lawajava
I had an issue once of this nature. It nagged at me and made me mad.

Somewhere along the experimentation route to quell it I put a $0.50 three to two prong grounding adapter on one of my power lines. It cleared immediately. Years later that little charmer still works.

Problem solved in my case for 50 cents.

If the ground line is "dirty," that can cause issues because those sensitive audio circuits connect to ground eventually. I reviewed an Equi-Tech balanced power unit (not cheap), and there was a measurable change in noise level at a mixer output when it was powered from the balanced power supply. I'm not recommending that as a solution necessarily, just confirming what lawajava points out about grounding issues causing potential problems, and that filtering can make a difference.
 
However, I think this kind of noise would tend to be more constant as opposed to the "bursts" you describe. As mentioned in a previous post, it could also be RF from a cell phone, iPad, etc.
 
Lifting the ground as lawajava suggests may provide a fix, HOWEVER that ground wire is there to protect you. If there are AC problems, fix them at the source.
 
Finally, I've had better luck eliminating ground loops by adding an extra ground rather than removing a ground. For example there's a grounding screw on the back of the Roland Octa-Capture...when I connected that to the metal screw that affixes the plastic plate to an outlet (that screw connects to ground), various noises, hums, and whines went away instantly. That's because it now had a low-resistance path to ground as opposed to traveling across various audio cable grounds. Remember Ohm's Law: a voltage drop can occur across a resistance...and there's your signal.

Wow! First off, many thanks to all who have contributed, really appreciated.
 
To clear up a few suggestions. It's not residual demo white noise, it's too low level for that although the noise does have a "white noise" sound to it.
The noise bursts are "regular" i.e.about 2 seconds apart and each lasts for about 500mS which would seem to indicate that they are being "clocked" somehow - the PC perhaps?
The noise is recorded if I open a clean track with no other audio - the noise is too low to be heard on a track with audio on it unless there is a silent period in the track.
I can also hear noise while moving the mouse - only while it is moving.
To try and minimise potential external sources, only the A/D is connected to the DAW and I am monitoring the noise directly on the headphone output of the Sapphire.
 
All the points about earth/ground  loops are noted and I have a made up a socket box with a switch to remove the power earth connections individually. All my outboard expanders are powered on of course so maybe I need to remove the power from them all.
 
Another point I had overlooked is that the DAW and monitor run from a UPS, and obviously I need to remove that from the equation as well.
 
I do have a scope to look at the noise waveform. I can also record a few seconds of the noise, but not quite sure how to put that into the forum???
 
Again many many thanks to you all for your suggestions. Over the holiday period I will have a bit more "exclusive" time to investigate and report back.
Happy new year to you all too!!
 
 
 
 

Has a di box been suggested yet? It might help with hums or other noise. Here is some hopefully helpful info.
http://www.native-instrum...ng-and-buzzing-sounds/

You are connected to the UPS via USB for clean shutdown when the battery runs out..?

Hmmm

Not a solution, but potential work around in the interim. Is the noise level low enough to be gated out of the signal effectively?

Have you checked background processes running on the machine? Something that cyclical could be in the box, or possibly the UPS... Sounds similar to a capacitor charging and then bleeding off, so the cycle could relate to voltage as well as a clock.

If you take the computer out of the equation, do you still hear the noise?

A few more possible sources of "clocked" noise -

Hard drive motors.
PC cooling fan motors.
Coils in the power supply or on the motherboard or graphics card.
EMI emmissions from LCD screens (not as bad as the old cathode ray tube monitors though).

I've also found the situation where there was something transmitting radio interference that was picked up by guitars as a whine whenever the audio interface was opened for recording by Windows/Sonar/Audition/anything. Whether the interface used was USB or PCI made no difference, nor did whether the driver was ASIO, WDM or WASAPI. The whine changed in pitch and volume depending on the sample rate and buffer settings. Never did get to the bottom of that, the only solution was to move the PC further away from guitars.

It is also possible to get an earth loop between a USB device and the PC, though if that happens it's usually a steady hum or whine.

Use DPCLatency checker and Latmon Latency monitor.
If there is a process that is causing it, you should be able to get an idea via those programs, it may well show up as spikes in the DPC prog.

Other than that there could be a dodgy component on your motherboard or in the psu.

Corrected:::: I figger this does happen with Sonar not open. Right?  You can just listen and hear it regularly?
 
If so, sounds like an input into the system.
 
Does it happen without the audio interface connected and Sonar closed?
 
Test: enable the on-board audio device, connect to speakers. Can you hear it?
 
Got windows sounds disabled?
 
For what prices will you be selling the "old" audio interfaces?

FWIW, At one point it was useful for me to create a new user profile. My old user profile ended up slightly corrupted with the DRM stuff. IDK what did it, by the guy from Syntorial.com suggested it as a fix and it did the trick for me. 
 
I'm meaning at the windows level, I created a new login account.