When I was doing my Phase I testing I got the fuel flow calibration pretty close with a K factor somewhere between 70,000 and 71,000. In the last month I started noticing the calculated fuel consumed was significantly more that fuel gauges and actual consumption. Most recent flight I had bumped the K factor up to 74,000, but still had a calculated fuel consumed of 47+ gallons, and actually pumped in 36 gallons. I suspected a loose connection, and just inspected the connectors. I was able to measure 0 ohms across each connection, and black wire having 0 ohms to engine ground. The sensor and airframe have approximately 195 hours. It is the "red cube".