Forum » EFI Tuning Fundamentals » Ford factory widebands

Ford factory widebands

EFI Tuning Fundamentals

Discussion and questions related to the course EFI Tuning Fundamentals

Page 1

My car, a gt350 has upstream (of cats) and downstream sensors, narrow and wide. I have been told by several ford tuners the widebands are excellent. Does anyone have experience with these. I know when I put it on a dyno I will use an external also as a reference. I have been watching fuel trims and everything looks pretty good so far NA but when I put on turbos I am concerned that the factory widebands may fall short, or I need to be very careful where they are located in the system.

I don't have personal experience sorry, however I'd always confirm the logged wideband data with data from an external wideband that you know. This will confirm their accuracy and assuming they check out then you can use them for your tuning. I've found several factory wideband sensors tend to be very accurate around stoich but less so at the sort of AFR we run at WOT. That's not a rule to live by but I'm always slightly suspicious of factory wideband data.

Rule of thumb is, the newer the car, the more accurate the factory wideband, due to emissions regulations. Denso and Bosch are the main suppliers for modern widebands on modern cars. They use a different principle for operation, but certainly anything from the last 5 years is going to be good. The older sensors lost accuracy in the very rich operation range you find at high load and higher rpm. Newer engines need more accurate sensors to control carbon monoxide emissions.

Sensor readings are all going to vary a little bit according to location due to differences in pressure, temperature, and mixing of gases.

I bought spares for bank one and 2 should I find issues I will first replace them, and retest. The Voodoo has different widebands than even the 5.0 coyote of the same year and generation. I do believe that the factory 8250RPM limit, combined with the nature of the stock tune to get to 526hp required the use of a fairly accurate component. But time will tell. Thanks guys.