Carl - Ironic you should ask about a CAT/Krell match. I have not heard the "Ultimate" version of the CAT but I am acquainted with numerous previous versions of the CAT preamp going back 10-12 years or so. I've always loved the CAT preamps. Unlike the 50SE they do have that warm, romantic and seductive sound that is so musical and enjoyable. Anyway I was "hot" to but a CAT preamp about 2 years ago. At that time I was using an Audio Research LS 5 Mark II preamp with Krell Audio Standard-2 monoblocks. I wanted to change preamps because the LS 5 was just so restrictive. Anyway I did my homework/due dilegence back then when I considered switching from the LS 5 to a CAT preamp. The problem, as you indicated in your posting is that the CAT only has RCA inputs and outputs. At that time I considered the CAT I re-confirmed my comparative analysis of the output signal from the DACs that I owned which included both the Sonic SFD2 Mark II and the Theta Pro Gen V. Both the Sonic and the Theta sounded considerably better when the signal was fed from their balanced as opposed to their unbalanced outputs. I also called Krell to inquire as to the degredation I could expect if I used an RCA to XLR adaptor in hooking up the CAT to the Krell AS-2s. The explanation was highly technical and way above my head. However, my somewhat simplistic understanding of the matter is that when you feed an unbalanced signal into a balanced component (without the use of an active device such as a phase splitter,) only the positive side of the balanced component's circuitry will carry the signal in the balanced component. Accordingly, when the balanced component goes through its Common Mode Rejection Ratio "gig" - there is no signal coming through the negative side for the balanced component to balance. I think this adds noise and degredation to the signal. Finally, I have also used relatively long interconnecting cable runs from my preamp to amp. These runs have varied from 16 to 24 feet. I have been advised that balanced signals are better (i.e. the lack of noise) becomes an issue when the cable run begins to exceed 10 feet. For the reasons stated above I decided to pass on the CAT preamp. If your system is primarily digital and if your preamp needs to drive long runs of interconnecting cable and if you are feeding the signal into a fully balanced component such as the Krell 600 amp - I would be reluctant to use an unbalanced preamp. Beleive me I've tried to rationalize it myself numerous times in the past. I love the CAT preamp (I also love the Conrad Johnson Premier 8 monoblocks). But they are wrong for my digital based system that uses long runs of interconnecting cables and balanced power amps. I think that is why so many of the CAT lovers are analog guys. Digital sounds so much better in balanced mode. I don't think the same is true of phono. Anyway, if you haven't already done so, I would compare balanced and the unbalanced outputs of my digital front end and see if there is a difference. I would also call Krell and see if ypou understand what they are saying about using an adaptor for the input to your amp. Oh yeah - one last point. In order to get around the balanced is better for long cable runs idea I also considered going against "conventional wisdom" and begin using short runs of interconnecting cables with a relatively long run of speaker wire. I thought this idea might work for Transparent, MIT or any other network type speaker wire. It seemed to me logical that if the signal was going through an active device such as a network box, then the network box would "correct" the degredation from a long speaker wire run in the same manner in "corrected" the signal from a short speaker wire run and yield no difference. I called Transparent about this and got hold of Karen Sumner. Karen told me that in listening tests they had at Transparent, their test revealed no "audible" difference in the length of their speaker wire up to runs of at least 30 feet. I then called Krell and asked their technical people what they thought. In essence they gave the classic reponse that the longer the wire the greater the impedance and their "agin" the long speaker wire idea - with or without network boxes. Hope this helps a little.
