If it is mounted to the firewall and if the grounding is through the npt threads, then is the ground path through the hose/fittings???

Perhaps the fittings are sufficiently dirty/cruddy to provide a poor ground?

 

 

My first suspicion would be an electrical problem with or around the oil pressure sensor.  You mentioned that you rigged a ground wire.  Depeding on how it's hooked up it may or may not affect the problem.  I've seen some engine sensors read erratically because of where the ground lead from the engine interface box (the box all of the engine sensors are wired to) was physically attached.  The problem may affect multiple sensors, or only one.  For example on a Legacy using a JPI engine monitor the installation manual said to attach the ground wire from the engine interace box to the engine block.  We first tried an unused stud (it was real convenient) on the back of the engine where the vacuum pump would be, if it had a vacuum pump.  The tach reading jumped all over the place.  We removed the ground wire from that stud and attached it to the large ground stud running through the firewall to the engine.  The tach worked fine after that.  After thinking about it I suspect attaching the ground wire to the vacuum pump pad created a secondary return path from the engine back to the rest of the system that was in parallel with the primary return path (a "ground loop").  Bad. 

 

What type of oil pressure sensor do you have?  Is it a 4-wire sensor?  A 3-wire sensor?  Single wire sensor?  If it's a multi-wire sensor does its ground wire run back to the engine interface box?  If it's a single-wire sensor, and the sensor case isn't attached to the engine, then the case of the sensor should probably have a wire running back to the same ground as the engine interface box.  Does the installation manual specify how to ground the sensor?

 

Tom Gourley

 

Hello Jeff-

               A few WAGS:

 

  Do the connectors have a wire sleeve and are they crimped with a ratchet crimper? I had a fluctuating EI fuel pressure gauge and the shop had used automotive crimpers. I recrimped or replaced the connectors and the problem went away.

 

  Is the engine new as opposed to reman? I bought a new replacement engine from TCM and discovered the seat for the pressure relief was rough. It had to be faced with a hand tool supplied by the tech rep. If it was a reman I think the odds are against that having gone unnoticed in previous runs.

 

  I have an aerobatic Continental IO-360 in my Chipmunk. When I bought the plane the oil pressure would be stable when cold but fluctuate erractically when hot. I did a lot of of testing looking for a correlation to power settings, tried adjusting the pressure relief, even changed the spring and plunger; no good. En route to OSH the pressure dipped to 0 briefly then came back up. Quick landing, fiddle with the relief, ground run was fine. A while later in cruise it happened again. Landed, replaced the engine with new. On test flight, the SAME fluctuations! Refacing the seat didn't help (see above). TCM didn't know what to do. Turns out it was cavitation- the Christen valve and hoses used for Lycomings (-10) is too small for the Continental. At cruise settings there just wasn't enough oil getting to the pump. The engine was originally used for a military trainer down under- the manufacturer discovered they had to make their own valve body with -16 fittings. That solved my problem. Do the fluctuations go away at lower RPM settings? If so there might be a restriction in the pump supply.

 

 The TSIO-550 Overhaul Manual shows that the pump body is sealed with silk thread. I know nothing about it but if there was a pinhole leak in the right area?

 

  I think the mechanical gauge makes sense as a check but it's another hole to seal for pressurization.  -Bill Wade