Oxygen sensor placement for twin turbo?

[IBLOWN]

As some of you know im doing a twin turbo setup on my ecotec.

The turbo's are mounted on the same side which means a crossover pipe is used.

My question is: Where should i put the sensors? There are 2 of them. If i put them on the dumps the ecu will be out of whack because the gasses will be alot hotter in that position.

To me they need to be at the same length as where they were factory whcih means before the turbo's.

The drivers side is easy, all i have to do is put one sensor in the crossover pipe but the other side will be a mission. I was planning on putting it in the merge collector but i dont think the gasses will have mixed enough from all 3 ports to give an accurate reading.










Any suggestions?

[Boonz]

maybe use an ECU that uses a wideband o2 sensor and have it back in the merge pipe further back in the exhaust? like next to the trans in the same position the VR's use?
thn you will be able to tune it more acurately and only have one sensor

[VN_Luke]

put em both in the dump pipes! just make sure they're not too low/on the bottom so that condensation/water build up doesn't get at them.

heat shouldn't be a problem - the sensors need to be hot to work anyway...

and any factory turbo cars i've seen have them in the dump pipe too

[IBLOWN]

put em both in the dump pipes! just make sure they're not too low/on the bottom so that condensation/water build up doesn't get at them.

heat shouldn't be a problem - the sensors need to be hot to work anyway...

and any factory turbo cars i've seen have them in the dump pipe too

That was my initial thought but isnt the ecu calibrated for the sensors to be at a certain length/temperature down the pipe? Too close and its too hot, too far away and its too cold. The exhaust coming out of the dump is gonna be HOT which will make the ecu think that the engine is running leaner/richer than it really is!

Im not worried about sensor life, need the readings to work with the stock calibration of the ecu.

When you said you've seenfactory cars like that do you mean they were N/A first then made turbo?

[VN_Luke]

That was my initial thought but isnt the ecu calibrated for the sensors to be at a certain length/temperature down the pipe? Too close and its too hot, too far away and its too cold. The exhaust coming out of the dump is gonna be HOT which will make the ecu think that the engine is running leaner/richer than it really is!

Im not worried about sensor life, need the readings to work with the stock calibration of the ecu.

When you said you've seenfactory cars like that do you mean they were N/A first then made turbo?

i wouldn't stress about the sensor calibration - to be honest I doubt the temperature of the sensor has any actual effect on the reading that it gives. I think the normal operating temperature is ~300-350 degrees C - but work till a much higher temp! - for reference, Innovate (who make my wideband setup) specify 18-24 inches away from the turbo outlet.

All the jap cars with turbos are running their o2 sensors in a cast dump pipe, right behind the turbo outlet (search for 'dump pipe' on ebay and looksee at all the aftermarket ones and where the bung is)

also, for more reference; on the dyno i can put the wideband in the dump pipe or in the end of the tailpipe, and it reads the same-ish.

In either case - the o2 feedback when the ECU is in closed loop mode isn't a drastic trim - it only operates when it should be cruising at 14.7:1 - so no actual difference on full noise runs etc...

[IBLOWN]

I just thought the sensor purely worked on temperature.

Ok cool, thanks Luke!

[Boonz]

i think your thinking of the cat converter... the further away they are from the manifolds the less effective they become. they need to be hotter to work more efficiently

[VN_Luke]

i think your thinking of the cat converter... the further away they are from the manifolds the less effective they become. they need to be hotter to work more efficiently

Yeah I reckon you're right! - every "how to pass an emissions test in your TT ls1 - by Dodgy bros. performance and shoe repair" article I've ever read says to get brand spanking cats and put them as close as you can to the engine so they start working quicker and better!


As for the o2 sensor, I believe they work on a chemical reaction - they generate voltage based on how much oxygen they can sense in the exhaust pipe. I think they compare oxygen inside the pipe vs oxygen outside or something crazy like that. In either case, they don't actually start working till they get to about 300 degrees - which is why newer O2 sensors all have heaters - so they can work from the word go! (as opposed to the older 'single wire' sensors, which heated up with the zorst and THEN they worked)

also you will find that with a set of extractors on an otherwise stock V6 manual - when you're coasting down a hill engine breaking, you will get an error code because the o2 sensor drops below it's operating temp!

[danja]

As for the o2 sensor, I believe they work on a chemical reaction - they generate voltage based on how much oxygen they can sense in the exhaust pipe. I think they compare oxygen inside the pipe vs oxygen outside or something crazy like that. In either case, they don't actually start working till they get to about 300 degrees - which is why newer O2 sensors all have heaters - so they can work from the word go! (as opposed to the older 'single wire' sensors, which heated up with the zorst and THEN they worked)

Spot on, they do need to be hot to operate (thus heated sensors as Luke mentioned, they get to operating temp much quicker)

[DannyboyDS]

I thought the oxy sensor worked on thermal resistance, the hotter it gets the more resistance it causes across the circuit, since exhaust temps are directly related to air fuel ratios that was my understanding of it anyway.

[VN_Luke]

I thought the oxy sensor worked on thermal resistance, the hotter it gets the more resistance it causes across the circuit, since exhaust temps are directly related to air fuel ratios that was my understanding of it anyway.

then you'd be reporting a linearly leaner mixture when say... doing a burnout.

i.e. max load, max rpm, for a period of time - the exhaust will get RED hot, etc - but the AFR's stay the same!

[immortality]

I thought the oxy sensor worked on thermal resistance, the hotter it gets the more resistance it causes across the circuit, since exhaust temps are directly related to air fuel ratios that was my understanding of it anyway.

then you'd be reporting a linearly leaner mixture when say... doing a burnout.

i.e. max load, max rpm, for a period of time - the exhaust will get RED hot, etc - but the AFR's stay the same!

temp does effect the O2 sensor reading, thats why it is best to measure both for tuning purposes. i don't think doing burnouts is a good example, as engine temps rise above normal the ecu actually injects more fuel in, i.e. it makes it richer. helps stop detonation. saw it somewhere in the tunes using tunerpro

[VN_Luke]

temp does effect the O2 sensor reading, thats why it is best to measure both for tuning purposes. i don't think doing burnouts is a good example, as engine temps rise above normal the ecu actually injects more fuel in, i.e. it makes it richer. helps stop detonation. saw it somewhere in the tunes using tunerpro

If you wanna play technical - burnouts is a bad example as the car will never be running in 'closed loop' mode with that much throttle anyway

i would have thought people measure both exhaust temp and afr as a way to find the leanest possible mixture they can run before things turn a bit pearshaped - i.e. read the wideband for the AFR, and slowly lean it out bit by bit until the EGT's start going a bit high. - thereby producing peak power and best fuel efficiency.

Though this isn't really a practise you'd use on an every-day car anyway, because something as simple as a bad batch of fuel could make a mess of your motor! - along with naughty emissions from running lean etc

I'm still not believing that sensor heat has an effect on AFR output, though (mind you - I'm happy to be educated/proven wrong!). Another example - the wideband on the dyno reads same same @ idle before and after a few decent power runs (which can get the exhaust mani. flowing red hot).

A cold sensor won't work - as there's appraently heat required to generate the chemical reaction that causes the sensor to produce voltage - I accept that.

but a sensor at 300deg C and 400deg C would read the same I reckon! - however if the actual EXHAUST temperature was 300 or 400 degrees, chances are that there is actually a difference in AFR's - so it would read differently if that's the case.

essentially what i'm saying is - run a car at idle - say the sensor reads 14.0...... then apply blowtorch to the exhaust around the sensor without changing anything else..... sensor should still read 14.0!

(and remember that the 'narrowband' sensor only reads pretty much two values... i.e. 'above 14.7' and 'below 14.7')



*information above relates to the common narrowband style sensor - i.e. the 'zirconia' style http://en.wikipedia.org/wiki/Oxygen_sensor

[vt1538]

Voltage output from a narrowband sensor is changed by exhaust temps. Hence the reason they are used as a binary type switch around stoich & not for an absolute AFR value. See attached pic.
A wideband however is accurate at most temps but becomes inaccurate as exhaust pressure rises away from atmospheric pressure. This is why you should always sample downstream of the turbo.

[VN_Luke]

Voltage output from a narrowband sensor is changed by exhaust temps. Hence the reason they are used as a binary type switch around stoich & not for an absolute AFR value. See attached pic.
A wideband however is accurate at most temps but becomes inaccurate as exhaust pressure rises away from atmospheric pressure. This is why you should always sample downstream of the turbo.

I see (and stand humbly corrected)

Although the graph with the 400 degree change in temperature in the pic illustrates that the ecu fuel correction would function the same at 500 degrees sensor temp and 900 degrees.


Which in turn suggests IBLOWN can mount em pretty much anywhere and they'll work fine (pre cat)

[vt1538]

I see (and stand humbly corrected)

Although the graph with the 400 degree change in temperature in the pic illustrates that the ecu fuel correction would function the same at 500 degrees sensor temp and 900 degrees.


Which in turn suggests IBLOWN can mount em pretty much anywhere and they'll work fine (pre cat)

Exactly. Though I would personally still place them downstream of the compressor to keep them at a more constant temp.

[VT-565]

Bang.

How oxygen sensors work

[Cheap6]

Already covered.