So my 2014 Maxima SV with 145k is showing a reading of High H hot on the dash instrument temp needle in the hot summer heat when I stop at lights and in rush hour stop and go traffic. When I pull off it goes back to normal in 5 secs give or take. Does this over and over. On the highway no problem till I stop. Had the Coolant checked for exhaust gas = head/block crack. Shop said no. Its been bled 2 times now. One shop did a digital Infrared thermometer check read on the large and small coolant hoses. One showed 120 degrees the other showed 180. They said it was normal temp, but at that time it was showing all way way up on the gauge to hot.

They could not figure it out so I took the car back. I was thinking about replacing the Temp sensor, but read that maybe that is Not the sensor that sends the gauge needle the signal? Read somewhere it's a "thermal transmitter" that sends the signal? I have no idea where that's located or if that is even correct info? Also not sure if that's really the problem but thought it was worth a shot at changing. I also saw somewhere that someone changed that had the same problem and it turned out to be the connector or something right there at it?

At this time I have not had any trouble check codes. They also told me there are no overheat codes in the history scans at all. There is Zero coolant leaks, never any at all. Anyone have a idea of what may be going on? Thanks.

That has to be a bad cluster. There's only one temp sender on a '14 and that's read by the ECM, which then passes the data to the cluster via the CANbus. That means the error has to be internal to the cluster, something (probably a bad stepper driver) is causing an incorrect needle position.

PS - That's assuming at least one of the places you took it streamed the data from the ECM with a scanner and the reading from the ECM isn't showing hot. If the pinheads didn't scan the car then you could have an actual problem with flow through part of the cooling circuit. Just because there's a temperature drop across the radiator doesn't mean flow is sufficient through all parts of the system. The block and heads have separate cooling loops, and the ECT sensor is reading temperature at the coolant manifold coming out of the heads. So it's entirely possible for the heads to run hot from a blockage in that loop without showing anything abnormal at the radiator.

It was a Nissan dealer of course. I have no confidence that they read or monitored any scan data. So the 1 temp sensor goes into the head not the block? I think I am going to change the temp sensor just to eliminate 1st.

If it's not the temp sensor, what would be the best way to check for blockage and where it is? Just to explain further, it will be doing ok, then when you get to a point in stop and go, or at a stop light it will just start to climb. I can cut the engine off. Wait until the light changes, then take off and the needle will go back to normal in 3 - 10 sec at most.

This morning after talking with the Nissan dealer and asking pin point questions, I am being told that it's borderline on the exhaust in the coolant test that it could be possibly a crack in the head, head gasket or block crack, some kind of internal engine crack. At 1st test they told me it was not. I got this car back in Feb 2023. Paid very little fort it. I seems to be in good mechanical condition and service. I drove it about 4 hours back to home with no indications and no gauge overheating until it hit rush hour traffic, then discovered this problem.

I have not had time to look into it until now it's been sitting. I 1st changed and drained and refilled the coolant and bled it. Still same problem, so I had a radiator shop bleed it again. Still same problem. It's confirmed that no coolant is leaking any where. I don't see any indication that coolant is in the oil ( i changed it), nor burning or smoking out back. I just can't see a engine crack driving for 3-4 hours with normal temp, then only overheating when coming to stop and go?

Any help or experience with this 3.5 engine would be appreciated. Thanks.

The first thing you should do before throwing parts at it is get a cheap scanner that can stream and see if the gauge reading corresponds to what the ECM is seeing. A China-cheap ELM327 and a copy of Torque Pro or NDS-II will set you back about $25. Working blind will just cost you lots of money and time and possibly get you noplace at all.