After pulling what little hair I had out, I found the short. It was the right rear marker lamp.

I took a blown fuse and used it to solder in wires to a miniature 8 amp circuit breaker so I could test under load without wasting fuses. Anyway, the right rear marker lamp positive contact is clamped in a piece of insulating plastic. It had cracked and allowed the contact to occasionally touch the lamp base (ground) and short out. I used JBWeld to repair the plastic part and all is fine.

Previously, I found out while trying to troubleshoot using only resistance to ground measurements that cold lamp filament resistance goes up by a factor of about 15 in the first 0.1 second or so when power is applied. So, what looks like a possible low resistance, say two or three ohms, will really draw much less current than it appears.

Brent Smith
Atlanta, GA
1986 300E 274,700 miles