I have been involved with several large studio construction projects that were designed and built with the spoke-and-hub philosophy, and I've inherited rack rooms (and studios) that were wired on a point-to-point basis. My experience is that the point-to-point method is inferior in almost every way.
You could make the case that the point-to-point method is superior in one way -- it's quicker initially. I know there are instances in which you (the designer and builder) find yourself up against an exceedingly short time line for a build project -- and there may not be time or budget to do it right (in other words, the spoke-and-hub way). You slap stuff in a rack and connect the dots with wire. When finished you stand back and say, "Voila."
Now, there is plenty I hate about this. Firstly, when a wire is put in from equipment A to equipment B and they are in different racks, you can bet that before long, you'll need to move item A or item B, and of course the cable won't be long enough then. So, for the second time, you're installing a cable. Oh, and don't forget to remove the old one, right? Is it in a bundle? Have to do wire management all over again?
Secondly, I don't think that the point-to-point method lends itself to documentation well. You can label them in a serial fashion; but when old wires come out, do you re-assign the old number to a new cable, or simply retire that number? If you re-use old numbers, better make sure that any and all old versions of wire-lists are found and deleted.
So let's talk now about the spoke-and-hub method. Each rack in your rack room will have an associated punch-block or some type of terminating method. One or more punch-blocks will placed in the rack at some unobtrusive point. I put those in the back of the rack (yes -- you must have rear rack rails) at a height that is convenient for use of the punch tool. Too low is hard on your back and knees; too high means you'll be using a ladder all the time. In the rack room will be a backboard upon which are also installed the same type of punch-blocks. Trunk cables are then installed between the punch-block on the backboard, and its complement in the back of the rack.
Now, inside the rack itself, everything is connected by way of the punch-block; and to connect equipment A to equipment B (if they are in different racks) you'll make a cross-connect between the associated punch-blocks on the backboard.
So now what are the advantages to this obviously slower method? Well, they're many-fold. When it's time to move equipment A or equipment B, instead of running yet another cable between racks overhead in a cable tray or under the floor, you will instead just put in a new cross-connect, conveniently, on the backboard. And, when installing new equipment in the rack, instead of having to run cables up and out wherever, now you'll make a connection inside of the rack, to the block; and on the backboard, you'll add a cross-connect. The time you invest in doing it this way will pay you back with time savings, over and over and over again during the life of the rack room.
Documentation using this method makes a lot more sense in my opinion. Instead of just labeling the cables in a serial fashion (1, 2, 3…n) I refer to the punch-block in the documentation because they don't change. In other words: a cable starts off at (as an example) block 1, column 1, pair 1. If you use a spreadsheet program such as excel to keep track of these, it's simple also to put hyperlinks in so that you can bounce over to the document for the other end of the cable in question.
As far as the types of blocks, there are not that many choices. Stay way away from 66 blocks. I like the ADC QPC style of punch blocks myself, mainly because of their capacity for multiple wire punches on the same posts; but, Krone makes great blocks as well. These are nice because you can insert a shoe in to each pair of connectors to facilitate testing.
Keep it at bay
This may seem a little too old-school for some, but in the rack itself, I still like to use patchbays (even for AES, with the appropriate type of patchbay). This facilitates testing and emergency response. Even the least technical person can usually be talked through inserting a patch cord when it's needed.
Irwin is RF engineer/project manager for Clear Channel Los Angeles. Contact him at email@example.com.