Wideband O2 Sensor Placement on 930
 

I'm interested in installing a wideband O2 sensor (Bosch LSU4.2 5-wire).

I've read that on turbocharged engines the sensor must be installed after the turbo charger, if not, the pressure differential will greatly effect the accuracy of the unit.

Has anyone had any direct experience with this? If you have installed the sensor upstream of your turbo (using the stock location), how much do the readings fluctuate?

put 4-6 inches after the turbine outlet

It's a heated sensor so placing isn't critical. I have exhaust clamp so I attach it at tail-pipe and it works just as well (as long as you don't have exhaust leaks). It will work before and after the turbo. Easiest is to attach it where OEM lambda sits, where tubes from exchangers merge before going into turbo.

From what I read/understand, its not accuracy you are sacrifacing but the life span of the sensor.

Mine is after the turbo.

Same here

OK, thanks, everyone.

I bought some headers that have the O2 fitting at the plenum just at the turbo intake, so I'll place it there and see what happens.

A fellow UK member on the turbo911 board has investigated this on his Motec motor. He has 2 sensors before and 1 after the turbo(running headers). He found that the "stock" location shows .03 of lambda richer than actual. Doing some math of my own, that equates to almost 1/2 point of A/F ratio.(.44 actually) The Bosch LSU is a pump cell design lambda sensor, "pressure will effect the reading. The sensor does not measure an A/F ratio directly, rather, the energy required to return the mixture in the diffusion gap back normal."

more info can be found here:

http://wbo2.com/lsu/default.htm

from that site:

"Another effect is that higher pressures contain more oxygen and fuel molecules and thus require a larger pump current to come to equilibrium. This gives the indication the mixture is richer than it would be if measured at normal atmospheric pressure. This effect is particularly important for turbo engines where poor sensor placement can fool you into thinking you have a richer mixture than you really have. "

Heat not only affects life of the sensor but more importantly it's ability to accurately measure. I have the LM1 installed on on my fabspeed single-out and had to add a heatsink where it mounts to the bung to prevent the sensor from overheating which caused a "sensor error"

disclaimer: I'm not an expert, just an enthusiast:)