I have posted similar material previously but given the recurring nature of the problem and that some of the advice I've seen posted exposes owners of these devices to both serious physical and financial risk, I'd like to reiterate the points I've made previously.
 
NEVER EVER use ground-lifting adaptors or disconnect ground wires on the AC mains side. To do so exposes you to two risks. The first is simply that in disconnecting the protective ground you are disabling a vital safety measure intended to ensure that if a power adaptor etc. fails, then any live mains voltage that makes its way through to the equipment chassis will effectively be shorted to ground and blow a protective fuse or trip a circuit breaker before any harm is done elsewhere.
 
If you have never had the experience of being across either 110VAC or (elsewhere in the world, up to 240VAC) then I can assure you from personal experience that it is at best excruciatingly painful. Should you be unlucky enough to trigger ventricular fibrillation, then you will also have several seconds of remaining consciousness during which you will be aware that you have just killed yourself. (there are plenty of reports of such experiences, narrated obviously by onlookers). You may also be surprised to learn that much smaller voltages can be fatal if the contact points are close to the heart and/or the skin is moist. One researcher died after applying only 12V to his body during a research project.
 
The financial risks are twofold. Firstly that an electrical failure is more likely to trigger a fire if the ground path is interrupted, and secondly that when grounds are lifted, equipment tends to 'float' and there is a substantial voltage between chassis and ground. This voltage exists because of capacitive  leakage paths within power adaptors (particularly via transformers, which have capacitance between the primary and secondary windings). This voltage is associated with a high impedance so it isn't immediately dangerous - you'll probably have experienced this when touching an improperly grounded device; the surface feels like it's vibrating when you run your finger across it.
 
But this voltage, while associated only with a very small current, poses a serious danger to your equipment. If you connect or disconnect signal cables associated with an improperly grounded device, it's quite possible for a hundred volts or more to appear at sensitive preamp inputs or outputs. Since inputs, in particular, have quite high impedances, this voltage can go on to destroy semiconductor devices or degrade them so that they either fail prematurely or exhibit seriously reduced gain or noise performance. I've seen all manner of equipment damaged or destroyed this way, including laptops where a projector was plugged in to the VGA port and destroyed the laptop.
 
So, what to do?. Well, the basic problem with bus-powered devices is that an already noisy power source (taken from the computer) is being fed into a digital device whose power consumption is also 'spiky' as various digital devices inside the audio interface switch on and off.
 
Although the USB cable ground connection is fairly substantial, due to the high currents flowing in the cable (up to half an amp), a voltage will be induced between one end of the cable and the other. Although the spikes and noise associated with this fluctuating voltage are small, a few millivolts at most, this is of course the kind of sensitivity that mic preamps are designed for. So when another piece of equipment is plugged into the interface, it's quite possible for this voltage to reappear across the preamp input, leading to hum, buzzes, and various noises, which not uncommonly change as other equipment such as a mouse is moved around. (this is the contribution from the already noisy power supply inside the computer).
 
When this happens, the safest mitigation is to connect a thick cable - and by this I mean something with a conductor size of 3mm or thereabouts, between the audio interface and an appropriate ground reference, which might be the ground on the mixing desk it's plugged into, or the computer, or possibly a ground directly at the mains plug itself (you make a cable with just a ground wire that plugs into the electrical outlet). The best kind of cable I've found is the flexible insulated copper braid cable sold to wire up car audio systems - it can be obtained in nice solid thicknesses but is still flexible and easy to deal with.
 
What you have done here is to give that noise current a much lower impedance path than the USB cable. This should substantially reduce the induced noise.
 
You may also - if the interface supports balanced inputs/outputs - use special cables for the ins and outs where the ground is lifted at one end. This is perfectly safe since you are only dealing here with audio level signals and there are no safety implications.
 
To be honest however, none of these mitigations works all of the time. Bus powering an audio interface is a fundamentally flawed idea. Cut your losses and purchase an interface that uses an external power adaptor, such as the FocusRite Scarlett 6i6 etc. Note that some of the smaller FocusRite units ARE bus powered. The Roland quad capture is bus powered, hence the prevalence of posts regarding noise issues. It's just a plain stupid idea and although it works in a lot of cases, it's just guaranteed to cause problems in other cases.
 
 
 
 
 
 

I've got a bus powered interface that isn't particularly noisy, hasn't yet caused any life-threatining instances nor blown up my computer in the last few years.
 
Should I be worried?
 
The only time I get noise is when using Hi-Z and I get expected RF interference when the TV is turned on or I'm in a position that is sensitive to the reception of any other kind of RF interference.  It's normally easy enough to find a good spot like you would with any guitar amp which is equally susceptible to RF noise.
 
Conversely I sometimes get access to a factory supplied wall-wart powered interface where the Earth (ground) pin isn't even implemented it's merely a plastic blank for the purpose of allowing the 'live' and 'neutral' pins to be accesed, and it imparts AC hum to the interface which disappears when I run it off a USB hub instead.  In the UK there's a third pin called 'Earth' which means ground and the computer unlike the wall-wart is Earthed (grounded) from the PSU anyway.  It's part of the UK's electrical standard and it exceeds your specification given above for a ground wire anyway.
 
If bus powered interfaces didn't work they simply wouldn't sell.  I do appreciate you are trying to protect us all from danger though.

The reason I would never use a bus powered interface is there's no way your going to have proper Phantom power for condensor mikes. There's just not enough juice in that cheap little supply. Then you use the 2 prong wall warts and you get hum.. No look for an interface with a A/C power cable or next best a gounded/ earthed wall wart. 

The danger from the ungrounded system is more closely related to the electrical system in the building being ungrounded, improperly grounded, or from defects in the wiring such as reversed polarity on the hot and neutral/ground wires.
 
Yes, it is very important NOT to remove a ground intentionally. If having the ground is causing issues, buzzing or hum, it is best to look for and locate the real cause of the problem. There could be reversed polarity hot/neutral wiring, swapped neutral/grounds, loose or bad wiring connections, missing or bad ground connection that is the real culprit. Finding it and resolving it is the correct way to deal with grounding..... not bypassing the system and component grounds.
 
Having said that..... a ground life adapter (several of them) was standard gear in the wire and cable box for every band I played in when we went to do a gig. They were "life savers" in many clubs where the hum in the PA of the guitar amp was unbearable..... and the foam windscreen on the mic kept you from getting zapped. You know we all did it....  However, in a studio, the exact opposite should be the rule of the day. Do it right or don't do it at all.
 
A few weeks back I posted this: http://forum.cakewalk.com/Electrical-Grounding-issues-and-why-they-are-dangerous-m2875220.aspx
 
Take a few minutes and read the background stories I link to for an explanation of why and how this can be dangerous. This link in the OP has several stories that you should read. One is about the issue of faulty wiring that doesn't show up easily using many of the common testers and testing methods and the second story is music related where a studio was set up and the building wiring was faulty, but didn't appear to be faulty. The result was thousands of dollars of gear being damaged when it was connected and operated. Good read for studio owners like the folks here who may set up a studio in a spare room, attic or basement of their home.
 
While a good ground is vital to proper operation of the electrical system and to personal safety, it's simply not true that it's there to protect the gear. The current needed to trip a breaker is hundreds of times greater than the current needed to destroy the circuits in most computers and audio devices. The ground simply is there to allow sufficient current to flow to trip the breaker to remove the voltage from the device to prevent a person from being electrocuted by touching it.
 
The device may have a fuse or CB in it that IS designed to attempt to save the circuit board...... sometimes that even works like it's supposed to.
 
Yes, so called "low voltage" can be dangerous under the wrong conditions. 12vdc will hurt you if you have wet or sweaty hands AND the power source is capable of delivering sufficient current to the load. Voltage is pressure, it is the current that does the dirty work and all it takes is around 60ma across your heart to create problems.  Ground fault breakers and outlets are designed to trip with 7ma of current going to ground through them. You will feel that 7ma. The higher the voltage of the source, the more easily the current will be pushed through the load. But don't let the low voltage supplies fool you.
 
Voltage induced.... is a calculable number. Using ohms law.....E=IxR.   Current times the resistance will give you a voltage..... but you also have to consider the rest of the circuit, including the load, not just that wire.  In any short cable, the resistance and current is low so the voltage drop across the cable will be negligible under most circumstances. Probably not even readable on the average DMM.  I agree 100%.... use the best cables you can get. Larger diameter wire, solid quality solder connections, good quality connectors on the ends. The larger the cable conductor is, the lower it's resistance to current flow and the more efficient it becomes.
 
If an interface or other audio device is supplied with a factory power supply, it is good advice to use that power supply as opposed to using the USB bus power since that is limited in many computers to a maximum per USB port and in some cases in not sufficient to power the device properly. Circuits that operate in a current/voltage deficit mode as a result may have issues including hum and other things that would be problematic in a studio situation.
 
 

i had one plugged into the hub built into my laptop cooling station just to test it out.  how does all that hum show up in the audio through the usb??? lmao!

swamptooth
i had one plugged into the hub built into my laptop cooling station just to test it out.  how does all that hum show up in the audio through the usb??? lmao!

hubs are not recommended for interfaces. Printers and other things like that are OK on a hub, but not the interface or any other audio related things.  the hub can not generally provide the power needed to run the interface and more importantly, the interface is sharing one USB port with everything else on that hub. Bandwidth is critical to the proper operation of the interface, and sharing the port on the hub can severely limit the bandwidth for the interface.
 
I'd have to go back and look through the manuals I have for my devices but I do recall seeing the warning NOT to use the device on a hub..... not that it won't work, but it may not work properly.
 
as to the hum.... and why it may be there. The more you stress a power source, the more ripple tends to appear in the DC source from the power supply. Devices that are engineered to use fairly pure DC with low ripple are not designed with additional filtering in their power input stages so any ripple present can easily pass through to the audio stages where it gets amplified.
 
In my day job business, I use an intercom system that runs on 12vdc. I thought I would try to save the cost of the power supply for that intercom and use one I had in the shop for a 12vdc alarm system.  It worked, but the hum in the system was unbearable. I connected it to a 12vdc battery and it was perfectly quiet. The hum from the cheap alarm power supply had too much ripple in it.... I ended up having to spring the money to get the engineered power supply and it worked perfectly quiet since it was heavily filtered.

You know, these threads are why I really love this place!!
 
Thanks guys.