Wideband O2 Sensors for 940

[Pete940]

There doesn't seem to be a lot of info on the net regarding what wideband sensors fit the B230 engine.

Is the universal Bosch wideband suitable for these engines?

If so, could anyone provide a description of how to wire up the 2 extra wires to the ECU? (Since standard is 3-wire and Bosch wideband is 5-wire)

Additionally, what's involved in logging the afr? I'm guessing you need to just splice up some of the sensor's wires?

Cheers.

[baggy798]

The stock oxygen sensor is a narrowband type. You can't replace it with a wideband variety as the ECU won't understand the output.

Universal Bosch - do you mean the LSU 2.4 ?

If you want to log the AFR's you need a proper wideband system with a separate sensor. The narrowband gauges are effectively junk.

You will need to weld a bung onto the exhaust, but if you don't want to pay someone to do that you could use an AEM no-weld clamp like I did. There is just enough straight pipe at the top of the downpipe for it:

[TonyS9]

What does yours tell you? I know how they work but I'm interested in MPG these days and think that the LH2.4 over-fuels somewhat. By playing with the accelerator you can get major differences in MPG (using an MPGuino on the injector line + speed line) without a perceptable change in acceleration.

How quicky do the widebands react? How accurate are they? Is it possible to wire up an offset for the narrowband to make the LH2.4 fuel on the lean side.

[Pete940]

Originally Posted by baggy798

If you want to log the AFR's you need a proper wideband system with a separate sensor.

I see. So what does this "proper wideband system" involve? More than a sensor and gauge I assume?

Do you know why the stock ECU cannot understand the wideband? From what little I've read, it would be unusual if it indeed cannot understand the wideband.

Seems bizarre that there's little info on the net regarding this type of installation for any car.

Thanks for the info.

[baggy798]

TonyS9,
I only had it connected for a short while as I've got to redo the wiring, but from what little I saw LH2.4 sucks in standard form. As reported on turbobricks, it always tries to stay at 14.7:1 during cruise conditions. It does turn the injectors off if you take your foot off over 1900 rpm, and then it kicks them back in at about 1200 rpm. (pegs the gauge at 18:1)

People on turbobricks have messed around with it using an 'Ostrich 2.0', and have been able to lean it out to 15.5:1 during cruise, so there are gains to be had there.

Response time depends how far from the turbo you place the sensor, it was only 4 inches away on mine so the response was very quick. I searched around a bit and some people say if you have it 36 inches back (what it says in the manual) then there is a 1-2 second delay.

Pete940,
Not necessarily. There's a few gauges with built-in datalogging and all that good stuff. I went for the AEM Failsafe Wideband which I think is pretty entry level as these things go. It allows you to datalog using AFR, boost, and an RPM signal. There are of course more expensive options, a popular one at the moment being Zeitronix stuff. Supposedly, Innovate Motorsport widebands are a bit crap and unreliable.

I don't know much, but I believe a narrowband sensor outputs 0-1v and it constantly flits between that range, whereas a wideband sensor outputs 0-5v.

There are some wideband gauges that do emulate a narrowband signal as well, allowing you to replace the stock sensor. However, certain types are hit and miss when interacting with LH2.4. I know the AEM widebands are rubbish in this regard as they don't emulate the signal fast enough, sending LH into a fit and triggering the check engine light.

[volvoz2000]

I went for the AEM failsafe Wideband guage too...... works well

[TonyS9]

I plan to use an arduino for logging lots of stuff, very low cost but you need to get software or write yourself, aswell as a bit of hardware build. I'm sure there is some opensource IP out there.

The MPGuino shows the fuel cut-off well, its not that consistent and its much much better to coast in neutral if you are not trying to slow down. Tickover is very efficient, and you can see the different fuel flow when the lights are on.

The real problem is constant light throttle is very rich and inefficient sometimes, but there are efficient settings you can find. That leads to >37mpg (average on a longish trip), which is impossible to get without the meter. I'm getting about 43 peak I think.