Need help identifying and eliminating this noise!

[Ben Asaro]

I hope this is the correct forum for this question. I have a really harsh noise which appears on certain pieces of kit and I am having a devil of a time identifying what it actually is. Is this USB noise? In the clip below, you can hear that it is incredibly loud on some things, sometimes it's barely noticeable, and other times it's non existent.

Signal chain:
Input
Focusrite OctoPre > ADAT > RME Digiface > USB > Mac Mini M2

Output
Mac Mini M2 > RME Digiface > Behringer ADA8200 > Neve RNHP (headphone output from the RME)/Adam audio monitors (from Behringer)

Audio clip:

https://audio.com/don-music/audio/noise3

Thanks in advance!

[Kwackman]

Does anything in your chain have a connection to earth? Maybe it needs one?

EDIT - ignore the above, it looks like the OctoPre DOES have an earth.

[James Perrett]

Since much of the kit is already optically isolated from the computer it won't be USB noise. The big question is - where is this noise appearing? Does it happen on the inputs or are you hearing it on the outputs. Are the cases of the computer and any of the other devices touching? Is the case of the Digiface USB touching anything?

What happens if you disconnect everything from the Octopre and Behringer and then re-connect things one at a time?

[Ben Asaro]

Since much of the kit is already optically isolated from the computer it won't be USB noise.

Thanks, James!

Unfortunately, I agree; I think this is 5G interference from cell transmitters on the roof.
The reason I say this is because I can even hear it from portable, battery-powered speakers in the house as well with no source attached at all, just with the power on.

The big question is - where is this noise appearing? Does it happen on the inputs or are you hearing it on the outputs.

I am not 100% sure I can answer this question, but my system is generally very quiet -- for example, right now I have the monitors turned up full and only have a touch of thermal noise coming from them.

If I plug in an offending device, I see the noise straight away; the OctoPre's LEDs will light up, even before sound is being generated. The sound clip I posted is basically me just turning the gain up, then down again, on the OctoPre.

Are the cases of the computer and any of the other devices touching? Is the case of the Digiface USB touching anything?

The OctoPre and ADA8200 are rack mounted into my desk bridge. The Digiface USB is under my desk, not physically touching anything else making sound.

What happens if you disconnect everything from the Octopre and Behringer and then re-connect things one at a time?

I don't hear anything unless an offending source is connected; for example, the JV-1010 picks up a lot of the noise, as does the Lofiatron. See my above comment about the portable speaker (powered by AA batteries).

[Ben Asaro]

I just conducted an experiment where I switched off all of the breakers in the house, and then turned the portable speaker on, and sure enough -- same noise.

[ajay_m]

In my experience, (which in my case was being line of sight about 5 miles away from Auckland's primary TV transmitter tower), RF interference is very difficult to deal with. You may find clipping ferrite beads onto all the cables helps (you can buy these on eBay, they clip over the cable) but sometimes you have to construct a Faraday cage to really shield stuff.

[Ben Asaro]

In my experience, (which in my case was being line of sight about 5 miles away from Auckland's primary TV transmitter tower), RF interference is very difficult to deal with. You may find clipping ferrite beads onto all the cables helps (you can buy these on eBay, they clip over the cable) but sometimes you have to construct a Faraday cage to really shield stuff.

Thanks Ajay! I have ferrite beads everywhere; they DO help, but it's not a panacea. Unfortunately, I think it depends on the power supply.

Sadly, I'm on the top floor of an apartment building in NYC; so I'm VERY close!!

I wonder if I try using a better power supply with some pieces, like the JV-1010, if that will help? Hmmm ...

[Wonks]

Time to contact the FCC (if they still have any staff)?

[Ben Asaro]

Time to contact the FCC (if they still have any staff)?

HA ... I don't think they will be much help.

[James Perrett]

So this looks like RF interference. To eliminate it you are going to have to be using properly designed gear with properly designed cables and a decent grounding scheme. Gear that is powered from a wall wart (like the JV1010 isn't likely to be well grounded so you would need to think about how to improve the grounding.

I'd start by going through all your cables - are they decent cables from a broadcast supplier or do they come from a music or hifi shop? Music shop cables often look impressive but when you open them up they turn out to be a flimsy cable hiding in a large rubber jacket. Given the frequencies involved, you may need to use foil screened cables rather than braided screens. You may even have to think about starquad cables as they are designed for high RFI environments.

I would also make up some shorting XLR connectors which should help you decide whether a device is sufficiently RF immune. Plug a shorting connector into the input of the Focusrite and turn up the gain on that channel. Is the noise there? If not, try plugging a cable into the Focusrite and shorting the other end. You may well be able to find which cables are good and which are bad like this.

[Hugh Robjohns]

It could well be RF interference, but it sounds a lot like computer ground noise too.

Plus, 5G masts are generally designed to radiate as a cylinder, meaning not much energy goes directly downwards into the building... although obviously you are very close.

So... what to do?

You need to start by stripping the system right back to its simplest form — interface and computer — and check that is noise free on record and playback. In particular, make sure the computer and interface have good grounds via the mains safety earth.

If you're confident of good grounding, but still getting noise, it might be worth wrapping the interface (and possibly computer) loosely in aluminium foil which will exclude and RF fields and should help identify where the noise is getting in.

Once the core system is satisfactory, you can start adding back the other I/O devices, one by one, and retesting.

Anything connected via ADAT won't get its ground from the computer/interface, so you'll need to make sure it has a good ground via its mains supply or some other bespoke route.

Again, check for noise when properly grounded, and try the aluminium wrap if you're still getting noise....

If the wrapping reduces the noise it indicates that RF is the problem. If so, Ferrites on all connecting leads (including power) might help... but it's usually a fundamental design issue, and you either need a better engineered product, or build a Faraday cage into the room!

If you can get the core hardware quiet, you can move onto connecting the sound sources and destinations. Again, connect one by one, checking as you go.

Where possible, use balanced connections with decent cables. As explained, James' shorting XLR is a useful tool (just link pins 2 and 3) for identifying where noise is getting in.

Hopefully, by working through each part of the system sequentially, you'll be able to identify where the problem(s) lie and stand a better chance of resolving it/them.

[Jimmy B]

It could well be RF interference, but it sounds a lot like computer ground noise too.
-------------------------------------------------------------
In particular, make sure the computer and interface have good grounds via the mains safety earth.

Another question could be if the cell site on the roof is polluting the AC supply and ground in the building.

It would be interesting to know more about the cell site installation. Is there also a cabinet full of base station electronics on the roof? If so, how is it powered? And where does it get its ground reference?

Either way, I think that I would be tempted to ask the cellular company to investigate. Most companies don't want a bad press, and the worst that they can do is say "no". Then you can escalate to the FCC, as Wonks said.

[resistorman]

I have a friend who was experiencing noise from a nearby radio station. Here in the US we use grounding rods on the mains, currently 2 of them 10' apart. I checked his and the outside connecting wire had corroded off, a simple fix.

[Ben Asaro]

So this looks like RF interference. To eliminate it you are going to have to be using properly designed gear with properly designed cables and a decent grounding scheme. Gear that is powered from a wall wart (like the JV1010 isn't likely to be well grounded so you would need to think about how to improve the grounding.

I'd start by going through all your cables - are they decent cables from a broadcast supplier or do they come from a music or hifi shop? Music shop cables often look impressive but when you open them up they turn out to be a flimsy cable hiding in a large rubber jacket. Given the frequencies involved, you may need to use foil screened cables rather than braided screens. You may even have to think about starquad cables as they are designed for high RFI environments.

I would also make up some shorting XLR connectors which should help you decide whether a device is sufficiently RF immune. Plug a shorting connector into the input of the Focusrite and turn up the gain on that channel. Is the noise there? If not, try plugging a cable into the Focusrite and shorting the other end. You may well be able to find which cables are good and which are bad like this.

Thanks, James!

[Ben Asaro]

It could well be RF interference, but it sounds a lot like computer ground noise too.

Plus, 5G masts are generally designed to radiate as a cylinder, meaning not much energy goes directly downwards into the building... although obviously you are very close.

So... what to do?

You need to start by stripping the system right back to its simplest form — interface and computer — and check that is noise free on record and playback. In particular, make sure the computer and interface have good grounds via the mains safety earth.

If you're confident of good grounding, but still getting noise, it might be worth wrapping the interface (and possibly computer) loosely in aluminium foil which will exclude and RF fields and should help identify where the noise is getting in.

Once the core system is satisfactory, you can start adding back the other I/O devices, one by one, and retesting.

Anything connected via ADAT won't get its ground from the computer/interface, so you'll need to make sure it has a good ground via its mains supply or some other bespoke route.

Again, check for noise when properly grounded, and try the aluminium wrap if you're still getting noise....

If the wrapping reduces the noise it indicates that RF is the problem. If so, Ferrites on all connecting leads (including power) might help... but it's usually a fundamental design issue, and you either need a better engineered product, or build a Faraday cage into the room!

If you can get the core hardware quiet, you can move onto connecting the sound sources and destinations. Again, connect one by one, checking as you go.

Where possible, use balanced connections with decent cables. As explained, James' shorting XLR is a useful tool (just link pins 2 and 3) for identifying where noise is getting in.

Hopefully, by working through each part of the system sequentially, you'll be able to identify where the problem(s) lie and stand a better chance of resolving it/them.

Thanks, Hugh!

The reason I think this is airborne is bc I hear the noise on my portable Blackstar loudspeaker with nothing connected and running on batteries.

The noise persists when I leave the apartment and walk into the hallway. It finally stops when I enter the stairwell to the building.

Also, I heard the noise on the Blackstar (again, nothing connected to it, and running on batteries) in the apartment with every breaker in the house turned off.

[resistorman]

Either something is directly polluting the AC or it's in the ether. Either way, the building ground is suspect.

[Ben Asaro]

Either something is directly polluting the AC or it's in the ether. Either way, the building ground is suspect.

There is absolutely no doubt about the grounding. The building is old, the wiring is ancient, and the electrician the building hired to put in our last electrical panel was a joke. I will probably end up hiring an electrician to put in a better panel.

[merlyn]

As the JV1010 is an offender if you concentrate on that and find a solution you can apply that to the other gear. Is the noise there if you use headphones straight into the JV1010?

I would assume it is. Is it an all metal case? I think it is, so tinfoil wouldn't help there I don't think. Are you using the Roland power supply? I have a Roland power supply for my guitar synth and it's pretty good.

5G has bands up to 66GHz. That's not audible so what you're hearing is demodulated into the audio band. Two things have to happen: (1) a piece of gear acts as an aerial (every piece of wire in your house is acting as an aerial) (2) some non-linearity demodulates the high frequencies into audible frequencies.

RF is not like audio frequencies, which are relatively well-behaved, and a non-linearity could be down to geometry, or something like the rusty bolt effect.

Try a few things, but it is possible that due to a quirk of its design, the JV1010 just can't be around 5G. (Without going full tinfoil wallpaper).

[Ben Asaro]

Try a few things, but it is possible that due to a quirk of its design, the JV1010 just can't be around 5G. (Without going full tinfoil wallpaper).

Thanks for these insights!

[merlyn]

Something to try is grounding the case of the JV1010. That is connecting the case to the mains ground. Bad though your ground may be, it appears to work for the other gear.

[Ben Asaro]

Update: Purchased a USB isolator to place after the Digiface just to see if it made any difference. If it did, it's negligible.

I conducted a little experiment yesterday and got the worst offenders onto DC power banks and did some recording outside .... AND -- a block away from my building -- I was still picking up trace amounts of the 5G RFI on the tape cassette recorder!

However, this only seemed to happen when I engaged fast forward or reverse? In other words, the noise didn't make it onto the tape, which does happen when I'm at home.

So the testing and experimentation continues!

[Ben Asaro]

An interesting development. The AC power bank arrived yesterday and I gave it a quick test today with the L-8 and a single skiff.



With the skiff plugged into the AC receptacle and the L-8 plugged into one of the USB ports, there was a lot of USB noise. Switching the L-8 to a an AC adapter almost totally eliminated any noise. Using a 5V DC power bank w the L-8 and the AC power bank on the eurorack, the only noise was self-noise from the eurorack, and then was only audible when I had the input trim SUPER high going into the L-8.

I'm not sure if this will be a fix for recording in the house, I was mostly testing this for use remotely/outdoors, but it's certainly interesting. Will see how this develops!

[Hugh Robjohns]

Have a think about the different grounding arrangements in your different scenarios. There's logic in the madness...

[Sam Spoons]

I have one of those though mine has different branding and UK power sockets.* My intended usage was for busking/live sound rather than recording so not so critical and while I haven't really used it in earnest in testing with various single bits of audio gear it has been quiet as a mouse. As the intention was to use it to power just the mixer when driving a pair of Bose S1 Pro's earth loops should not be an issue.

* After buying a dodgy one which claimed to have a pure sine wave inverter but proved to be very noisy and since the inverter has subsequently failed I can't stick the oscilloscope on it to see for sure. I got a refund from Amazon on that one and, as the "UK" seller refused to give me a UK return address, I got to keep it. The USB outlets still work fine as does the wireless charging pad so it is still has it's uses.