‎Engine Start CB-Oct2010-testing02

  EDITORIAL NOTE: 
  Dave emailed the version below to Thom at 21:42 on 04Nov.
  Then we e-chatted a bit and Dave said he'd send another version.
  Alas, Thom seems not to have a newer one to rush into the wiki for the Saturday workers. 
  However, he's afraid to engineer the updates from our e-discussions (especially 
  because Dave talked him out of one idea.)

• 40A breaker problem (starting solenoid and glow plug circuit)

This doc describes measurements taken to diagnose the 40A breaker tripping problem. The work was done on <2010-11-03 Wed> by Thom S, Dave W, and Ken S.

According to both versions of the wiring diagram, the 40A breaker provides +12 only for the cockpit starting solenoid and the glow plug circuits. As far as we know, no other circuits take power from this line, including the start switch in the engine compartment.

First, to isolate the cockpit start/glow plug lines, we unplugged the three-wire connector in the engine compartment. The signals in this cable are +12, start, and glow.

On this isolated cable, we took a number of resistance measurements between ground and the lines using both an ohmmeter and an audible continuity tester. All measurements were done at both the engine end and the control panel end. To detect an intermittent short to ground, we moved bound cables, loose wires, and the stop plunger while taking measurements. Before every measurement, we checked the setup by probing another known ground point.

Everything was normal:

   |-------+---------------+---------|
   | line  | test location | reading |
   |-------+---------------+---------|
   | +12  | engine        | open    |
   | start | engine        | open    |
   | glow  | engine        | open    |
   |      |                 |        |
   | +12  | control panel  | open    |
   | start | control panel | open    |
   | glow  | control panel | open    |
   |-------+---------------+---------|

At both ends of the cable and in the engine compartment, we looked for obvious shorts, insulation cuts, loose connections, and evidence of sparking; everything appeared normal.

To estimate the current running through the 40A breaker, we took resistance measurements of the starting solenoid to ground: 0.3 ohms, and the glow plugs (all four in parallel) to ground: 0.6 ohms. Unfortunately these are at the edge of the ohmmeter's range and resolution, so the accuracy and precision are extremely doubtful.

To get actual current readings with reasonable accuracy and precision, we measured the current running into the glow plugs and starting solenoid using a clamp-on ammeter. All readings were below the 40A breaker rating:

   |---------+------------------+---------------+--------------------|
   | pressed | clamped wire    | test location | amps (approx mean) |
   |---------+------------------+---------------+--------------------|
   | glow    | +12              | control panel |                23 |
   | glow    | yellow glow line | top of engine |                21 |
   | glow    | big main +12    | starter        |                22 |
   |        |                  |                |                    |
   | start  | red +12          | control panel  |                10 |
   | start  | big main 12v    | starter         |                10 |
   |---------+------------------+---------------+--------------------|

To see whether ~30A would trip the breaker, we pressed both glow and start concurrently; the breaker remained closed.

The wiring to each glow plug appeared normal. The glow plugs didn't feel loose, and there wasn't any obvious cracking. We measured current while lightly rocking each glow plug and the reading was always constant, so there is no evidence of an intermittently-shorted glow plug.

Finally, we exercised the system repeatedly by running the glow plugs (load detected by a drop in the DVM at the nav desk), and running the starter. The system operated in all tests; we never observed the breaker tripping. The engine cold-started and ran normally.

So, at this point we know that the problem is intermittent, but we haven't isolated it to the wiring, the glow plugs, the starting solenoid, or the 40A circuit breaker.

If the breaker is triggering at too low a current, and if it is a typical heat-measuring device with a bimetallic strip, then we could perturb it further by warming it up with a hair dryer and testing under load. If it trips under normal load with relatively little heat, then that would be evidence the breaker is defective.

It would also be useful to run tests with the engine compartment hot. It is possible that an intermittent short would occur more frequently in a different temperature range.