As everyone (including yourself) already said in their suggestions for things to try: "Process of elimination" is the way to figure out where the issue is coming from.
So, if the noise happened when using your "typical" signal path, minimize the signal path and see if it still happens. If not, add stuff back into the signal path that you are normally using, and test when/if the issue comes back.
Time (e.g. for thermal issues) is also an important factor, of course. Just for the sake of not having to deal with a variable that is difficult to control (e.g. how long exactly until heat might introduce the issue), I'd do my first set of tests with a "warm" console that's been running for a few hours. If I can't reproduce the issue then, I'd try again with a cold system, and try to repeat the tests quickly while it hasn't reached full operating temperature, yet.
And I agree with the various ways that were suggested to eliminate portions of the signal chain, to see if the issue is still there, without that component or path. Such as, trying a different channel, "bypassing" the D8B by recording directly to the HDR from a different ADAT source... or bypassing the HDR, by recording from the D8B to a different recording medium (e.g. a DAW or just a DAT recorder, for example).
If you had any other analog gear in the chain (e.g. I almost always add a touch of compression to vocals on the way in, before hitting the converter... so, IF I were to use the D8B pres, I'd almost certainly have a comp inserted... and with that unbalanced connection, that could introduce interference like that noise... but if your signal chain is fully digital, then this should mean that you do NOT have anything on the insert, right?).
Another issue that I had a number of times with DAW computers in the 90's with different soundcards, was that they would pick up interference from inside the computer case. This could be a high pitch wining that changes either with CPU or graphics activity, such as mouse movement.
Clearly, the D8B doesn't suffer from that sort of issue if everything is working right... but I wonder if anything could either have gone bad that could cause that... or maybe it it could be something like not having re-installed some sort of shield or metal cover/barrier back into the D8B or the rack unit when having them apart, and now interference is impacting a component, that was previously better shielded?
Or sometimes (I think), hardware devices with an issue can introduce interference via ground. So, IMO for the sake of troubleshooting, it would also be worthwhile to disconnect all outboard gear that is usually "permanently" connected, and check if the noise still happens.
Anyway... just trying to throw out some thoughts.
Part of the reason why I have more than one D8B, and why I keep more than one in working order, is so I can replace the whole unit, and check if a problem I'm observing, is also present with another unit. So, if that is an option (and IMO it is, since these desks sometimes pop up pretty cheap on craigslist, etc.), I'd keep a spare around, just for troubleshooting and eventually, for parts.
Not to sound like a hillbilly, but also used to have 3 cars of the same model, for the exact same reason why I keep 3 D8Bs around, hahaha. It can help take some guesswork out, if you have an option to compare between units