Amek Angela - 8 kHz Noise From Master Channel

[smilan]

Hi, I'm looking at the working noise of my Blue Amek Angela (all the channels are muted by the "master mute button on the master channel) and I have this 8K noise equally on both L and R channels.
With the master fader at unity gain the peeks are at -80dbfs and with the master fader all the way down I can still see this 8, 16 and 24 kHz peeks, but under -90dbfs.
Does it a mutter of "dirty mains power"? or modern (and quieter) OP amps will give me a better result?
I've recapped the entire desk 3 years ago.

 

[smilan]

It seems like the noise is coming from the USB cable that runs from my PS to the Motu 16A. With a different cable the noise reduced from -70dbfs to -85dbfs. Do you think that changing the PSU will reduce the noise even more?

 

[ccaudle]

USB cable that runs from my PS to the Motu 16A

The 16A has an AC power input, why is there a USB power supply involved?

 

[smilan]

The 16A has an AC power input, why is there a USB power supply involved?

That make sense, can I just chop the red and black wires of the USB cable?

 

[ccaudle]

can I just chop the red and black wires of the USB cable?

I believe device detection will not work properly if you cut out the power cables. There may be a way to make that work, I think user Tubetec did some pretty extensive surgery on the power supply of a USB audio interface he had.

Earlier you wrote "the USB cable that runs from my PS to the Motu 16A." I assumed PS meant power supply, are you now saying that it is not explicitly a USB power supply used to power the MOTU, but is actually a USB data cable? Was that a typo, and you meant to write from your PC to the MOTU?

In that case you would need to verify that it is actually power supply noise, and not leakage of data signals into the analog audio causing the problem. Power supply noise seems more likely.

Do you have the AC power cord installed already? That interface seems to large to run from just USB power. It is possible that the USB power only powers the USB interface, and the AC power powers all the audio circuitry, in which case disconnecting the USB power would prevent the interface from being used via USB. Just a wild guess, I am not saying that is likely, but you would need to dig up (or reverse engineer) more information about the design before doing anything too non-standard.

Is the noise present on the MOTU without it being connected to the AMEK?

 

[smilan]

Earlier you wrote "the USB cable that runs from my PS to the Motu 16A." I assumed PS meant power supply, are you now saying that it is not explicitly a USB power supply used to power the MOTU, but is actually a USB data cable? Was that a typo, and you meant to write from your PC to the MOTU?

Sorry for this typo. the USB cable runs from my PC to the Moto.

Do you have the AC power cord installed already? That interface seems to large to run from just USB power. It is possible that the USB power only powers the USB interface, and the AC power powers all the audio circuitry, in which case disconnecting the USB power would prevent the interface from being used via USB. Just a wild guess, I am not saying that is likely, but you would need to dig up (or reverse engineer) more information about the design before doing anything too non-standard.

It sounds right, The Motu has its own PSU that converts the mains power to DC to operate the audio circuitry and the USB power only powers the USB interface.
No, this 8kHz frequency can be heard only when the Moto inputs are connected to the Amek outputs and only when Cubase working (even with no active project), the 8kHz starts at some point when Cubase boots up.
When I changed the sample rate from 88K to 44K the noise reduced from -85dbfs to below -90dbfs.
Does it means that the noise caused by a leakage of data signals into the analog audio?

 

[ccaudle]

Does it means that the noise caused by a leakage of data signals into the analog audio?

Not enough information to say. It could be related to the data signal since the noise only starts once Cubase starts.

What is the shield configuration of the audio cables, connected on both ends? That is usually best for high frequency noise, or connect the shield only at the device output if you have power frequency noise (50Hz or 60Hz) and want to experiment with one-end-only shielding.

 

[scott2000]

What is the shield configuration of the audio cables, connected on both ends?

Was wondering about maybe how any balanced and unbalanced connections might be lurking too. And how the motu wants to handle those.... tying pins or not..etc....

Is this a new issue ?

 

[smilan]

What is the shield configuration of the audio cables, connected on both ends? That is usually best for high frequency noise, or connect the shield only at the device output if you have power frequency noise (50Hz or 60Hz) and want to experiment with one-end-only shielding.

The audio cables are shielded on both sides. I tried now to connect it with a one-end-only shielding connection with no change in the noise level.

Was wondering about maybe how any balanced and unbalanced connections might be lurking too. And how the motu wants to handle those.... tying pins or not..etc....

The Angela outputs are all unbalanced. I tried to get out from the Angela to the Moto from both TT patch panel and from the XLR output at the bottom of the desk with the same noise level.
I've also tried to connect the audio cable with the ring and screen shorted on the Motu input with no change in the noise level.

Is this a new issue ?

I replaced my old PC few months ago, I'm pretty sure this issue started after the replacement.

Some other test Imade:
I connected the USB cable to an external 5V PSU with no change in the noise level.

When I connected the Motu to my laptop with the same USB cable the noise reduced significantly.
Does it mean that the noise does not com from the PC PSU, but something else in the PC creating this noise?

 

[ccaudle]

The Angela outputs are all unbalanced.

If they are truly unbalanced then it may be that common mode noise from the PC is showing up on the output. That is the disadvantage of unbalanced connections, the common mode connection point is one of the signal conductors as well.

 

[smilan]

If they are truly unbalanced then it may be that common mode noise from the PC is showing up on the output. That is the disadvantage of unbalanced connections, the common mode connection point is one of the signal conductors as well.

The question is why this noise appears with the desktop computer and not with the laptop? There's a way to find out if the noise is coming from the PSU or something else in the desktop producing this noise?
So far I tried to disable all the plugins, removing the UAD pci card, disconnect all the hard drives except for the system drive, all of those actions did not made any change in the noise level.

 

[Khron]

The question is why this noise appears with the desktop computer and not with the laptop?

Some funky earth-loop going on there. Even if the laptop is connected to the charger, and that is earthed, those are usually earthed via a 1meg resistor (as opposed to a "zero ohm" connection). Your desktop PC is definitely zero-ohm-earthed.

 

[ccaudle]

The PC is probably a lot higher power as well, which has an effect on the magnitude of chassis leakage current (if that is the cause).

The Angela outputs are all unbalanced.

and from the XLR output at the bottom of the desk with the same noise level

The XLR outputs are unbalanced? What does pin 3 connect to on the Angela male XLR connector? That schematic in the first post doesn't show the connector configuration.

 

[smilan]

Some funky earth-loop going on there. Even if the laptop is connected to the charger, and that is earthed, those are usually earthed via a 1meg resistor (as opposed to a "zero ohm" connection). Your desktop PC is definitely zero-ohm-earthed.

The laptop was connected to the charger during my test.
There's a way to convert the desktop PC PSU to "1 meg" earth connection?

The XLR outputs are unbalanced? What does pin 3 connect to on the Angela male XLR connector? That schematic in the first post doesn't show the connector configuration.

As the manual implays he XLR outputs are unbalanced, pin 1 and 3 are shorted together

 

[mbgrace]

Sounds like computer power supply leaking noise in to the grounds and imposing noise voltage across the audio cables from the console to the MOTU. You can try isolating the ground on the computer power cord (remove ground pin) as an experiment but not recommended as a permanent solution.
You could also try to wire the unbalanced output of the console to the balanced in of the MOTU with the ground connection to the inverting input of the MOTU. That could give you some common mode rejection and break the ground loop.
Ultimately a modification to provide balanced output from the console would be best. Either active or with a high quality transformer.
Soundcraft used to use a pseudo balanced output with a single op amp to help with this. Might be something you could retrofit. Basically the "negative" output was connected to a positive feedback node on the output amp. This would attempt to add ground noise to the output amplifier to make it common mode.

 

[mbgrace]

Here is the circuit mentioned above.

 

[Khron]

Soundcraft used to use a pseudo balanced output with a single op amp to help with this. Might be something you could retrofit.

That sounds complicated, and for however many outputs, may end up adding quite the extra load on the power supply.

Why not just go with impedance-balanced outputs?

https://www.modwiggler.com/forum/vi...sid=895c7283631abde88829e98e1cc9d9d4#p1085128
No need to add any active circuitry, only another resistor and capacitor as are already on the outputs, by design.

 

[ccaudle]

You could also try to wire the unbalanced output of the console to the balanced in of the MOTU with the ground connection to the inverting input of the MOTU.

I think that should be the configuration when using the XLR outputs:

XLR outputs are unbalanced, pin 1 and 3 are shorted together

With ground potential differences on unbalanced equipment you can reduce the voltage between the chassis and directly reduce the voltage in the audio.
The post with the noise FFT did not specify the gain, but if that conforms to dBu or something close to it, -80dBu is less than 0.08mV; just under 78 microvolts.
Not very high, so maybe a few copper braid straps between the Amek and the MOTU could get it down into the -95dbu range.

Maybe that in conjunction with making the outputs impedance balanced can get it even lower, although 8kHz gets into the range where the balance of most differential inputs is beginning to fall off so the CMRR will not be nearly as good as it is at the below 1kHz range.

 

[smilan]

Thank you for your answers, I made a test with the electronic balanced circuit (with 330R resistors for R1 and R2). With the electronic balanced the 8K peak reduced a little, but the 16K peak and all the above noise increased compared to the unbalanced connection.

 

[Khron]

330R is arguably a bit high of a value; no capacitors though?

Also, with the impedance-balanced connection, you have the option of disconnecting the ground wire from either end of the cable (check which end disconnected gives you the least amount of noise).