Scarlett 2i2 oddity

[ruffrecords]

I have been testing some transformer using REW with my Scarlett 2i2 as the input. Because I am testing at line levels I use the Scarlett's TRS input with the input set to line. This should give me a 10K input. However on a balanced floating input this doe not work. The level is about 40dB below what it should be. After a lot of faffing around and head scratching I tried connecting pin3 to pin one on the XLR connected to the transformer output (I use and XLR to TRS cable to theScarlett). Bingo, up pops the level to the expected reading. So, it looks to me that the ring of the TRS is left unconnected. I can imagine it is OK if you plug in an unbalanced synth with a TS plug or even and electronically balanced source but transformer expect both hot and cold inputs even at line level.

Before I take the top off to check, has anyone else seen this issue?

Cheers

Ian

 

[ruffrecords]

Oops. I found out what it was. Dodgy cable.All fixed now.

Cheers

Ian

 

[Tubetec]

I have limited experience with the Scarlet 2I2  ,a buddy of mine owns one ,he does sometimes get this odd fault with his ,where upon plugging the interface in the green led on front blinks fast and the driver wont load up ,usually just re-pluging it fixes the issue or maybe a re-boot with the device attached .

I gave up on cheap xlr cables long ago ,especially the ones without any form of packing inside the cable slieve, the geometry ,and capacitance of these is subject to change , I tend not to bother using anything other than neutrik for connectors as the cable strain relief on the 'elcheapo' xlr's and jacks isnt much good ,especially in live situations .

 

[cyrano]

If the Scarlet 2i2 doesn't start up reliably, get a new/better USB cable or try another USB port on the computer.

USB power is sometimes just sufficient and a thin USB cable can make that worse.

 

[Tubetec]

Yeah very often Usb is dirty power at best ,interesting conclusion you make Cyrano .
I guess a usb Y cable with off board power could be used to make a more reliable  set up .

 

[cyrano]

In theory, USB provides 5V, 500 mA. Not much for two mic preamps, AD/DA, headphone amp and phantom power.

Most of the times, it's the computer and not the interface. These Focusrites are reliable.

 

[Newmarket]

Yeah very often Usb is dirty power at best ,interesting conclusion you make Cyrano .
I guess a usb Y cable with off board power could be used to make a more reliable  set up .

Yes. That's what I found powering a DAC off USB. USB is the only power input available. Lots of 'digital' whirrs and noises coming through - can see it matching hard disk operations etc. I am talking about winding the output up by 60+ dB for testing but still...

I made up a board / box that takes in 5V DC power and puts that onto a USB A connector so simple USB A-B cable from there into DAC. All good then and I was just using a wall wart switcher (originally for a USB Hub) for the 5V.
When using USB as the audio source (it also does SPDIF and TOSLink) I simply have a USB B socket to take the USB from the PC but take only the Data Pins to the USB A output connector.

 

[DerEber]

Cool, such a thing would be helpful in so many Laptop situations!

 

[RuudNL]

Yeah very often Usb is dirty power at best.
I guess a usb Y cable with off board power could be used to make a more reliable  set up .

Yes, true! I had this problem before witn an external USB audio interface.
I could 'hear' the mouse moving and the harddisk...
I made a split cable, data lines and ground to the PC, +5V and ground to an external power supply.
Now everything is completely silent!

 

[Tubetec]

Hi Guys,

Thanks for the thumbs up on that, I think most good devices should have opto isolators on the data ,so once you break the ground and power to the laptop or computer ,your good galvanically as well .

I remember down through the years ,while testing out various audio recording packages ,hearing that exact same digital mush ,mouse causing audio interference etc ,and I simply refuse to use it ,no way am I commiting to tape a performers precious efforts  with that crap going on in the backround  . My  solution was old fashioned optical audio toslink  from a small external usb sound module .
Its manditory to  seperate power and ground from the computer and your audio interface if you have analog connections  in my opinion  ,otherwise spitty splatty digital dross in the backround .Maybe a battery 5 volt supply for the interface would be cool, Y usb cabled in ,maybe eliminate screen connection to computer completely, only connecting at one end  ,and of course no chance of any of the noise associated with switcher psu's in your audio rails ,actually thats probably wrong ,as dc-dc will be used inside the device to provide proper op amp volts ,at least youd hope it was ,still one less source of noise than switchmode if you use battery as well as the inherently groundless nature of the battery , a powerbank like used for remote off mains phone charge with four AA Nimh cells could work  ,but I dont even see the need for a regulator/dc-dc on the battery output  ,4x1.2 volts say 2200mah ,might be a days  running time (not sure of mA draw of 2i2),be a simple fool proof solution ,unless of course the batteries go ballistic and pierces your heart while charging  ,surely a modern charger can have a temp probe on each cell to monitor in case of thermal runaway  ,maybe even a polycarbonate lid to stop projectiles in worst case scenarios. 8)

 

[Newmarket]

Yes. That's what I found powering a DAC off USB. USB is the only power input available. Lots of 'digital' whirrs and noises coming through - can see it matching hard disk operations etc. I am talking about winding the output up by 60+ dB for testing but still...

I made up a board / box that takes in 5V DC power and puts that onto a USB A connector so simple USB A-B cable from there into DAC. All good then and I was just using a wall wart switcher (originally for a USB Hub) for the 5V.
When using USB as the audio source (it also does SPDIF and TOSLink) I simply have a USB B socket to take the USB from the PC but take only the Data Pins to the USB A output connector.

Replying to my own post as I got a few things incorrect.
Going back to it on the weekend I remembered that I still suffered from the 'digital' computer noise if I connected only the USB 0V line (and not the +5). I guess you can try not connecting the 0V but there's a danger that the Common Mode Voltage of the USB D+ and D- signals might get away from the Common Mode range of the DAC input. ? Maybe not due to pull up / down terminations but not sure  I want to chance that. Maybe parallel diodes between PC USB 0V and DAC 0V to limit voltage difference to a diode drop worst case ?
Guess could try connecting DAC 0V to a 'clean' Chassis ground point (PC is earthed).

But I did get clean audio from the DAC - remembering how - I used my DSP based soundcard (Soundscape / Sydec Mixtreme 192 (now obsolete and not really useable post XP  with the SPDIF option. SPDIF set to output at 96kHz. SPDIF into DAC. Not sure if SPDIF DAC is galvanicaly isolated from the PC ground but all seemed good.

 

[Newmarket]

Thanks for the thumbs up on that, I think most good devices should have opto isolators on the data ,so once you break the ground and power to the laptop or computer ,your good galvanically as well .

With USB it's not so simple to implement opto isolators due to the bidirectional nature of it and the protocol involved. Although there are commercial solutions available. Not sure about USB 3.0.
I do have somewhere  a 'digital' isolator IC from the the Analog Devices ADUM range specifically for USB isolation but not done anything with that as of yet.

 

[Tubetec]

I did a couple of tests with a transit USB interface ,
I put together a splitter cable, power +/- to small two cell power bank , green and white wires from the device into a usb plug to host , it wouldn't initialise the device until the 0V black wire was connected from the host to battery - ,but it still worked fine with sheild disconnected .In the case of the MaudioTransit output no difference in noise was found on either battery or host power .Line/mic Inputs could be a different story in that regard.

I was thinking an easy quick and cheap solution would be a USB active extender with isolated data outputs and seperate input jack for 5 volts , you could locate the interface a distance from the pc without trouble if needs be and power it off any supply without interactions such as mouse fuzz or Hd activity related chirps. Alternatively I think Ive seen someone, in the case of a signal generator who used a small pc board to add an optocouple for the very purpose of  prevention of contamination of psu currents from pc to 0V on the device via usb ground .

https://www.circuitsathome.com/measurements/usb-isolator/

 

[Newmarket]

I did a couple of tests with a transit USB interface ,
I put together a splitter cable, power +/- to small two cell power bank , green and white wires from the device into a usb plug to host , it wouldn't initialise the device until the 0V black wire was connected from the host to battery - ,but it still worked fine with sheild disconnected .In the case of the MaudioTransit output no difference in noise was found on either battery or host power .Line/mic Inputs could be a different story in that regard.

I was thinking an easy quick and cheap solution would be a USB active extender with isolated data outputs and seperate input jack for 5 volts , you could locate the interface a distance from the pc without trouble if needs be and power it off any supply without interactions such as mouse fuzz or Hd activity related chirps. Alternatively I think Ive seen someone, in the case of a signal generator who used a small pc board to add an optocouple for the very purpose of  prevention of contamination of psu currents from pc to 0V on the device via usb ground .

https://www.circuitsathome.com/measurements/usb-isolator/

The ADuM4160  iso;ator ic used in the linked article is what I have I think. I use ADUM parts in my commercial designs but not used the USB application device yet.