Temp Gauge Spiratic and Jumpy
 

2002 vvt miata
speedyefi pnp

Ever since rebuilding this motor ~ 300 miles ago, the temp gauge has been giving me grief. It will randomly jump around and then go to its actual temperature on tunerstudio.
the gauge on the dash is fine, but my readings are getting thrown off, namely when I'm doing a pull it seems.
Ill attach some logs of it doing it.
I went ahead and followed the fsm repair and diagnosis of the issue and it led me to finding pin B having no continuity to pin 4o in the ecu. It had continuity to ground when the ecu was plugged in, my guess backfeeding through other sensor grounds? Either way, no cont to ground when unplugged. I took apart the entire loom engine side, and found the wire that had no ground connection. I followed it back to find a break all the way to the bend before it entered the firewall. I had to get to work, so instead of delooming my entire car, I jumped that sensor ground to another sensor ground (good idea or bad not sure.) This worked swell for all of yesterday, but today doing a small pull, the reading in TS went down to 95 and played around there. I got home and soldered the jumper wire on thinking it was just a bad connection. This did nothing.

the sensor tested fine according to the repair and diagnosis procedure.

Im a bit confused. When I unplug the sensor with it running, it will go down to about -27, then jump back up to around that 95 number, -+ some number randomly.
Im not entirely sure what sensor ground I jumpered to, but it was any of the sensors past the middle of the valve cover towardss the front of the engine. TPS, Crank sensor, ect.

Logs attached,.

Jumpering to another sensor ground isn't really bad, there is typically only 1 or 2 sensor ground pins on any given ECU that all the sensors gang together onto at some point. Are you positive the sensor ground tree is fully isolated from chassis ground when the ECU is unplugged? Are you saying the number in TS drops to -27 when unplugged, but then moves back to ~95 still without the sensor plugged in? If so it sounds like the signal wire is getting some interference/backfeeding/etc assuming open circuit is read as -27 in TS.

Only other thing I would suggest would be using non-insulated butt crimps and heat shrink for connecting the new sensor ground, solder can be problematic with vibrations long term.

Not positive it is, I went ahead and redid the jumper earlier, made sure everything was as it was supposed to aside from that, when I took the ecu connector out again today, it hadnt been snapped in and came out pretty easily. Im assuming redoing the jumper and making sure that plug was all the way seated fixed it, because it acted normal for the 3 hours i was driving.