Fuel Tuning Woes... Grand Canyon of Fuel Maps

Hey All,

Vehicle: 2003 Nissan 350z (VQ35DE) - Basic bolt ons

ECU: Link G4x w/ WB2 Widebands + Map + IAT sensor (Tuned on a Mainline Hub Dyno)

My car was tuned on a mainline dyno about a year ago. It was tuned cell by cell and created a nicely shaped fuel map.

A week or so later after the initial tune, I got a CAN guage which shows live Lambda for banks 1 & 2 and noticed it was quite lean around 2250 rpm at light throttle cruise. I ended up road tuning this out over a few weeks and got the car to drive fairly nicely, albeit probably a bit too rich.

Fast forward to yesterday, where I am getting the car retuned from the traditional to modelled fuel equation and the lambdas were reading incredibly rich in the 2000-2250 rpm range... We gave the engine what it was asking and it ended up creating a huge trench all the way through this area of the map. At about 2500rpm, the map starts to become normal again.

We ran the general area of the map over and over again, and nothing changed. We actually went back to the traditional mode and tried to tune the same area, lo and behold it was asking for the same trench in the fuel map?

What is really strange, is when I go on the street with the 'trench' in the fuel map, the lambdas will show it's extremely lean in this area (which makes sense as we pulled out a bunch of fuel), but on the dyno when tuning cell by cell, the engine is rich and is basically asking for this trench in the map?...

I've had the suggestion to add a Fuel Pressure Sensor closer to the fuel rails which I will do today.

Has anyone had experience with rapid changes in VE like this? Or have you seen the dyno ask for one thing, but on the street it's demanding something else?

Thanks for any input. I've attached the image of what the map looks like below