[F] Very loud hum in audio output when connected to computer

[Holyver]

Status: this is solved by updating hardware in Nome II.
If you still experience noise issues in your setup with a Nome II, there is very little chance it is cause by the Nome - so check your other devices / cables / etc. And avoid ground loops generally if you can.

For Nome I (Midronome), this is unfortunately not solvable, connecting the USB cable to the same ground (say a computer) as the audio output (connected to the computer through an audio interface) will create a ground loop which will induce noise in the audio output.

If you experience this issue please read this thread for the many ideas on how to work around the noise. Here is a summary (try these in this order):

• Make sure the audio volume is high, ideally maximum (on the Midronome but also on your other devices) to get the best signal-to-noise ratio
• Use TRS cables in your setup, in particular on the Midronome Audio output (set the switch to "speaker" = Balanced Line Out)
• Power the Midronome from a USB wall plug / USB charger - if needed use a splitter USB cable like the one coming with the Arturia Keystep
• Use a hum eliminator / USB isolator on the USB power supply

Feel free to post in this thread if you struggle with it.
---


Short description of the issue:
When the Midronome is connected via USB to the computer, there is a very loud hum in the audio output when connected via an unbalanced (TS) audio cable. It's so loud I would consider it unusable to use as a click track.

Setup (devices connected to the Midronome and how):
PC > USB Cable > Midronome
Midronome > Unbalanced Audio cable (TS) > Mixer input

I've tried multiple USB and audio cables, and multiple USB-ports, with or without (powered) USB-hub.

Steps to Reproduce:
Connect as described above and listen.

In the following cases I have no noise:

• When connecting the Midronome with a TRS-cable straight into the mixer.

• When listening to the audio output via headphones.

• When powering the unit with a separate USB-charger, without connection to the computer, there is also no noise, even when using an unbalanced audio connection.

I have also connected another USB-powered audio device (Roland VT-4) in place of the midronome (so exactly the same cabling), and it has much lower noise (I would not call it perfect, but it's not disturbing at all). That makes me feel like something is clearly different (and less optimal) in the way the connections and signal flow are set up in the Midronome.

Expected Behavior:
I'd love to only hear the click track, no noise (or way less).
I can accept that I have a suboptimal setup, with most of the cables being unbalanced, but midronome is the first device that really gave me troubles with that.

Actual Behavior:
I made a recording:
https://drive.google.com/file/d/1WvjKrW ... sp=sharing

Note:
The noise varies slightly in volume and timbre depending on what is shown in the LED-display: the more segments are lit, the louder the noise becomes (the audio recording starts at 71 BPM, which has only 5 segments in the display, and ends at 128, which has 15 segments).
Playing with the brightness setting also gives you a nice change in timbre.
The volume of the click is 5/4 in the recording.

ErrorLog (from the ErrorLogExtractor - https://midronome.com/support ):
N/A

Has the Midronome firmware been recently upgraded, from which version to which version?
I upgraded to the 1.0 firmware when I received the unit.

[Simon]

Hi Holyver

Yes I'm afraid that's all correct It's a ground loop, it's like when you are at a gig and draw an unbalanced cable from the mixing table to the stage, you get 60Hz or 50Hz noise from the AC transformer. Except that here the noise is from the display. But it's on the ground path, not on the signal line.

This will be improved in a future version of the hardware, changing it for this firmware would have delayed the shipping by too much.

But as you said yourself there are many workarounds:

• Power the Midronome from a separate source (plug on the wall, separate computer, etc.)
• Use a TRS cable
• use headphones

For other people reading this - the same kind of noise can appear when sending clock to synth via the ANLG plug, if you are using unbalanced audio cable, but it should be much lower.

[Holyver]

Thanks for your reply, Simon.
That's too bad. I'm not yet sure how the midronome will end up being used in my setup (but it will most likely involve DAW Audio Sync), so it's hard to say if that will prove to be a problem. Time will tell.

[beatbastelbaer]

Hi,
this is a groundloop problem.
If you need the USB connection to your computer you can solve this with an USB Y adapter cable like the one that comes with the Arturia Keystep 37, this one has USB B connectors.
https://synthanatomy.com/wp-content/upl ... 37-Hub.png

I couldn´t find this as a spare part like the micro USB Y cable for the Beatstep Pro that is easy available
https://www.arturia.com/store/accessori ... ntcablekit

but it should be possible to DIY make one.
Greetings,
Frank

[Holyver]

Thanks for the additional information.
That Arturia Anti-ground loop adapter bundled with the keystep 37 seems interesting. Too bad they're not selling it as a spare part.

Greetings,
Olivier

[Dave]

I connected the CV out to my Erica Synths Acidbox 3 CTRL IN port in order to sync the Acidbox 3 LFO. However, the connection generates a lot of noise. It's like the AB3 not Midronome, but....

I have a MIDI2CV device that I send quarter notes to from the DAW that converts it to a gate pulse, which also can be used to sync the AB3 LFO. I was excited to replace that little thing and free up a MIDI channel, remove a device, etc. But, unfortunately, using the Midronome CV out in this case is a no go.

Am I using the wrong type of cable? Is there a way to reduce the "noise" on the Midronome side to just the clock pulses? I tried a couple different patch cables and they all have the same issue, so it's one, or both, boxes.

[Simon]

Am I using the wrong type of cable? Is there a way to reduce the "noise" on the Midronome side to just the clock pulses?

Hi Dave

Yes that's unfortunately true, it's caused by a ground loop, which makes the Midronome's display noise take the wrong path.

Basically that noise should be on the USB cable (=power cable). Problem is the ANLG clock connects the ground to your modular synth as well, which is then connected to your computer.

So there is your ground loop: Midronome -> USB cable -> computer -> audio cable -> modular synth -> jack cable to ANLG plug -> Midronome

That means the ground current which should normally go through the USB cable is also going through the other part of the loop (ANLG plug -> modular synth -> computer), hence anything unbalanced on that path will make the noise hearable.

Notice that the noise is already tamed a lot and only a small amount should take the wrong path. Then it depends how loud your audio is, if somewhere you have very low audio which is heavily amplified - that would amplify the noise as well.

To solve it, you can:

• Try to have your audio source (your synth I guess) play louder, so you do not need to amplify it so much
• Use Balanced audio cables anywhere on the loop
• Ideally, break the loop completely: power the Midronome from a different power source (see the suggestion above for the funky USB cable splitter) or put a DI box with a ground lift somewhere on the loop

Simon

PS: I've merged your post with the other topic about the same issue.

[Dave]

Am I using the wrong type of cable? Is there a way to reduce the "noise" on the Midronome side to just the clock pulses?

Hi Dave

Yes that's unfortunately true, it's caused by a ground loop, which makes the Midronome's display noise take the wrong path.

Basically that noise should be on the USB cable (=power cable). Problem is the ANLG clock connects the ground to your modular synth as well, which is then connected to your computer.

So there is your ground loop: Midronome -> USB cable -> computer -> audio cable -> modular synth -> jack cable to ANLG plug -> Midronome

That means the ground current which should normally go through the USB cable is also going through the other part of the loop (ANLG plug -> modular synth -> computer), hence anything unbalanced on that path will make the noise hearable.

Notice that the noise is already tamed a lot and only a small amount should take the wrong path. Then it depends how loud your audio is, if somewhere you have very low audio which is heavily amplified - that would amplify the noise as well.

To solve it, you can:

• Try to have your audio source (your synth I guess) play louder, so you do not need to amplify it so much
• Use Balanced audio cables anywhere on the loop
• Ideally, break the loop completely: power the Midronome from a different power source (see the suggestion above for the funky USB cable splitter) or put a DI box with a ground lift somewhere on the loop

Simon

PS: I've merged your post with the other topic about the same issue.

Thanks! For now I simply used a USB power adapter and plugged it into the power strip instead of computer. That solved the issue for now.

Worst case scenario is that this will just ok, though I know it removes a little potential functionality - but it will still free up a MIDI channel, which outweighs the loss of Midronome functionality over USB. I'm going to look to see if I have a USB cable with that knobby thing on it that would work to reduce/remove the noise. Otherwise... I'll eventually get a dongle thing to remove USB ground loops (I just sent one back as I'm trying to reduce and eliminate other noise so my budget is shot as I buy more gear, like the Radial Stagebug and an iFi Silent Power,etc.).

[muZikman]

Hey kids,

I ran across this issue as well, with my intent being to send line out 1 to a monitor controller (for use in control room) and line out 2 to a headphone amplifier (for tracking space). Both worked, but were very noisy. My Arturia Drum Brute does the same thing. It's the USB thing... not anyone's fault here.

If you're going back into a board/interface, sending through a d.i., first fixes things. I just used a cheapo passive DI and that seemed to do the trick. It was dead quiet, despite needing a little preamp gain to boost back up to line level (just enough to blend in w/ playback audio)

This works for my intended purpose, but doesn't seem to be loud enough w/out a preamp, which is too bad.

Maybe this will help some of you out - but what I'm really hoping is that someone either reports back the USB splitter dongle works well, or that someone finds a better fix than direct boxes.

Cool device, Simon. Cool device!

-Z

[Simon]

Ha awesome thank you for the feedback, Z

[Dave]

Thanks for the additional information.
That Arturia Anti-ground loop adapter bundled with the keystep 37 seems interesting. Too bad they're not selling it as a spare part.

Greetings,
Olivier

The Keystep 37 adapter works well - plus, by having the Midronome connected to the computer, it will also power off when the computer is turned off. Without an on/off switch, if you connect it to a power socket it’ll stay on permanently.

[muZikman]

/

The Keystep 37 adapter works well

Great! Thank you for the report!

Now, next step - who can find one of these doodads for sale? I have a micro-usb version that came w/ my BSP, but I can't find it's big brother anywhere out there. I kinda doubt Arturia invented that dongle...but maybe?

Anyone ever come across something this for sale, or know what it's equal would be?

Thankya,

Z

[beatbastelbaer]

Howdy friends of Midi,

here are some thoughts and ideas about the USB ground loop thing.

The Arturia USB splitter is not exactly rocket science, in the end it is a very simple thingie that devides the USB Power supply and the data signal.
USB uses 4 wires/pins
+5V
Ground
data +
data -
You can easily DIY a splitter cable, see this drawing (replace DAW with Midronome)


just take a USB A to USB B cable (eg. like the one that came with the Midronome)
remove the coating somewhere to free the 4 wires,
identify the signal and power wires,
see this schematic of USB plugs


cut any USB cable with a USB A connector apart, keep the male connector and throw away the other piece (or keep ist for other fancy projects),
free the wires and identify them,
be awaere that some suppliers use different colors for the wires or mix them,
now cut 2 wires from the first cable and connect the the USB B side with the half cut USB A cable.

I made a similar one with micro USB connectors for the Korg NTS-1 and it works perfect.

Good Luck !

Just anotther way for the decadent customer:
buy a used Arturia Keystep37,
keep the splitter,
sell the rest on ebay

[mattbassg]

I've started using my Midronome in a live setting at church. Some of the team have reported the very loud underlying hum as a major problem in their in-ears. I've done some work-round testing and this is what I found - sharing here in case it's of interest / help to anyone.

My Midronome is powered via a USB power supply and connected to the main (live) mixing desk.

Initially I was using an unbalanced (TS) jack to jack cable to connect the Midronome to the desk via a DI box - hum was present and (since we know that the ground loop occurs within the device itself) not surprisingly the ground lift switch on the DI made no difference to the hum.

Caveat: The following results are based on reports from the tech team member who was running the desk. 'Your ears may vary'.

Some improvement
I tried inserting an EBTech Hum Eliminator (now owned and marketed by Morley as this device: https://www.morleyproducts.com/hum-eliminator/) and that did reduce the level of hum, but not stop it altogether (again, nor entirely surprising given the source of the noise).
Signal chain Midronome / TS-TS cable / di box / standard xlr cable / desk

More improvement
I tried switching out the jack to jack cable for a TRS to XLR balanced cable (with a Phantom power blocker - https://www.tritonaudio.com/phantom-blocker - inline just in case Midronome would suffer harm from unexpected voltage back down that cable). This significantly reduced the level of hum, but did not stop it altogether
Signal chain Midronome / TRS-XLR cable / Phantom Blocker / standard XLR cable / desk

No noise
Finally I put everything back to the first setting (Midronome / TS-TS cable / di box / XLR cable / desk) but inserted a USB ground isolator into the power supply to the Midronome. The device I used was a 'Binchil Adum3160 Digital Signal Audio Power Isolator Usb To Usb Digital Isolator' and it goes inline between the USB power supply and the USB cable powering the Midronome - power signal chain USB power supply / Adum3160 device / USB cable / Midronome.This seemed to stop the hum / noise completely. I may try a more expensive and robust alternative to my Binchil device - it was the solution I, wrongly, thought would be least effective - so bought the cheapest item I could find and it works fine, but has a flimsy looking plastic body... hardly 'road worthy'.

Hope that's helpful info for someone.

[Simon]

Wow, thanks Matt and beatbastelbaer for the detailed explanation - I hope this will help those of you who are fighting with this issue as well!

@Matt: I am actually quite surprised by your results I must say? Could you tell us a bit more how your device is connected and to what, and in particular where are the grounds connected?

I just tried the following setup:

Code: Select all

Midronome AUDIO plug -> TS cable -> Audio interface -> ground plug on wall
Same ground plug on wall -> USB power adapter -> Midronome USB plug

And there the noise is completely gone, even with full volume on both speakers and audio interface I don't hear a thing.

And if you need to connect the USB cable of the Midronome to your mixing Desk or a laptop for example (for USB communication), then you can use one of the USB cable splitters as mentioned before (either the Arturia one or a DIY one as pointed out by beatbastelbaer ).

By the way, even if you are using a TS cable, make sure to put the switch next to the Audio plug to "speaker" position, you will probably get much more signal (i.e. better signal-to-noise ratio).

Simon

[mattbassg]

> I am actually quite surprised by your results I must say? Could you tell us a bit more how your device is connected and to what, and in particular where are the grounds connected?

I agree with you - in all my testing of the Midronome at home it has been completely noise free. However in my current 'live' playing environment I am encountering some noise*


Setup
The use-case I'm describing doesn't involve connecting the Midronome (or any devices being clocked by it) to a computer or recording interface, nor even (yet) clocking any external devices over MIDI - I'm simply using Midronome to provide a rock solid AUDIO clock through a FOH PA system for musicians to hear a tempo 'click' through their ine-ear / foldback monitors.
(It is my intention to add MIDI clocked devices to this live setup in the future; the whole reason I wanted the Midronome was to link audio 'click' with MIDI clock so what I'm describing as my current use case is the first stage in that process).
In my opinion the hum noise is quite low in volume relative to 'click' itself - but some people like to have the click very loud in their in-ears so are hearing it.


On the power side

Code: Select all

Normally:
USB 'wall-wart' -> supplied USB cable -> Midronome

'Hum free' workaround:
USB 'wall-wart' -> USB ground lift adapter -> supplied USB cable -> Midronome

I may try some other USB power adapters to see if the one I'm using is inherently noiser than others
I have the Midronome set to 'speaker' output.


On the audio side

Code: Select all

Midronome audio out -> TS to TS cable -> DI box -> XLR to XLR ->Desk

I'm using a high quality TS to TS (mono jack to jack) cable. The DI box is runs on 'phantom power' from the Desk.

Output switch is set to speaker and the level on the Midronome was set to 8 (I note from your ealier posts that I should increase that to 9 which I will try, though I imagine this will increase the difference between the audio and the noise floor, not remove the noise entirely.

Toggling the 'ground lift' on the DI box makes no difference.
I may try using a passive DI to see if that helps

As mentioned in my previous posts I've tried:

Code: Select all

Midronome -> TRS to XLR cable -> Phantom power blocker -> XLR to XLR ->Desk

and

Midronome audio out -> TS to TS cable -> Hum Eliminator -> TS to TS cable -> DI box -> XLR to XLR ->Desk

both of these significantly reduced, but did not entirely remove, the hum.


Ground
So basically other than on the USB power supply there are no ground-lifts in place.


*Comment
I already *love* Midronome - so far it's lived up to (or exceeded) all my expectations and is everything I hoped it would be - I can't wait to move on to using it to help live musicians stay in time with MIDI clocked instruments - which is the goal of my journey with it.

I recognise and accept the possibility that my experiences may have everything to do with the idiosyncrasies of where I'm using the Midronome rather than anything inherent in the device itself; my intent in sharing is nothing more than to add to the pool of 'in live use' knowledge that might help anyone else who encounters similar things.

[beatbastelbaer]

Hello Simon and Midronomists,
lets get things sorted.
We have two different scenarios here.
1 USB ground loops when the Midronome is connected via USB to a PC/Laptop/DAW
this one can be eliminated by separating the USB power and data as I wrote above
2 Power supply caused ground loops coming from an external power supply
that one comes from poorely designed switch mode power supplies and can be eliminated by these magic ground loop isolators.
I have a couple of wall mount PS here, some of them cause more or less hum, others don´t hum at all.
If you power up the Midronome with a powerbank, the noise should be gone, but we need a powerbank with <always on mode> because the power consumption of the device is so small that most powerbanks switch off after a short while.

This problem is not Midronome specific, you can find tons of complaints about this behaviour with many devices.
Interesting points:
some devices do not produce any noise, I assume there is a significant difference in the internal circuit design with common ground.
Arturia adds these splitters to their devices because they had the exactly same issue.
I am wondering why nobody produces something similar, you could sells millions of these.

Frank

[Simon]

Thanks Matt for the feedback and the kind words! Happy to hear the device works for you

Output switch is set to speaker and the level on the Midronome was set to 8 (I note from your ealier posts that I should increase that to 9 which I will try, though I imagine this will increase the difference between the audio and the noise floor, not remove the noise entirely.

Correct.

• breaking the ground on the XLR connection to the desk (toggling the DIs ground lift switch) made no difference to the noise. It also occurs to me to try using a passive DI rather than one requiring phantom power.

This is probably the most surprising of the results, a ground lift should definitely solve the problem.

The overall problem is:

1. The display generates noise
2. That noise is finding a way back to the power supply, which should only be on the black wire inside the USB cable
3. But it can find another path, for example when powering the device from a computer (=power supply is the computer), it can go via ground on audio cable, and then back to the computer
4. And noise on ground + unbalanced cable = noise on audio

But in your case, if you have a ground lift at the DI, then there should be no other path for the ground than the USB cable... Are you sure that ground lift works? Or maybe then the noise comes from something else (if it is the display noise, it will sound like a buzz and its tone will change when the display changes) or from somewhere else (anything else connected to the Midronome, in particular on the ANLG plug?).

Just to verify this, I just tested with the following setup:

Code: Select all

Midronome AUDIO plug -> TS cable -> DI box -> XLR cable -> Audio interface -> computer
Computer -> USB cable -> Midronome USB plug

The DI box I use is a Radial StageBug SB-1. Without the ground lift the buzz is very loud, but engaging it make a huge difference. And I know for a fact that this DI does not use a "real" ground lift (for safety reasons), so even with the lift, the ground is still connected via a 100 ohm resistor, which is why the buzz is still hearable at very high volume. With a full ground lift it should be gone completely.

Anyways - it sounds like you solved the problem, so no worries, I fully understand if you do not want to spend time investigating it anymore

[mattbassg]

My situation definitely falls into the second of the two scenarios helpfully summarised by @beatbastelbaer; Power supply caused ground loops coming from an external power supply.

In my earlier reply I said, "I may try using a passive DI to see if that helps"
I used my own Radial SB2 and Radial JDI di boxes to test my situation

Code: Select all

Midronome AUDIO plug -> TS cable -> DI box -> XLR cable -> mixer -> amplifier

Noise was present in both cases, toggling ground lift made no difference. This is consistent with my previous experience (and a solution in the form of a 'magic ground loop isolator')

I also said earlier "I may try some other USB power adapters to see if the one I'm using is inherently noiser than others"
My findings are totally consistent with @beatbastelbaer's conclusion; I tried 7 different power supplies and found huge differences in 'the hum' from one to another including 2 that were almost completely silent. There is still some residual noise but this is extremely low level, consistent with your description of 'display' noise, and very unlikely to become an issue in my use-case.

Thank you (both) for your engagement, wisdom and experience.

[mattbassg]

Just to add to the previous I thought it might be helpful to start a list of 'suitably quiet' (i.e. isolated) power supplies in case anyone else is using Midronome the same way as me; live and connected to a PA system, but not connected to a USB host or computer, so running on power from a USB 'charge / wall wart' power supply.

So far the best ones I've found have been a couple of old iPad / iPhone charges, an Apple A1357 and also an A1401

Every 'noisy' power supply I've tested has been made quiet by adding one of the ADUM3160 based USB Isolators in line between the power supply and the USB A->B cable.

[Simon]

Good point!
And Matt, if you can, I would really like to "hear" how the noise sounds like when you use the problematic power supplies, if you don't mind recording it?
Cheers
Simon

EDIT: just FYI Matt and I have been talking over private message, what he experiences is not related to this noise issue (which makes sense since his device is not connected to a computer), but rather to general noise floor issues when raising (a lot) the volume of the Midronome's output.