sensors in exhausts

[Adsport]

Hi, i own an AW11 MR2 supercharged with a couple of bolt ons, after installing HKS type-II headers. they are a 4 to 1 design with primaries 1100mm long with 4 balance pipes (yes they look very cool) , i put the o2 sensor where it was (at the junction of the primaries, 1000mm down the line) .the headers were getting VERY hot and red, so we figured it was leaning out and bumped up the fuel via SAFC.

im forseeing this as a potential problem as when the o2 sensor was in its original position (closer to the head) the issue wasnt there and now i think its getting an inappropriate signal and causing bad/wrong readings.

having played with many jap cars with exhaust temp sensors in the dump pipes of turbos etc and seeing some that are really close to the head or IN a turbo i cant help but wonder what the proper orientation for temperature and oxygen sensors is? is anyone able to shed some light onto the subject??

im thinking of getting the headers ceramic coated, so before i go and do that i would like to put sensor bungs and flange(s) in for oxygen sensors (narrow and wideband) before the coating or its going to be a partial waste of time.

so here are the actual questions:

1/ is there an actual appropriate distance from the beginning of the heat source (head) that all sensors (o2 and temp) should be so that they are accurate?

2/ will the above sensors get the same signal being in one runner in the right place as at the runner junction no matter where it is?

[oldcorollas]

1. yes.. non-heated sensors should be maybe 30-40cm from head, but in turbo apps, the heat before turbo is too high.
heated sensors can be anywhere. you can run them a few feet back at least. they are best as you no longer have a signal that varies with temperature of exhaust (at least from low temps)
if the sensors are too close to head, or temp is too high, then can fail...i think usual temp is 3-600C perhaps?

2. if you get signal from one cylinder, you only get info from that cylinder. putting at junction will get info from all cylinders.
if things are normal, it makes no difference (as long as you have heated sensor at junction, or it stays up to temp), but if you start dropping one cylinder, you can see the change if you are datalogging fast enough when the sensor is at the junction.

that help?

footnote.. heated sensors ftw!

[TooF]

i would have gone down the lines that there isnt a single *correct* placement for them, obviously further down the runner the temp will drop so your tuning will have to compensate for this.

the factory computer will be expecting the temp sensor in a particular location and your new sensor will have to correspond or you will need to tune differently..

just my initial thought.

[oldcorollas]

oh, and for temp sensors, close the head as possible, so then you are measuring the gas output temp....
O2 sensors have minimum temps for working properly, but above that, they work fairly stable up to their maximum/failure temp

as toof says, there is no one correct place.....
exhaust temp sensor close to head
heated O2 sensor in your collector (will be fine, i had one in similar position)
non-heated O2 sensor within say, 30-50cm (but the exact position could be checked with a temp sensor at different positions )

[Adsport]

oldcorollas: thats pretty much all that i was thinking..

toof: i had the same kind of idea regarding the factory sensor , then thought that it must be in that position for a reason whether it be that its factory position either be the correct distance form the head to get the calibrated reading or that its in the junction because thats simply where it needs to go. i doubt that the exhaust temp would remain the same along the whole exhaust and thats whats given ECU the odd reading to cause the lean out.

basically the conditions of tune didnt change, i moved my o2 sensor about 800mm further away and it must be a reading a lower temp reading and it thinks its running richer and the ECU is leaning it out , causing headers to heat up. i might measure some stock headers and i think it should go back in the same distance away , even if its in one runner it should get the right temperature from there?


if i cant get a correct reading on the o2 sensor on narrow or wideband the tune will be all out of whack whether is factory of aftermarket .. hrm geez seems like a big can of worms

[Merudo]

if i cant get a correct reading on the o2 sensor on narrow or wideband the tune will be all out of whack whether is factory of aftermarket .. hrm geez seems like a big can of worms

Isn't the o2 sensor only referenced to in closed loop/cruise conditions?

[Adsport]

typically in a stock setup i beleive so, but im looking at the insertion of a wideband for tuning of aftermarket ecu in the long run aswell as exhaust temp sensor output for instruments. good point though

[thechuckster]

i have an Innovate LC1 with the windeband sensor just inches after the turbo. But it could have also been plumbed in at the base of the dump pipe and work just as well.

Main thing to avoid is the sensor haveing water condence in it or the external part protrude ino the path of things like rocks and other road debris.

Also, the wideband sensors have their own heaters so you could (if required) put them halfway down the exhaust system. They have to sit before any catalytic converter.


edit: the only thing position sensative is a temp sensor. O2 sensor can go just about anywhere. No can of worms here...

[off-road]

Dyno dynamics usually have a sensor on a bracket that they shove up the tailpipe so position of sensor not that important for their tuning purposes.
What are other peoples experiances with dyno tuners ? Do they screw their sensors into the bung in the manifold or shove it up the tail pipe.
I always get them to screw it into the manifold but that is because I like to make sure the exhaust extraction pipework can be connected properly to ensure a nice 'clean' dyno room.

[thechuckster]

as long as you dont suck normal atmo into the zorst system (e.g. back up the tailpipe or via leaks and dodgy flanges, etc) a wideband sensor can go anywhere... but in the manifold or dump pipe is ideal.

[nick.parker]

Main thing to avoid is the sensor having water condence in it or the external part protrude ino the path of things like rocks and other road debris.

Yes, install the sensor on a higher than hoziontal angle. 45 degs from vertical is best so sensor body doesn't overheat and water (if condensation occurs) will drain out (according to Bosch, not me) this is how I installed mine, its been fine for 18 months.

Dyno dynamics usually have a sensor on a bracket that they shove up the tailpipe so position of sensor not that important for their tuning purposes.

Yep true, but at idle and low revs reversion at exhaust tip mixes in fresh air and your WB will read lean. If you intend to use the wideband at low revs / closed loop cruise this will not do.
But we all should have cats right right? So the sensor should be upstream anyway!!

[Jonny Rochester]

I will add, the factory computer, or other computers like MicroTech will normaly be set to work with a narrowband O2 sensor. The narrow band gives a 0 - 1volt signal.

A 5-wire wideband sensor gives a 0 - 5volt signal and will normaly need a wideband contoler that has a 0-1v output if you want that signal to go into your old MicroTech or factory ECU. A new MicroTech can be ordered to read a 0-5v signal.

[oldcorollas]

just to add to that.. all widebands need a controller.. otherwise there is nothing to tell the oxygen pump what to do..
most aftermarket wideband controllers give you the option of a 0-1 or 0-5V output too (or they should )