While running a train around after ballasting the ‘sawmill’ side of the layout, a section of track appeared to be dead. The section at the bottom of the image. This was frustrating, because I’d made a point of testing track as it was wired, and running trains for a while before ballasting. I knew that once the ballast was in, tearing up track to fix problems was going to be much harder, and there’s the inclination not to tear up finished work. Most conductivity problems have been caused by paint/glue/ballast getting between the points and the stock rails. Attention with an abrasive block on the tops of the rails and sandpaper where the points meet the stock rails has usually fixed the problem. This section was dead. I put a multi-meter across the rails and got zip.
Aside: yes, I know multi-meters won’t correctly read DCC waveforms, but on live track you will get some number on the display. I wasn’t looking for a true voltage reading, just some indication that there was power.
I decided to solder the joints on the turnout that weren’t required to be insulated (the running rails from the frog). I’ve read that some modelers solder joints while laying track, and now I see why. It’s significantly harder to do that after track has been laid and ballasted. For good measure I soldered the joints on both ends, and the joints of the next turnout. Of course, you’re going to end up with blobs of solder on the rail, so the next step was to chuck a cutoff disk in the motor tool and carefully remove solder while shaping the rail profile. In N scale, this isn’t an operation easy on middle-aged eyes. However, the results were fairly credible: Everything works. The track is smooth enough that even the lightest cars may wobble (somewhat prototypically), but no derailments.