I removed the spurious db_R.vdc.dat in the database and this seems to have fixed the problem. I replayed run 1526 and it looks good. There was no signal in run 1489 because I was testing on this run and the missing channels were on slot 13.
http://www.jlab.org/~adaq/halog/html/1102_archive/110217143231.html
So this was a database issue because of the type.
Good catch from Ole.