› Forums › Music › Sound Engineering › Ground loop illustration / Cheeseweasel deserves a medal (as does this Italian dude)
- This topic has 28 replies, 7 voices, and was last updated May 30, 2012 at 9:27 pm by Pat McDonald.
-
AuthorPosts
-
October 26, 2011 at 8:15 am #1051748
As I mentioned in other posts I was facing a issue with noise and ground loops when connecting two laptop sound cards together (even USB sound cards) from different laptops, caused by a variety of issues.
On one laptop for instance had such bad interference on the PSU I have quarantined it from any of the other audio equipment (no amount of ground loop removers, ferrite beads or the like would stop the interference, and I checked the PSU without load and it was clearly coming from there and much of it in the audio frequencies)…
The better one has much less interference (its not totally non existent as you will see below) but using the ground loop adaptor I can get it to a point I am happy with…
I also found this excellent software which this brainy Italian dude has written and given away as freeware – it is hardcore stuff – you’d pay about £20000 for separate test equipment what could do this and it can also do other electronic circuit design like building passive filters etc (TBH I dont completely understand some of the functions myself!)
Managed to set up a basic test rig using a Soundblaster USB Card as line input (not the best, but I’ve had it hanging around since 2005) and an Alesis IO2 as the output. (in case anyone is curious about why I don’t use the IO2 as the input monitor, its because it seems to use exactly the same pre-amp for mic and line without much attentuation so it gets overloaded way too easily).
Anyway here is an illustration of the effect of the ground loop isolator..
This is the recording output of the amp without any signal going through the mixer feeding it
The first bit contains the harmonics and subharmonics of 50Hz which is of course our electric mains frequency in Europe, and some of this getting into your audio equipment is unavoidable – but if kept to this level can hardly be heard.
But there are plenty of nasty spikes at harmonics of 1kHZ! These are right in the sensitive part of human hearing – and if you turn the amplifier up these are heard as a ticking noise. If you (or someone else in your house) are careless and send a distorted signal to your amp as well as all this interference, you may well blow your speakers!
The noise comes from the circuits in the computer, and get sents in from the ground of the USB cable (which is also the audio ground) as thats the ground wire of the metal case of the computer and goes to earth connection in the mains plug – and its international safety law that this must be done. (if it didn’t you can get electrocuted, as a piece of faulty equipment might put 230V on everything metal)
Now here is the signal again with the ground loop isolator in..
notice that all the bad noise is mostly gone, and the remaining amount buried below the level of the mains hum raaaraaa
Now there are lots of expensive devices you can also use to achieve this, but the one cheeseweasel found was barely more than a fiver…
October 26, 2011 at 8:35 am #1245926I had some issues with wi-fi noises lately( airport extreme base station too close to one of the monitors)…have you checked for that?
October 26, 2011 at 8:48 am #1245915router is at least 2 metres away from any audio equipment or other IT kit – TBH if I am using a laptop as a desktop replacement I tend to shut off the Wi-fi and connect it directly to the LAN as its faster..tried turning off the Wifi out of curiosity but it makes no difference to the noise level (which TBH I am now happy with having added the ground loop isolator).
The real nasty was the DC power supply on that Toshiba :yakk:
I do need to shield the CRT monitor though (at least until I can get a new LCD one…) – that will be old skool cardboard and tinfoil job..
October 26, 2011 at 3:53 pm #1245933I’d like to see a graph of what that ground loop isolator does to music, that’d be interesting.
October 26, 2011 at 10:57 pm #1245930Nice one, glad it sorted the problem!
@DaftFader 455246 wrote:
I’d like to see a graph of what that ground loop isolator does to music, that’d be interesting.
I guess the ground loop isolator is just an inductor placed in-line with the ground (black) wire? Hopefully it shouldn’t have an effect on the wanted signal.
October 26, 2011 at 11:20 pm #1245934@cheeseweasel 455295 wrote:
Nice one, glad it sorted the problem!
I guess the ground loop isolator is just an inductor placed in-line with the ground (black) wire? Hopefully it shouldn’t have an effect on the wanted signal.
Ah ok, I read that it has some kind of filter (LP,HP and/or BP/BR within it iirc) … and assumed it was placed in the signal path. I guess as this could apply to a general electronic signal, that they were purely refering to blocking electronic frequencies as opposed to specific sound waves (I know they are the same thing esentualy … hence the confusion).
October 27, 2011 at 11:06 am #1245916OK here’s whats inside it.. its is just two proper audio frequency transformers on a decent bit of PCB and shielded (was pleasantly surprised, expecting far worse. Ok the build quality is your average Chinese standard (though neater than what I would have come up with) – and it worked perfectly after I put it all together again.
Apologies for the colour balance being all over the place on these photos, was rushing a bit trying to get something useable online..
and this is how it all fits together in the box (note the metal screening box to keep out radiated interference).
I have no idea what the transformers are / their specifications but I’m old enough to remember when all hi-fi equipment had many more audio frequency transformers (this practice only fell out of favour by the late 1970s). They do not appear to be filtering excessively as I am currently feeding a FM audio signal into a soundcard via one of these and the 19KHZ stereo pilot tone is getting through…
Daftfader is correct inasmuch that there will be an insertion loss and perhaps change in frequency response, but It all sounds OK to me, and I don’t have the brains to do the proper test (even though that italian software should do it – it hurt my brain enough just getting it to work!)
What I expect it does do is present high impedance to frequencies well over 30 kHZ (including all the nasties from the computer soundcard) which isn’t a bad thing as thats the noise you want to get rid of!
Frankly its a bargain for what you get. In the old days folk would have built one inside a tobacco tin, but its hard to get both baccy in tins and decent AF transformers and my old mate what told me about this was in addenbrookes for some time with throat cancer :yakk: (he recovered though)
October 27, 2011 at 2:01 pm #1245935OK so it does go in the signal path … GL can you show me a graph of it in the loop and then out with the same signal going though please? (I’d buy one my self and do it but I can’t afford one with beer expences in toe – that’s my second toe btw)
I’m gonna be a pain but ask for a 20-20K flat responce signal (i’m not even sure it’s possable without alot of fucking about …. if it isn’t then don’t worry .. but if you know a way of easily doing it then please do!!!)
I’ve got a fucking terrible problem with noise in my set up .. I knew it was from my PC … some of which is static interfearance from my grafics card seaping into my sound card what’s right next to it on my mother board (and my gfx card it way to fucking big and probably has it’s own gravataional pull, what’s not good, and this is without any electric current going thru it!)
I’m now, after reading this thread, 90+% it’s my power supply for my PC that’s fucking ruining everything that comes out my speakers … I only have to be playing a game and i can hear what my grafics card/HDD is doing out my speakers … I can actualy speak fluent PCIE and SATA now!
However, what I don’t want, is this device taking out frequencies that will effect my music production. I’ve only just got to grips with, for a start, my buggered up ears from DJing too much at loud volumes, and the overly coloured bass responce from my monitors … I don’t think my brain could handle any more adjustments that I need to make to the sound!
October 27, 2011 at 2:37 pm #1245931@DaftFader 455297 wrote:
Ah ok, I read that it has some kind of filter (LP,HP and/or BP/BR within it iirc) … and assumed it was placed in the signal path. I guess as this could apply to a general electronic signal, that they were purely refering to blocking electronic frequencies as opposed to specific sound waves (I know they are the same thing esentualy … hence the confusion).
Tbh I’m surprised that cheap box even has transformers in it. They will be cheap-ass ones though and chances are they will affect the frequency content of the audio passing through them a bit. I know you can pay a lot of money for good quality audio transformers (Jensen are a well-known manufacturer found in expensive DI boxes), and these really have a negligible effect on sound quality (transformers crop up in all sorts of audio equipment and were the only way to create a ‘balanced’ line before integrated op-amps became available – on an old analogue desk you’ll find a heavy balancing transformer on every input and output).
October 27, 2011 at 2:50 pm #1245936Probably totaly irrelevent to what we are talking about, but it strikes me that not only are musicians hard done by when it comes to the digital domaine but so are techies as well!
Not good man! 😥
it’s all about learning java script/html or better yet hex (or the matrix as I like to call it :laugh_at:)
October 27, 2011 at 2:53 pm #1245937P.S. I know none of the above apart from minor java scripts and a tiny bit of C# what I’ve probably forgot now anyway.
October 27, 2011 at 7:41 pm #1245917OK this is the spectrum without the filter
and with the filter added
apologies for not having the scale zoomed in but TBH I’m not 100% sure how to do that and get consistent results for both tests!
So yes there is some insertion loss and a slight change in frequency response (it actually slightly boosts the treble frequencies) but TBH I’d rather put up with that instead of the interference from the PC noise!
October 27, 2011 at 7:44 pm #1245918@DaftFader 455335 wrote:
I’m now, after reading this thread, 90+% it’s my power supply for my PC that’s fucking ruining everything that comes out my speakers … I only have to be playing a game and i can hear what my grafics card/HDD is doing out my speakers … I can actualy speak fluent PCIE and SATA now!
I’d also try moving the sound card as far away from the graphics card as possible (assuming that its internal and also it won’t cause you endless problems with IRQs etc…)
October 27, 2011 at 9:40 pm #1245938@General Lighting 455384 wrote:
I’d also try moving the sound card as far away from the graphics card as possible (assuming that its internal and also it won’t cause you endless problems with IRQs etc…)
Thanks for the pics! I have a high freq+/- on my monitors so could probably atenuate the slight addition to the highs and my speakers slightly boost the frequencies that it cuts in the low range anyway, so it looks like I might have to get one of these!
I know it’s not going to fix the gfx card hum totaly as it isn’t your “standard” noise frequncies (more a wining noise then crackle or hum) but I allso get some other interfearance that I think this will help get rid of.
Unfortunatly my GFX card/sound card combo will cause this noise where ever it’s plugged in (it is an internal one) and moveing them doesn’t fix my irq problem either which it a total pain in the arse (they are currently in the furthest slots apart possable, but my gfx card takes up the space of two pcie slots and there’s only one pci slot inbetween them).
hopefully this will go some way to soughting the issure till I can afford a proper external soundcard (I’m thinking focusright saffire 24 or somehting similer by the same people)
October 27, 2011 at 9:44 pm #1245939On another note .. the high frq. boost might be to try an compensate for the slightty lower mid/highs, so it doesn’t “dull” the sound. Could just be acsidental tho.
October 27, 2011 at 10:09 pm #1245932Interesting graphs. I think the main thing it tells us is that GL’s system has a lousy response above 13k!
The +3dB boost at high freq is quite significant and should be audible, though I’d take that over an annoying PSU whine any day.
I wouldn’t be concerned about the HF boost at all though – the combination of speakers and listening environment will have a far more significant effect on the frequency response than that transformer. If you put a measurement mic in the listening position and ran the same analysis on its output (play noise through the speakers and record it), you’d end up with a much messier graph!
Graphs only ever tell part of the story though – there are far more variables that affect sound quality than frequency response at a particular input level, and the best way to assess them is generally to listen.
October 28, 2011 at 12:28 am #1245919going by my ears I would say the frequency response is good up to about 18-19 kHZ (I can still hear up to about 18.5-19 kHz) – TBH that graph isn’t scaled very well anyway for looking at HF response – out of curiosity I managed to run one with a easier to read scale..
TBH the two soundcards used for the test haven’t been calibrated/balanced either so more errors are creeping in there. They definitely can pick up higher frequencies, as if I try and use the record output from the amp/receiver, the damn 19kHZ pilot tone from the FM tuner stage gets into it! (the obvious solution is not to record from the mixer output, unless you are of course recording from the radio (in which case there are ways of filtering it as FM audio only goes up to 15kHZ)
That said I’ve now got everything to a quality I’m happy with – even the basic stuff is way better than all the audio kit I’ve had in the past,…
October 29, 2011 at 11:39 am #1245929I have a very silly suggestion for the problem and tbh I have never tried this inside a pc but if you are game you could make up a card sheet with aluminium foil sandwiched inside it and see if that helps at all [mind dont leave the ali exposed at the edges – nothing worse than random loose metal in a pc]
My thinking on this runs along the lines – most interference is picked up as a radio signal from badly shielded electriconics and if you put a metal foil layer between the PSU and the problem PC card it may do the job of blocking the noise transfer; I know this technique will work on a badly shielded external soundcard and as long as the disruption of the airflow in the PC case is not drastic or can be compensated for its worth trying this idea as its a real cheap fix….
I dont have this trouble in my PC though so I cant try it out and tell you what happens or I would be doing that instead of typing it in :crazy_fre
October 29, 2011 at 12:14 pm #1245920many old style analogue modems have such a shiekl on them as well as isolation transformer (the Communications Ministry in most countries will refuse to approve them for connection if they are sending interference back down the telephone line, as it causes problems to other users in the multipair cable especially with ISDN and broadband being common nowadays),
What Daftfader may have is a dirty SMPSU though… this is what the Tosh PSU is putting out without any load on it (and with the DC filtered via the capacitor in the scope). This is only a cheap scope unfortunately, was a mission to get it to capture the spike properly (had to keep pressing “hold”) – I can’t afford a top range multi colour Agilent one and my sources at British Telecom have took early retirement so I couldn’t even try and blag the PSU into Adastral where they have loads of nice toys like that..
thats a square wave pulse of 40 mV if not 50 so definitely going to find its way into any audio..
November 21, 2011 at 1:06 pm #1245927@DaftFader 455297 wrote:
Ah ok, I read that it has some kind of filter (LP,HP and/or BP/BR within it iirc) … and assumed it was placed in the signal path. I guess as this could apply to a general electronic signal, that they were purely refering to blocking electronic frequencies as opposed to specific sound waves (I know they are the same thing esentualy … hence the confusion).
An inductor is a filter – inductor placed inline with a speaker = basic low pass filter (or 1st order crossover).
Inductor of the right value placed on a possibly dodgy ground line stops it from allowing any hum etc through…
November 21, 2011 at 1:16 pm #1245928@General Lighting 455326 wrote:
OK here’s whats inside it.. its is just two proper audio frequency transformers on a decent bit of PCB and shielded (was pleasantly surprised, expecting far worse. Ok the build quality is your average Chinese standard (though neater than what I would have come up with) – and it worked perfectly after I put it all together again.
Apologies for the colour balance being all over the place on these photos, was rushing a bit trying to get something useable online..
and this is how it all fits together in the box (note the metal screening box to keep out radiated interference).
I have no idea what the transformers are / their specifications but I’m old enough to remember when all hi-fi equipment had many more audio frequency transformers (this practice only fell out of favour by the late 1970s). They do not appear to be filtering excessively as I am currently feeding a FM audio signal into a soundcard via one of these and the 19KHZ stereo pilot tone is getting through…
Daftfader is correct inasmuch that there will be an insertion loss and perhaps change in frequency response, but It all sounds OK to me, and I don’t have the brains to do the proper test (even though that italian software should do it – it hurt my brain enough just getting it to work!)
What I expect it does do is present high impedance to frequencies well over 30 kHZ (including all the nasties from the computer soundcard) which isn’t a bad thing as thats the noise you want to get rid of!
Frankly its a bargain for what you get. In the old days folk would have built one inside a tobacco tin, but its hard to get both baccy in tins and decent AF transformers and my old mate what told me about this was in addenbrookes for some time with throat cancer :yakk: (he recovered though)
Looks awfully like the boxes I’m putting in the car when I eventually install the bass bins and amp – I’m using 2 to go from the head unit to the power amp in the boot (setting them up back to back at either end of the cable essentially gives us a balanced line between the 2 devices – doesn’t need an “earth” point like a normal sound system as the car doesn’t have a standard earth like a house or genny etc, it uses a negative grounding point system…)
No sure how well they’re gonna work yet (I’ll post here when we eventually get round to it), but I’m told they work really well like that (transformer balancing works as well as if not better than servo etc IMO – most of the reason it doesn’t get used much is the transformers are usually pretty expensive especially when compared to an SMD servo chip)…
December 29, 2011 at 2:11 am #1245921@noname 458334 wrote:
Looks awfully like the boxes I’m putting in the car when I eventually install the bass bins and amp – I’m using 2 to go from the head unit to the power amp in the boot (setting them up back to back at either end of the cable essentially gives us a balanced line between the 2 devices – doesn’t need an “earth” point like a normal sound system as the car doesn’t have a standard earth like a house or genny etc, it uses a negative grounding point system…) [/quote]
they appear to be sold for that purpose TBH. I even managed to get the bad PC usable with some audio via a USB soundcard but even on that and everything other than the amp and a newer desktop PC using balanced audio lines now, even the soundcard input needs to be sent via one of these. Otherwise all the crap gets straight back in.
Another use I’ve found for these devices (which are amazing for the price) is when connecting a HF (shortwave) radio receiver to a computer sound card (to decode such things as messages from the Coastguard, which round here also provide useful forewarnings of bad weather). These stop all the interference the computer sends down the mains earth getting into the radio and swamping the signal. however I’ve managed to find a old modem card with a 600 R isolation transformer on it which I will adapt for this purpose, as the radio has to be put at the other end of my house as UK Power Networks substation is about 20m away from my back garden which will makes any HF antenna useless as it picks up not just 50Hz harmonics but the whole estates electromagnetic interference, and it will be look better here to run the audio as a twisted pair using telecoms cable (of which there are usually a fair few offcuts at work) than a great big audio cable – at least until I can afford a decent Japanese RX – whilst VHF/UHF scanners are way cheaper today most decent HF kit costs as a car and the older models are hoarded by dads/grandads generation who wisely do not rely on mobile phones/internet for everything.
whats more worrying is they were the ones what set up todays mobile phone and broadband networks and know that in Blighty its all held together with string, gaffer tape, and the remains of cannibalised equipment almost as old as me – probably what we sold to foreigners in my youth and now rejected by the Chinese for as scrap metal.
Whatever new kit we get is what the Chinese have on special offer (whereas the Germans will at least buy the newer stuff) or worse, built in China at bargain basement prices and systems designed by the French with the Americans sticking their oar in (usually moaning about costs and “time to market”).
In a previous job I’ve had to personally counsel numerous French engineers who have actually burst into tears when faced with the problems their nations telecommunications network and electricity supplies present – including all manner of unpleasantness such as “broadcast quality” comms circuits which intermittently mangle any signal fed down them (with France Télécom blaming everything else connected to them), and worse, 400V appearing where only 230V is expected. in many cases it wasn’t at all surprising that surprising they ended up crying like babies…
March 11, 2012 at 11:41 pm #1245922@ShaftInvader 455389 wrote:
hopefully this will go some way to soughting the issure till I can afford a proper external soundcard (I’m thinking focusright saffire 24 or somehting similer by the same people)
just a heads up – even some external sound cards with supposedly balanced audio will also be plagued by this (and yes it is the PC power supply), the noise comes in from the USB power…I tried some tests and found that as my cable runs are short I am better off using unbalanced audio and those transformer boxes (only about £5 each!) on any audio circuit than the balanced feeds…. (there are more expensive transformers around, and maybe its worth investing in those for a project studio). TBH I suspect some of the more expensive PC sound cards contain them (rather than other ways of creating a balanced circuit which can let in interference)
also the radio enthusiasts’ sites confirm its a big problem with PC’s, in fact its worse as the noise can get also get into the RF section of the radio and swamp the signal they are trying to decode..
March 12, 2012 at 2:59 am #1245940@General Lighting 472132 wrote:
just a heads up – even some external sound cards with supposedly balanced audio will also be plagued by this (and yes it is the PC power supply), the noise comes in from the USB power…I tried some tests and found that as my cable runs are short I am better off using unbalanced audio and those transformer boxes (only about £5 each!) on any audio circuit than the balanced feeds…. (there are more expensive transformers around, and maybe its worth investing in those for a project studio). TBH I suspect some of the more expensive PC sound cards contain them (rather than other ways of creating a balanced circuit which can let in interference)
also the radio enthusiasts’ sites confirm its a big problem with PC’s, in fact its worse as the noise can get also get into the RF section of the radio and swamp the signal they are trying to decode..
I’ve soughed it out pretty much now, my speakers go +/-30db and I did have them set to 0db .. but i’ve put them at -30 and turned it up on my soundcard software and the signal to noise ratio is good enough that I can’t hear the noise unless i have it up really loud and have my head next to the speakers. so all is well.
I still need an external soundcard to fix the irq problem though and i think the one I want to get has a separate power supply and doesn’t draw power from the usb, altho my luck it’ll probably still leak through anyway if i turned my speakers up so i could hear the noise.
ta for the heads up though 🙂
March 12, 2012 at 12:28 pm #1245941@Mitch Conner 472156 wrote:
He he, tekkies..LOL
TBH i’m just jellous that I don’t know how to have fun tinkering with gizmos and gadgets. Maybe one day though.
ask you’re local friendly tech support guy. :laugh_at:
-
AuthorPosts
- You must be logged in to reply to this topic.
› Forums › Music › Sound Engineering › Ground loop illustration / Cheeseweasel deserves a medal (as does this Italian dude)