firstly thank you all for your helpful suggestions. Second I didn’t know I could do a spectral Analysis with Ardour, cool. My computer is a desktop Ac powered. My motu is an M2 and is usb powered. I find the most distressing thing is I can see the noise in the vu meters of the motu with everything off ecept the computer and the powered speakers. The only cables attached to the motu are the usb , and two xlr cables (brand new) from the rest of my system which is powered off. The powered speakers don’t matter as in the noise is still there if I unplug them. I believe I have always had dirty power with problems with ac hum. That’s why my latest try to fix this was to add a pyle power conditioner to all my equipment ie: drum machine , guitar amps, pa amp, mixing board, rack mount effects, and keyboard. As I said before the noise level is the best it has ever been, (really bad before) . it only goes away if I unplug my xlrs. My computer is not using the power conditioner. My setup uses very long cables as my computer setup is on the other side of the room from my music equipment. I switched to zlrs from 1/4 inch thinking they would be better shielded. Before I had an old peavey amp with bad filter caps so when I went digital that had to go. Now you have me thinking. My mancave/studio are in the basement below my kitchen but there is no kitchen electrics directly above me. The refrigerator is the only thing above me that runs constantly and it’s at least 20 feet away. I think I might need to investigate further. Thanks again all.
Andy
So IF this is in fact a ground loop (Not yet 100% convinced it is), the equipment being off won’t really make a difference. Unplug the XLRs and see if the problem disappears or not.
Hate to tell you, those Pyle conditioners don’t do crap. They are MOV based surge suppressors at best, and marketing BS.
Does it go away if you leave the XLRs plugged in and disconnect them at the other end? This does make me lean more strongly towards a ground loop as a possibility, or a gain structure issue.
Yep moving more towards ground loop.
Less about the shielding and more about the signal, is it balanced or not. What are the XLRs plugged into?
Seablade
The xlrs are plugged into the xlr outputs from my mixing board. Sad to find out my money on the power conditioner was wasted. I have only unplugged the xlrs from the motu and not the board yet. I can try that tomorrow it’s pretty late here and I need some sleep.
Andy
You can get a ground loop over USB. Even on data lines. There are MEMS isolation transformers that protect against this for speeds up to USB2.
https://www.microchip.com/en-us/development-tool/tket001
In addition to the single coil pickup sniffer mentioned upthread a cheap battery powered AM transistor radio tuned between broadcast channels is handy for locating sources of electronic noise.
And a cable tester like the Behringer ct100 is nice to have.
The first step is to find the source of the noise before deciding on an approach to attenuate it.
There’s a company called Morley who mostly make guitar pedals etc but they’ve a few products for identifying and removing hum. If you type Morley Youtube humno into Google, you’ll find some useful videos they’ve made.
The computer’s power supply could also be a cause of noise.
I have a thinkpad workstation and it makes instrument pickups hum from quite a distance, depending on the direction in which they are pointing, though generally the system is quiet.
Also long cables often cause noise problems.
Thirdly I found that my Roland UA22 audio interface picks up all sorts of noises (generally higher frequency) unless it’s case is grounded to a good mains earth from the same ac outlet as the computer and monitor speakers.
It may help to find the source of your problems by:
starting with just the computer, see if you can hear any mechanical or audio noise noise from it,
then add the MOTU soundcard with the input level knobs turned right down, (switched to line input level if there is any instrument or mic switch) and listen on headphones. Even try moving the MOTU around a bit to see if its picking up any hum from something external.
Then connect your monitor speakers, one at a time.
If that’s ok just connect the other stuff one at a time and listen
Hoping this may help you
Will be interested in the results. Based off all the things so far, it seems likely the noise will dissappear when you do this, and in which case 1:1 transformers on the XLR line shields should address your issue. There are a few not to expensive devices for this purpose. You could also cut the shield on one end of the XLR (as @x42 alluded to above) and it would likely fix it if the test above is successful, though the transformer are the technically ‘correct’ way to do it.
Seablade
I think the “too wordy” comment was for the lack of formatting more than the word count. There are lots of well-written long questions over there that work just fine, but a “big block of text” is hard to read. Try to break it up, put a blank line between each individual point, etc.
A “trickling stream of water” could be data-compression artifacts, or a noise suppressor/reducer/repellent working too hard. If you really need that aggressive noise reduction, and you can’t fix it acoustically or in the analog circuitry up front (gain structure…), then it tends to work better to have multiple less-aggressive noise removers in series. One feeds the next, and each does just a little bit.
But it’s still MUCH better to fix the acoustics and other physical stuff, so that you don’t have that noise to start with.
For the sample that you posted, I hear 3 different noise sources:
- Raw AC power: low frequency buzz, probably a ground loop
- Switching power supply: high frequencies, also probably a ground loop
- Generic white noise: analog circuitry just does that
For 1 and 2, the solution is probably the same, and you’ll probably fix both at the same time. Somehow or other, you need to break the signal ground, and still get the signal across. (DO NOT EVER BREAK A SAFETY GROUND! Always keep the ground pins on power cords, and never use a “ground lift adapter”. Those are meant to provide a ground, via the tab, not to remove it.) As mentioned before, the power state (on or off) of the connected gear is irrelevant for a ground loop. It’s all about the wiring.
And yes, USB and other digital stuff can be involved in a ground loop. The digital signal itself is immune to that noise, but if there’s anything analog involved at all, that’s where the ground loop noise gets in. So isolating the digital thing may in fact be the easiest way to solve it. Been there done that myself.
For #3 - the generic white noise - see where you are after the ground loop is fixed, but you may also need to tweak your gain structure. You can never get rid of the hiss completely, but it should be low enough that you don’t notice.
General best practice is to have all of the required gain right up front, as the first active thing at all, and then keep that same nominal level all throughout everything else. Hence the mic preamp that has a TON of gain available in just that one stage. Use that, all by itself, to set a good working level, with absolutely nothing else involved at all, and then see where you are.
Slight nitpick: analog snakes. 150 feet of both mic- and line-level signals (a few even have big wire for speakers!), all bundled together for the entire run, with a TON of amplification on the destination end…and they stay quiet.
The secret is that the whole thing is balanced. Not equal and opposite signals, but equal impedance. The long run itself doesn’t reject noise; it comes right in almost unimpeded! But because both signal wires for a given signal have the same impedance, they pick up that noise equally. So when the receiving end takes the difference, the noise drops out.
If you don’t pay attention to that, and end up with a long UNbalanced run somewhere (one signal wire per signal, either physically or effectively because one is shorted to ground), then yes, that probably will be noisy! But a long balanced run can go the full length of an auditorium, past some high-power triac-dimmed lights even, and still be perfectly okay.
That’s also how the analog telephone system worked: all balanced wiring for a VERY long distance, past who-knows-what…
Pro audio inherited a lot from that industry, which explains quite a few terms and conventions that seem odd now.
Oliver Heaviside. We stand on a giant’s shoulders.
You need to start doing elimination.
Disconnect the XLRs from the mixing desk, leave them connected at the MOTU side. You should have no noise then. (If you do, invest in a cable tester to verify it’s a properly balanced cable).
Then, connect only ONE of the XLRs to your mixing desk and listen. This eliminates at least one of the paths for a ground loop. Also, look up the manual for your mixer and find out if those XLR outs are balanced or not. They MUST be balanced!
Next thing I would try is to take out the mixing desk as possible source; buy a cheap DI box for $30, plug in your long XLR, and nothing else. You should have no noise now.
Next, take your guitar (ONLY your guitar; no FX, no amps!!) and plug it into the Hi-Z jack input of that DI-box with a known-good SHORT (max 10ft!) shielded jack cable. Your guitar sound should have little to no noise.
…
Ooof!! I listened to the audio just now. That’s crazy noise!! I think you can eliminate a ground loop as real cause, though ground loops exacerbate any noise. This not a 60Hz sine wave, if anything it’s a 60Hz square wave. This almost certainly originates from a very very bad switching power supply/inverter somewhere. Could be your mixer’s power supply, your computer’s power supply, but also a badly shielded solar panel inverter, charger, or some heavy industry equipment.
But since it’s not a sine wave, it must come from a switched mode power supply or similar. I would try to see if the direction your guitar faces makes a difference. If not, I’d highly suspect one of devices’ power supplies. Mixer, amp, USB hub, computer, etc.
Ground Loops involving computer power supplies and signaling can sound very different from ground loops in other cases, so I am not ready to eliminate ground loop at this point. This high distortion signal isn’t uncommon (Albiet acting a bit differently) in ground loops involving computers in my experience, though I haven’t heard this exact sound as a ground loop before so it isn’t a give either.
I’m with Seablade. Ground loops can sound like anything. They don’t have any sound of their own, but pick up whatever’s around them. Usually that’s an AC power line - and that really is what the power line sounds like - but it could be anything.
As for why the power line sounds like that, I think a lot of it has to do with power supplies, both switching and linear. Both kinds draw power intermittently:
- Linear and cheap switchers are both just before the voltage peak and nowhere else. Linear has a transformer up front to reduce the voltage, and then it’s rectified, but the power line still sees that through the transformer. Cheap switchers rectify the incoming AC immediately, and then chop that down. In both cases, the rectifier only conducts when the incoming AC is higher than the capacitor that it’s keeping charged. Otherwise, the entire circuit runs off the capacitor and draws nothing from the wall.
- Power-factor-corrected (PFC) switchers have a series of spikes all throughout the waveform. They try to approximate a current draw that follows the voltage at each instant in time, but it’s only an approximation. It does that by rectifying without a (significant) capacitor, so the voltage still fluctuates wildly, and then it boosts that to an intermediate DC supply that is higher than the highest AC peak. That booster must necessarily chop it up, and it tries to make each “packet” proportionally sized to the voltage at that moment. Then the intermediate high-voltage DC is chopped down as before, with no further concern.
- Proofreading this just before posting, I realize that this could be the source of the “generic white noise” or “hiss” that I hear in the noise sample. If a switching power supply also randomizes its switching frequency, in addition to everything else it’s doing, then it converts a strong discrete set of frequencies into weaker noise. Same total energy, but more spread out in frequency so that the peak level is lower. That’s often called “spread spectrum”, and used not only to reduce audible noise, but also to reduce radio noise. So the entire problem in this thread could be a ground loop.
If you put an oscilloscope on most power lines, you’ll see that it’s not a clean sine wave. The leading edge of the AC peak sags a bit, just before the peak, and the above is why. And anything that is not an exactly perfect sine, has harmonics. The resulting harmonic content…sounds like a classic ground loop because that’s usually what a ground loop picks up.
The strong audio-midrange - fairly high-order harmonics, given the low fundamental - are from the sharp edges of the rectifier diodes cutting on and off. There are ways to reduce that - to make a softer edge - but they involve additional parts in each and every power supply that don’t make the supply itself work any better. So it’s usually not done.
But like I said, ground loops can pick up anything, and sound like whatever that is instead. Some switching power supplies have low enough switching frequencies to dip into the audio range. If a ground loop picks up one of those, then it sounds like a high-pitched drone or warble as the power supply’s load changes, but it’s still a ground loop. (see my #2 in a previous post in this thread)
As long as you keep good “noise hygene” though, none of that matters. You don’t get signal noise from the output of a power supply and into the processing circuitry through its supply rails, no matter how filthy the incoming AC is, but you can pick up its normal operating noise roundaboutly, as we’re discussing.
I think that falls under my “though ground loops exacerbate any noise” remark. Meaning, a possible ground loop will act as a conduit for this noise, but the ground loop isn’t the main issue here, the issue is where that noise originates from.
I have never, ever seen or heard such a 50/60Hz waveform from a power supply of any type. Switching mode power supplies can well produce square waves, but they would do that at their switching frequencies, in the kilohertz range. Never at the base mains frequency. This is totally weird.
And anyway, I would still proceed the same: eliminate the mixer first, by using a DI-box at the end of those same long cable runs. See what result that yields first.
Balanced connections technically do not have a “signal ground,” they have signal hot, signal return, and shield. Much equipment improperly handles the shield connection and intermixes noise on the shield into the audio circuitry, which is why disconnecting the shield at the receiver end is often a way to avoid the problem.
For that to work you need to make sure the AC voltage difference between the devices on each end of the connection is low, which it usually is if both devices have three-wire power cords and are therefore connected to the power line protective earth.
If the device on one end only has a two wire power cord, then disconnecting the shield will make the noise much worse, and the best approach there is to run an additional conductor between the device chassis to decrease the resistance between the devices and keep the common mode noise as low as possible.
The picture I found online of an Xtuga mixer showed an external power supply with only a two wire power feed, so first it would be good to check the XLR cables and make sure that the shield to pin 1 connection is present on both ends of the cables, because if it is not it can result in wide frequency power noise.
I could not find any additional information about that mixer. There is a listing for an Xtuga CT120 12 channel mixer on Amazon, but the xtuga-audio.com web site does not have a listing for that mixer, so no way to check for any more detailed specs or a user guide.
There are several things happening at the same time here. Like I said when I first came in, I hear 3 different sources of noise all at the same time.
And yes, cheap switchers do produce mains-frequency noise for the same reason that linears do. Delete the big low-frequency transformer from a linear supply, so you end up with non-isolated high-voltage DC, and then switch that down through a much smaller high-frequency transformer and a second rectifier. It’s still a linear front end that produces mains-frequency noise, but called “switching” because of the back end.
My present thinking is (in the same order as my first post here):
- Linear-supply buzz, either from actual linear supplies or from cheap switchers.
- Actual switching noise that just barely dips down into the audible range.
- Spread-spectrum “hiss” from a more expensive switching supply that actually cares about something beyond itself. (or might be required to care because of its power rating)
ALL of that is perfectly normal operation, and not really solvable as such, but what makes it a problem is a ground loop that picks up all three. Fix that ground loop, and the same recording rig becomes (practically) silent.
I can’t imagine a buck-boost supply that doesn’t switch at high ultrasonic frequencies.
The whole point is to use the smallest possible transformers.
They’re supposed to, and yes, that is the whole point. (or at least a big part of the point) But cheap transistors don’t switch quickly, and so a lower frequency is needed to keep them out of the “in-between” region where they waste a lot of power. Apparently, it was cheaper to use a slightly bigger transformer…
Maybe an older engineer couldn’t hear it and called it good? Or maybe everyone could hear it, but the ones who cared got overruled? Several possibilities there.
Oh come on! Buck-boost regulators are as cheap as chips.
https://www.google.com/search?q=buck+boost+regulator+digikey
Yeah I know about ground loops. I even had to prepare a special “IEC/C13” power cable with ground wire cut(*) to supply one of my screens, as that turned out to cause my severe ground loop. Which is totally weird, because that is along a digital path carrying no audio whatsoever, to a device playing no part in the audio chain. And my other screen (dual screen setup) had no such issue. But it 100% caused the ground loop noise in my studio.
(*) Do NOT try this at home, folks