary to the sound of a balanced secondary (which, by definition, is grounded via the center-tap.)
I suspect the reason you prefer the floating secondary is because you've got a ground loop between one or more components that gets attenuated somewhat when you float the secondary that they are all sharing.
I'm not a big believer in balanced power, perhaps because I so strongly believe that the "magic" of common-mode noise reduction occurs via the Faraday shield that inhibits capacitive coupling from the Primary into the Secondary. In other words, what work is left to be done, by the time the 60 Hz AC power has passed unharmed from the Primary through to the Secondary (via magnetic induction), stripped of any DC signals (Common-Mode noise) that have, at that point, already been attenuated 20,000,000:1 (-146 dB), in the case of the 0.0005 pF Topaz units? Whether we ground the secondary at a center tap (balanced) or ground it at the Neutral (unbalanced), the Common-Mode noise that came into the Primary on the Mains has already been diverted to ground by the Faraday shield(s), converting any residual common-mode into transverse-mode (normal- or differential-mode) noise - which most 1:1 transformers would just pass into your loads, but the Topaz Ultra-Isolators have some kind of double or triple Faraday shield configuration that manages to provide -65 dB of normal-mode noise reduction without using modern filter circuits (i.e. Y-Caps at the input, an X-Cap across the outputs (for modest voltage surges) or one or more serially arranged MOVs across the inputs or outputs (for heavy surge protection). Keep in mind that MOVs actually create some common-mode noise when suppressing normal-mode transients, and the Topaz can already absorb a lot of surge, at least until the voltages are high enough to saturate the transformer and penetrate the insulation of the windings. :-)
OK, I got into a bit of a ramble, there, but to answer your first question... There are two reasons for putting only one component (one load) on a transformer with a floating secondary: 1) If you were to plug two components into one floating-secondary transformer, and the first component somehow shorted Hot to the case, you would be fine, until, by chance, at some point in the future, the other component shorted Neutral to its case -AND- you happened to touch both cases at the same time - allowing hot to flow into one hand over to the other hand. The probability of this happening is extremely low, but if the two components were plugged into the same transformer with a grounded secondary, you wouldn't have to worry about the improbable possibility. 2) Much more probable is the chance that you've got a DAC, CD Player, Blue-Ray player or other microprocessor-equipped component that's dumping digital "hash" common-mode noise back onto the power cord, where it could pollute the power going to any other component that's plugged into the same outlet, power strip, or transformer. You could also have one or more components that are equipped with SMPS, which are also known for dumping noise back onto the mains.
A well-made analog component, with a linear power supply, like an amplifier, is more likely to be a victim than a culprit, when sharing a power strip with other components, so these non-backwashing components can be plugged directly into a grounded-secondary transformer (whether balanced or unbalanced - your choice), but all the backwash suspects (DACs and other digital sources, plus anything with an SMPS) should be segregated from each other and from the analog components, by using floating-secondary transformers that themselves plugged into the one, larger capacity, grounded-secondary transformer. Each of the floating-secondary transformers will prevent any backwash noise from polluting other components - the common-mode noise attempting to flow upstream will be attenuated as it attempts to get from the secondary coil to the primary coil. (A transformer with Faraday shield can attenuate common-mode in both directions.)
It will also break ground loops - even if you have signal cables running between components - i.e. between a DAC and an amp - no loop can be formed to get back to a shared power ground when they are each plugged into independent, dedicated, floating-secondary transformers. Remember, each component with a three-prong plug still has a safety ground that goes all the way back to the wall outlet, through both transformers, per code, but neither the Hot nor the Neutral supplying power to each compoent is connected to that safety ground - at least not within their respective floating-secondary transformers. Joy!
You can, however suffer some leakage currents, caused by voltage differences coming out of their independent transformers - i.e. when an amp is plugged into the powerstrip of the big grounded-secondary transformer, but the DAC is plugged into a floating-secondary transformer that has been inserted between the DAC and the big transformer - the amp and the DAC will see different voltages, in which case, if you want to discourage leakage currents, you can really pull out all the stops and insert a high-bandwidth signal transformer, such as one of these, made by Jensen:
But to start with, I would encourage you to just ground the secondary of your main transformer (unbalanced is my preference), then insert Hammond 171 Series transformers (which range from 100 VA to 1000 VA and have floating secondaries) or a B&K Precision 1604A, ahead of each digital component or those with SMPS. Give that a try and see if it doesn't sound better than plugging everything into your single, floating-secondary transformer.
I would, however, caution you to use a Kil-A-Watt to measure the voltage going into each component when operating. Each transformer in a series will likely increase the voltage that came into it, even though they are technically 1:1, until you fully load them to their rated capacity - which you really don't want to do, especially for power amps, because you want to leave "headroom" for the transformer to supply instantaneous peak currents on demand, above rms current requirements.
If stacking two 1:1 transformers serially has your component seeing more than about 127V, you should consider getting an APC LE1200 voltage regulator, and setting it to an output voltage of 107V (the lowest of three available settings.) Plug your big grounded-secondary transformer into the voltage regulator, then the smaller floating-secondary transformer, and you'll measure something closer to 120V at the second transformer's output. :-)
Normal mode, common mode noise
AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.As I understand it from Mike a IT with floating secondary will attenuate normal-mode noise and a grounded secondary will attenuate common-mode noise. The perfect combo is actually to have a IT with floating secondary (downhill) followed by a IT with grounded secondary (uphill) or a perfect sinus wave UPS (like AG500 or AG1500). However, I have a question. What causes common-mode noise? Is it DC connected devices as I understand it...or is it other forces? If anyone knows about this please let me know!
Sent from my Nexus 7 using Computer Audiophile mobile app
First off lets define what common mode and normal mode are.
Normal mode is noise between the hot and neutral. It is assumed that you have a 50/60HZ at 120 or 240 as the "signal", so anything else is "noise" by definition. The low capacitance transformers achieve high attenuation of this noise by having very low capacitance which blocks high frequency signals, and a special magnetic circuit which is only efficient for the 50/60Hz and a couple harmonics above. Higher frequency noise does not pass through the magnetics, and they are not transferred through capacitance because it is so low.
Common mode noise is noise between some other reference and the both the hot and neutral together. If this noise causes a 1V rise on the hot it will also cause a 1v rise on the neutral. Usually in three wire AC systems the "reference" for common mode noise is taken as the safety ground.
A system without a secondary connection to safety ground (ie floating secondary) has essentially no coupling to safety ground so there can be pretty much any voltage between them. Any common mode noise that was on the input is blocked, but what happens on the output side is the same as just holding a piece of wire in one hand and another piece in the other hand. They are free to float all over the place due to picking up radio signals etc.
So if you put a scope between safety ground and either secondary wire you will see random noise, mostly radio pickup of some sort. IF your audio equipment is designed so its input is referenced to safety ground then this may be a problem. The tie between safety ground and one of the secondary wires (secondary neutral grounding), forces them to be connected so they can't float relative to each other.
If the AC for your whole audio system is connected together and run off the secondary of the transformer, then it shouldn't make any difference which way it is hooked up. BUT if you have an input which comes from a source that is not AC powered from the transformer you may find a difference in sound between the two configurations.
From the safety side of things, IF you have the grounded secondary configuration then you should have a GFCI on the output of the transformer because now a sneak path can exist.
Balanced Power floated secondary
AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.
Here is good news from Peaktech support!
Dear Mr. Corn,
regarding your request I would like to inform you that the isolation transformer has a “floating secondary” according to your sent picture.
We hope that would be helpful for you.
- Technical Support -
PeakTech Prüf- und Messtechnik GmbH
So this means that the Peaktech 2240 have a floating secondary....not a grounded secondary.
I have also asked them for the inter-winding capacitance specs. I will post it here if they have this value.
I hope Mike @zilch0md does´nt mind me posting his drawing here that clearly illustrates the differences between the two!
Panel wiring ground fault
AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.
By the way, I believe the owner here wired the two dedicated lines on vertically adjacent slots and I'm not sure this is the recommended best way.
If you are speaking of two vertically adjacent circuit slots in a AC mains load panel, then you are correct, it is not a good idea for people in 100/115/120V countries (where the feed to their house is 120/0/120V) to hook their amp and front end gear to opposite lines (though there is something to be said for putting the computer on the opposite side of the line).
I get the impression that some here think I am joking about line sides coming from the street transformer sounding different. I am not. I live out in the country and have a transformer on the power pole that is dedicated to our house (there is not even another transformer on that pole and the nearest is a quarter mile away).
What I do to determine best line side (to run my gear from the dedicated 50A sub panel in the wall of my studio) is to turn every other house breaker on the main panel off so that other house loads (refrigerator, lighting, etc.) are not an influence. While in past homes I would then move the breaker for the stereo up and down one position (which puts it on the alternate side of the line), my current room has both sides of the line run to each quad of outlets--so one duplex outlet is L1 and the other is L2 (and someday I will rewire them with a 240V plug and resolder the transformer jumpers in my amps to run from that for even better results).
Anyway, it then takes but a minute to listen to all gear on L1 and then move it to L2. The more "upfront", "exciting", and "in-your-face" sounding side is NOT the side I choose. The other side seems more dull and slow at first, but it is actually the more musically correct and less fatiguing in the long run.
But yeah, I was kidding about the copper lines from the pole, though I have wondered…
That said, the problems with termination of aluminum wire are well documented: differing coefficients of expansion, galvanic corrosion, loosening of connections, and the fact that aluminum oxide does not conduct.
Once every couple of years I retighten all the breakers in my panels, but I stopped doing so for the big terminals of the down-feed from the weatherhead after I slipped one year--arcing and throwing some metal slag!
When we moved into the house 23 years ago the download from the weathered to the meter/load panel was aluminum. I replaced that with copper (and the entire panel) a couple years later when we developed a neutral fault that caused damage to a lot of gear in our house. The original tip-off what the huge bounce the lights would take when my laser printer (think ancient 50 pound early HP energy hog beast) would print a page.
Perhaps surprisingly, while I put some surge protectors on my work computer gear, all my hi-fi goes straight into the wall. Well, that's not quite true, I have two smallish custom 1:1 isolation transformers for my front end gear. But having tried a lot of "power conditioning" and protection units over the past 3 decades, I've found most do more sonic harm than good. Then again, we have pretty steady and clean power out here--many folks don't.
Twisting AC Mains wire pairs
AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.A twisted pair (Hot & Neutral) is the best way to go./QUOTE]
And always twist to the left please. Same for interconnect and speaker cable. Not kidding.
Go ahead and make an extension cord with a pair of solid #10 two ways: One twisted to the left, one twisted to the right. In a sensitive system they will sound different. (And no, it is not the reverse in the Southern Hemisphere, we checked that long ago.)
I'm really not into audio voodoo, and I don't pretend to understand why this is so. And please, nobody take my word it.
Hello HIFI, you don’t provide any details about what you’re exactly looking for, so this suggestion may or may not meet your needs.
About a year ago I decided to implement a new digital system based on local and remote streaming, with Qobuz as a remote source to an Innuos Zenith SE and more recently a Statement. Fortunately I have 2 branches of Curries PC World in close proximity so I was able to try lots of different combinations of routers, internet-over-power line, extenders, mesh systems and direct ethernet connectors, using their 30 day policy to buy, try and return.
My hi-fi is one floor above my router, with a staircase in between, so a 15m ethernet cable was temporarily installed bungee style (flung over the stairs) and would have been a lot longer if I routed the cable properly. The original router was a Virgin Superhub3 which combines cable modem and router and is a weak-kneed piece of crap with flawed firmware. My first trial was ethernet cable vs wi-fi. Fortunately the wi-fi connection sounded considerably better than the hardwired ethernet solution.
Analysing my network needs, they break down into the following categories:
- General fixed location wi-fi devices (PC, Naim MuSo QB, etc
- Mobile wi-fi devices (iPads, iPhones etc)
- Hi-res video
- Hi-res audio
Essentially I tried several different combinations listed above until I found the ideal.....
- No disconnects or instability on fixed location devices
- No disconnects on mobile devices when changing locations
- No buffering or drop-outs on video
- Ultra high speed connection for hi-res audio with no drop-outs
- Best sound quality
Virgin mandates the use of Superhub as a cable modem to connect to their network, so I switched it to modem mode and used a seperate router for all connections within my house.
After trialing lots of alternatives the best I found was the following:
TPLInk Archer AC5400 tri-band router with MU-MIMO, which includes:
communication with 4 devices simultaneously; ultrafast 1,4GHz dual core CPU with 3 co-processors, maximised range through 8 beam forming antennae and an easy to use interface, giving access to all the router’s advanced features.
For the hi-fi, I added a TPLink RE640 extender, which offers 2.4Ghz and 5GHz dual bands and the same high speed capabilities as the router.
Both router and extender are controlled via TPLink’s Tether app, which allows you to set up your network and define what connects to what, who has access etc. In order to optimise my entire network I set up the following:
2.4MHz Band to connect distant and mobile devices
#1 5GHz band for Video streaming and local devices
#2 GHz band dedicated to the hi-fi
In this manner, the slower but more robust 2.4 GHz band is used to connect low demand and distant devices, while the 5GHz band is used to connect video, where robust, high bandwidth avoids the dreaded buffering. For hi-fi, the 2.4GHz band on the extender is switched off, and the 5GHz band is dedicated to the router’s 2nd 5GHz band, with no other clients. At over 200 Mbps, the hi-fi server’s iPad based user inter is so fast is feels like a local, hard wired interface and the music streaming is totally robust and absolutely free of drop outs. Sound quality was exceptional but did improve markedly with the addition of Sean Jacob’s DC3 LPSs for all network devices.
Austinpop server build
A novel way to massively improve the SQ of computer audio streamingCustom Roon Server - Initial ReportAs I've described here in the past, while I have upgraded my Roon endpoint to near-endgame with the sCLK-EX-modified NUC7i7DNBE/Plato X7D running AL in RAM, my server upgrade has been pending. In the meantime, I've managed to squeeze (pun intended) out a lot of performance from my generic Dell XPS 8700 desktop. However, it was always my plan to upgrade my server, following shamelessly in the footsteps of @romaz.Roy has been refining his server build, and while he is still tweaking, I acquired his first build to try in my system. I am just going to enumerate the system here, because most of the credit goes to Roy for the thought and effort he put into this. Here are the system details:
While Roy and I discussed these options at length, it's really he who drove this system selection, and the end result is truly special. Before I describe the sound, here is a brief rationale for these choices.
- HDPlex H3 chassis
- Asrock H370M-ITX/AC mobo
- Intel 8700T CPU
- HyperX Fury 2400MHz DDR4 8GB
- JCAT NET Card Femto
HDPLEX 400W ATX Linear Power Supply
- 19V rail modded to 12V to supply independent EPS input to CPU
- Ghent JSSG360 ATX-to-ATX and XLR-to-EPS cables
This report is just my initial listening impressions after a couple of days. For now, I've located this server adjacent to my Dell (in another room), so I can compare them head to head. For simplicity, the endpoint is configured with Roon Bridge.From the first note, there's really no contest. I've talked about how I was living with some harshness ever since I resumed using my Dell as the Roon Server. Well, the lack of that harshness is what hits you first. The sound is so much more natural, relaxed, and yet, much more dynamic. I don't have the ability to compare the NUC i7 as a server with this 8700T server, but I strongly suspect this is superior.Happy as I was with this, I next ran a 50ft Cat 6a generic cable from the second JCAT Net port from this server to the SOtM switch, to which my endpoint is connected. Wow, this was another uptick in SQ. I think this credit goes to bridging (in general) but more to the JCAT Net card in particular. Roy had already raved about this card to me, but I had to hear it for myself. The biggest improvement is in tonal richness and density. There is just more texture and bloom (in a good way). Most importantly, and personally very exciting, for the first time in my system, I heard the same album on Qobuz sound as good as local files on NAS. Kudos to @Marcin_gps et al. This is an outstanding product.I have a lot of interesting things to try next. I plan to move this server into my listening room, where it can benefit from better cabling and better power. We'll see how much of an uptick that provides. This is where the fanless chassis comes in really handy. There are some PSU experiments to try, but Roy has already explored these, and I'm inclined to trust him on these. I'll describe these later. Finally, there is all this buzz about Euphony, so at some point I should give that a whirl.I will say, this may be as close to endgame in the digital streaming chain as I've ever felt I've gotten!
- The 8700T CPU has 6 cores/12 threads/12MB Smart Cache, all for a TDP of 35W at a base frequency of 2.4GHz. This seemed to be a good balance of low power and hefty horsepower.
- The HDPlex 400W Linear PSU has independent rails for the ATX voltages (3.3, 5, and 12V), with no DC-ATX conversion. Plus, by modifying the 19V rail to 12V, this adds another rail solely to power the CPU via the EPS input.
- TDP 35W is well within the capability of the H3 fanless chassis, and we snake the cables into the chassis through a PCIe slot opening on the back. Not elegant, but effective.
- The JCAT Net card was chosen to provide a high quality bridging solution, and as you'll read, boy does this thing sound good!
A novel way to massively improve the SQ of computer audio streaming1 hour ago, BigAlMc said:
If someone can confirm whether it's the BBB version or another AL version that's best for endpoint patrol?
audiolinux_mini_05.img.gz thats the headless one just burn it with Rufus 2.18 DD image.
Chose UEFI boot not legacy.
turn off secure boot.
chose linux os.
JSSG360 Step by Step
A novel way to massively improve the SQ of computer audio streaming
OK gang - SUCCESS building my very first JSSG 360 cable!
To those DIY-phobic readers like me who've been hesitant to try this - come on in, the water's fine.
I started with this generic Pasternack RG-400 50Ω cable
- I first masked the protruding ends of the BNC connector into the cable with electric tape (green in the picture).
- I then used a pencil to shape the flat Electriduct 3/8" flat tinned copper braid into a cylindrical form, and then slid the braid over the BNC connector onto the cable.
Following which, I applied electric tape over the bulk of the braid, leaving only the ends exposed, as shown below. Note: I used some tension while applying the tape to mimic the effect of the heat shrink tubing. It seemed to work fine. Also, this particular brand tape as shown is quite flexible and did not cause the cable to become resistant to bending as I had feared.
Next, slide on the 2nd layer of braid:
- After this, I "encouraged" the exposed ends of the inner and outer braids to fuse together - not hard, just peel away from the cable and then finger twist the wires of both braids together. I then used the silicone rubber fusing tape - awesome stuff, excellent hint by @mozes - to seal the fused ends of the braid. This tape is stretchy, so it helps to apply some tension when applying it, and it just fuses nicely.
- Final step, slide on the wire sleeve for a nice finished look, and seal with the fusing tape.
I'll be doing some listening this evening, and will report on how it sounds. Exciting stuff!
- I started with this generic Pasternack RG-400 50Ω cable
DC Cable Recipe - DIY JSSG360
A novel way to massively improve the SQ of computer audio streaming
DC Cable Recipe
A novel way to massively improve the SQ of computer audio streaming
Ha, not I. Shall be interesting to see if Rob's claims (again) that his switching power supply coming with the m-scaler will be superior to any LPSU. Could be possible for the m-scaler performance but lost in kickback of noise to other components on the same circuit. Will the Paul Hynes SR7 prevail over Rob's switching power supply?
I probably won't find out till next summer, adding this last piece of the puzzle to make my system complete. Not that it doesn't sound great as is. That will complete for hopefully a long period of time any additional components or upgrades. Just minor tweaks like cables, etc. DIY.
Thanks Larry for the heads up on the shielding design for cables. I just made a nice one about 8 inches long from my SR7 (XLR) to NUC (barrel connector) using Canare 4S6, 1st shield 1/4" Metal Braid, 12mm heat shrink, 2nd shield 3/8" Metal Braid, 12mm heat shrink, a little 18mm heat shrink over the XLR. Came out nice. I got rid of those screw barrel connectors, they we're awful over time. Connectors all soldered. I'll probably do all my cables, including SATA and clock cables with the 360, to protect from RF noise penetration.