JPI EDM 900 FF erratic

[cwaters]

I have a EDM 900 in my M20J and recently (within the last month) the FF reading has become erratic. I first noticed it spiked from my leaned out 8.5 GPH to 20 (I've never seen a reading that high) and then returned to normal and stabilized. Then on recent flights I've noticed it being stable then beginning to fluctuate higher than the actual flow (I have not seen lower) and not returning to stable flow. All other parameters appear to be stable and the engine is not running any different. No recent engine maint, last annual was 15-20 hrs ago and the issue started rather abruptly about 2-3 hrs ago. Currently its an annoyance as the JPI is wrong on the high side of FF the estimated amount remaining is always low (not a bad thing) so i have more in the tanks than it thinks i have. 

 

The Google machine suggests it could be the transducer, I was wondering if anyone here has had a similar issue and what the fix was/ cost of fix? 

[PT20J]

JPI uses a Floscan transducer. The transducer is designed to create a vortex of fuel within the body and there is a little plastic pinwheel that floats on the vortex and rotates with it. As it rotates, it interrupts a light beam that causes pulses to be emitted by a photo transistor. The wires carry the pulses to the EDM 900 which counts the pulses per second. The K -factor is merely the number of pulses (in thousands) for a gallon of fuel. The EDM 900 converts pulses per second to gallons per hour using the K-factor for scaling.

The primary failure mode of the transducer is for the bearing to wear out creating drag on the pinwheel which causes it to read zero or low or intermittently. To cause it to read high, something would have to be making MORE pulses. That could be caused by a loose connection, either to the light source (causing it to flicker) or on the output. JPI uses a cheap Molex connector for the connection between the transducer and the EDM, so I would start there. 

Skip

[Aerodon]

JPI sell a fuel flow tester that you can plug into your harness to test the wiring from the transducer back to the instrument.  It simulates the pulses generated by the transducer and gives a steady state reading.

But before you waste your time and money on that, try determine how old the transducer in your plane is?  It's quite possible that it was reused from a previous JPI install, or even the factory hoskins or other unit?  When they fail, the do go 'erratic' both low and high.  

Then the quickest and easiest fix is to replace the transducer and reprogram the K-factor in your EDM900. 

Aerodon