Not the same engine, but I had a similar sound experience with my new Bush 4141 engine when I was testing it recently. My experience didn't seem to have to do with temperature though, and I recognized I had a particularly crappy DCS signal for whatever reason(probably 2 to maybe 4). (regular loop of tubular, with no gap insulating pin anywhere on the center pin, so I am asking for trouble with multi path going on with the DCS signal - though I've used it like this before with relative success (few missed commands/not on track once in a while) and never heard any odd garbling like this)
I was also simultaneously using Legacy to manipulate the power the ZW-L I was using to feed power to the track. Not 100% sure this specifically hurt things, but what I did next seemed to help, so I mention it (just in case you are using a ZW-L or maybe other TMCC track control items simultaneously).
I took the ZW-L out of command mode and dialed up the voltage manually from the handle and also added a lighted lock on the the track (I know it's not as good as doing proper block isolation to avoid the multi path, but I figured it was worth a shot).
After I did that, I was able to have a decent length operating session going for a bunch of laps around the track and activating all the special sounds in the engine, and I don't think I had a recurrence of the garbling. (Maybe I did when I entered the track signal test mode, but it made it through normal operation for a while).
Might not be relevant to your specific situation (I would assume your layout setup is way more "DCS friendly" than my floor loop is), but the random garbling of sounds was very close to what you show here.