I think this might be of some help to anyone who might encounter an N18 fault while trying to obtain program data.
This past week, I was being innundated with "your program data will run out in x days" messages. One of my S2 240 units hadn't successfully connected since the 18th.
I manually tried connecting (it's on a wired network with one of the originally specified Linksys wired USB dongles), and that failed with N18. I could successfully "test" the connection, and had no troubles connecting into the unit via the lan. As it is on the same switch, and from there, the same NAT box as another S2 in our entertainment room, I knew it wasn't a firewalling issue -- the other S2 hasn't missed a beat.
I rebooted the unit a few times, with no change in the ability to connect.
Finally, after reading some older threads here about N18 problems suggesting to unplug the unit, I initiated a reboot, then unplugged it for 20 seconds and went to plug it back in - and immediately got audio from the input (without a time delay - audio was up the instant I connected power), but no video, and no green power LED.
I cycled it a few more times with no change in its behaviour, so I disconnected everything and hauled it out to my workshop. After opening it, I disconnected the fan from the mainboard (as it was a bit on the noisy side), and when I appliied power, noted that the HD wasn't spinning up. If I disconnected the I/O cable from the mainboard, the HD would spin up, but the power LED still didn't come up. I don't have a schematic of the PS outputs (over the film ribbon to the PCB), but suspected that the PS had failed in some fashion. I pulled an old S2 540 unit from a box out in my workshop (something I hadn't brought online because I didn't want to pay the added service fee each month just for a networked TV), opened it and found the PS to be significantly different, but still had the same mounting and output configuration. So, I swapped the suspect duff PS to the 540 and confirmed that unit wouldn't spin up, then swapped the 540 PS to the 240, and had success booting it again.
Back in the house, hooked it up and although it took _2_ hours to deal with the data processing, it completed the connection (WHY are they so fsking slow?). I now need to source a replacement PS for the inactive 540 unit so it's not a heap of junk in the workshop, but the unit with service is back in play.
My take is that the original PS was getting tired (everything looked okay, though there was a slight heat halo in the fibreglass around a 1W or so resistor near the PCB connector, it wasn't a burn mark per-se, and the resistor itself was in fine shape), and perhaps some aspect of the connection was drawing more current than the USB subsystem could pull. Dunno how that could happen with the USB networking still able to communicate just fine on the LAN (and remember, this and another S2 ran through the same NAT *BEFORE* being presented to my firewall, so as far as any of my firewall rules go, they would treat both units the same), but the PS swap resolved things.
I'm exceptionally annoyed that even after the successful connection, the unit continues to blast me with a "new message" about the program data being about to run out (despite having DELETED THE DAMN MESSAGE REPEATEDLY, and also confirming that the guide shows data out for 9 days).