1984 DL, sunburned ruby red with 325K...

The failure light came on and no amount of stomping would reset it, yet both circuits have no leaks I can see. Against all recommendations I could find I decided to see if I could reset it through the switch hole.
Before loosening the switch (14mm 6pt) I used a continuity tester to verify it was still switched on - in the failure mode. Then, after removing the switch body, spring and pin this was staring at me via bifocals, flashlight and mirror:

Looked like a thing you'd unscrew with a standard screwdriver. But after screwing with it for a while and still seeing the same threads exposed, I decided to stick a hook in and pull it out:

Got all the switch parts out so I could now make out the shaft between the two actuator pistons, I guess are in the ends keeping the brake fluid where it belongs. No threads on that last part, but why it has a slot I have no idea! Cleaned up the switch and used a few Q-tips to clean the opening in the junction box. The probe shown above the switch parts in this photo is how I determined, with my continuity buzzer, the shaft was insulated.

I could see a dimple where the switch contact driven by spring pressure might have dented what looked like shrink sleeve over a steel rod:

Put it all back together, and like the probe predicted, no more brake failure light. Cliff is right -- I spent 10 minutes trying to start those nylon threads back into the junction block -- the spring fights you.
Here's my theory. When I changed a caliper a month ago I upset the balance temporarily. The piston moved, but the switch was dirty inside and took a month to make contact. The spring tension in the switch kept it stuck until I removed the switch, allowing the old piston springs to once again center it.
Either that, or that dimple is worn deeply enough to break through even though I could not with a sharp probe gently applied. Less likely I believe.
Hope I never have to replace this octopus.
--
Art Benstein near Baltimore
|