Interference from cassette player
in [UK Mac]
|
Prev: Complex page numbering in documents?
Next: Special characters in file names and cross-platform compatibility
From: Woody on 19 May 2010 07:24 Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > Woody <usenet(a)alienrat.co.uk> wrote: > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > > Woody <usenet(a)alienrat.co.uk> wrote: > > > > > >if it is a mains loop, > > > > > > Earth loop, you mean - via the mains wiring is the usual way it happens > > > in domestic audio. > > > > Earth loop, yes. > > For those who don't know: > > If you get a circle of wire and change the magnetic field inside the > loop, you'll induce a current in that loop. > > The bigger the rate of flux lines cutting the loop, the bigger the > current change (basically, the more changing magnetic field, the more > the induction). > > So that means the bigger the loop, the bigger the current change. > > Big loops pick up lots. > > Think about a loop involving your mains wiring and the signal wires > connecting units in your music gear. > > Big loops. > > So loops via the mains wiring have been known to add /loud/ hum. > > Professional audio gear uses what they call `balanced' lines, aka > differential signalling, where you don't refer the signal to earth but > refer it to the other line (one line goes up while the other goes down, > and vice-versa). This avoids any chance of an earth loop (and other > problems). Almost apart from the last bit. Balanced lines reduce the signal due to general interference, mains hum, florescent light buz, switching noise etc, especially as they tend to be low impedence. Earth loops are where two connecting pieces of audio equipment are connected to the mains. As they all tend to be earthed, you can get a problem where the earth that two pieces of equipment connect to are different or far away. That can cause an earth loop and mains hum regarless of how you connect, althouh a balanced lines can reduce it. -- Woody
From: Woody on 19 May 2010 07:36 Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > T i m <news(a)spaced.me.uk> wrote: > > > usenet(a)alienrat.co.uk (Woody) wrote: > [snip] > > > >> I'm thinking that even if you 'suppressed' one input (as you would > > >> with a mono jack in a stereo socket) it (the Mac) might ramp the gain > > >> up and therefore be picking up more 'noise'. Given a signal it might > > >> AGC back to a more normal sensitivity <shrug> > > > > > >There isn't any AGC on the input of the mac, it is nominally line level, > > >although adjustable. > > > > Ok ta, wasn't sure and I have had AGCs 'ramping up' and causing issues > > in the past. > > Hmm - is there a software AGC anywhere? > > Rowland. > (who likes AGCs, when properly applied) You can get software AGC in various apps, yes, but it would kill audio applications if it had been put in hardware. -- Woody
From: Rowland McDonnell on 19 May 2010 07:42 Adrian Tuddenham <adrian(a)poppyrecords.invalid.invalid> wrote: > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > Adrian Tuddenham <adrian(a)poppyrecords.invalid.invalid> wrote: > > [...] > > > A ferrite ring on the audio lead might help too. > > > > Sounds like a more fundamental problem than that. > > I was thinking that if there was a lot of common-mode HF hash coming > from the computer into the output socket of the recorder, it could be > being rectified by the output stage if there was an earth on the rest of > the recorder chassis. More sophisticated than my thinking, that - but of course you are quite right. People who might wonder about that sort of thing might like to consider this: when my dad's old Ferrograph amplifier is plugged into his old Lowther speakers and kept switched *OFF*[1] (and it's a real off switch), it'll work as an accidental radio receiver. We got to listen to the tranmissions of the radio ham over the road clearly audible in crystal radio set fashion - the sound energy being picked up by the speaker leads as EM energy inducing an electrical current aerial-fashion, converted to something the speakers could respond to by the output stage of the amp. (There was no problem audible with lesser speakers, but Lowthers are *efficient*) - until my dad asked him to move his aerial to the back of his house so as to give a wee bit of screening. That reduced the problem so you couldn't hear the words unless you shoved your ear right up against the speaker. > A ferrite ring might reduce it to the level where > it is no longer rectified (less than 0.6 volts). The ring wouldn't > affect wanted signals or differential-mode interference. Ah - yeah, okay. But I can't help feeling if that were the situation, then it would indicate a fault condition in the Mac. > > > Is the 'earth' of the system actually earthed at all? > > > > If the signal's inside a Faraday cage the whole way, that shouldn't > > matter. Should it? > > You're quite right, it shouldn't. In practice the screening of domestic > devices is far from perfect Yup. That's the bother, innit? I've seen people use kitchen foil to improve screening around the back where the connectors plug in (just wrap the whole lot up with an earth wire shoved in to the mess). - and when one of those people is doing the same physics with electronics degree that you're doing yourself, you say things like `Why not give it a differential input?' to which the answer is `This is dead quick, dead easy, dead cheap, and deals now with the problem I've got now'. > so, for instance, the exposed ends of the > wiring on the back of the tape head can 'see' the operator's hand and > other earthed objects in the vicinity. Now you mention it, I remember that, and really heavy screening around *that* ex-Beeb's reel-to-reel tape recorder's heads - shrouded as best could be done. > With capacitive coupling, the effect gets worse as the frequency goes > up, so the sharp edges of the square waveforms on computer bus bars or > from switched-mode power supplies at many Megacycles could be turning up > at the level of several volts in a circuit designed for a few kilocycles > at a few tens of microvolts. Oh lordie - that bad, is it? I've tended to take the line that if there's nasty pick-up, I'm gonna kill it and I've not investigated the details that carefully (on top of that, things like that don't half change when you hook the 'scope up). Well I never. But... I can't see that happening [inside|due to] Mac circuitry unless there's some sort of screening/earthing fault. I mean, they are designed moderately competently. Rowland. [1] This no longer happens, 'cos I've got the speakers and middle bro has the amp and the radio ham moved house anyway. -- Remove the animal for email address: rowland.mcdonnell(a)dog.physics.org Sorry - the spam got to me http://www.mag-uk.org http://www.bmf.co.uk UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
From: Rowland McDonnell on 19 May 2010 07:53 Woody <usenet(a)alienrat.co.uk> wrote: > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: [snip] > > Hmm - is there a software AGC anywhere? > > > > Rowland. > > (who likes AGCs, when properly applied) > > You can get software AGC in various apps, yes, but it would kill audio > applications if it had been put in hardware. I don't see why. I process audio on my Mac that was recorded using an AGC, so what do you mean? Rowland. -- Remove the animal for email address: rowland.mcdonnell(a)dog.physics.org Sorry - the spam got to me http://www.mag-uk.org http://www.bmf.co.uk UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
From: Rowland McDonnell on 19 May 2010 07:53
Woody <usenet(a)alienrat.co.uk> wrote: > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > Woody <usenet(a)alienrat.co.uk> wrote: > > > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > > > > Woody <usenet(a)alienrat.co.uk> wrote: > > > > > > > >if it is a mains loop, > > > > > > > > Earth loop, you mean - via the mains wiring is the usual way it happens > > > > in domestic audio. > > > > > > Earth loop, yes. > > > > For those who don't know: > > > > If you get a circle of wire and change the magnetic field inside the > > loop, you'll induce a current in that loop. > > > > The bigger the rate of flux lines cutting the loop, the bigger the > > current change (basically, the more changing magnetic field, the more > > the induction). > > > > So that means the bigger the loop, the bigger the current change. > > > > Big loops pick up lots. > > > > Think about a loop involving your mains wiring and the signal wires > > connecting units in your music gear. > > > > Big loops. > > > > So loops via the mains wiring have been known to add /loud/ hum. > > > > Professional audio gear uses what they call `balanced' lines, aka > > differential signalling, where you don't refer the signal to earth but > > refer it to the other line (one line goes up while the other goes down, > > and vice-versa). This avoids any chance of an earth loop (and other > > problems). > > Almost apart from the last bit. Eh? > Balanced lines reduce the signal due to > general interference, mains hum, florescent light buz, switching noise > etc, especially as they tend to be low impedence. That's not why. The benefit is that most interference that is picked up is - assuming twisted pair cabling - induced identically on both wires. But since the signal is the difference between the two wires, anything that adds `something the same' to both wires adds nothing to the actual signal, assuming perfect transceiver circuitry (which does of course not exist). > Earth loops are where two connecting pieces of audio equipment are > connected to the mains. No, earth loops are where two bits of audio equipment connected together result in a loop along the earth wiring. The fact that you get a loop enclosing a varying magnetic flux is the problem here. You can get round it by making up signal leads which are only earthed at one end - but that way, you're relying on referring to `mains earth connected via the power lead' as your signal reference point in the piece of gear that's not got `signal earth' via your signal lead. And that's not good, as you point out below. I've just read the Wikip article on what it calls ground loops (earth loops in English) - and it's not using terminology in a fashion that I think is correct. That is, the article correctly describes a real problem insofar as I've bothered to check, but it's not the one that I think should be called `earth loop' in the sense I'm using it. > As they all tend to be earthed, Oh yeah? I've got more than one bit which isn't - a cassette recorder's the only such that I've got wired up at the moment. >you can get a > problem where the earth that two pieces of equipment connect to are > different or far away. That can cause an earth loop and mains hum > regarless of how you connect, althouh a balanced lines can reduce it. The earth loop problem is the magnetic induction issue that I describe. The `earths at different potential' noise source is a different mechanism - one that I had neglected to mention because, erm, it's not the earth loop problem that I was talking about. Rowland. -- Remove the animal for email address: rowland.mcdonnell(a)dog.physics.org Sorry - the spam got to me http://www.mag-uk.org http://www.bmf.co.uk UK biker? Join MAG and the BMF and stop the Eurocrats banning biking |