how close is too close for wideband sensor?
 

n/a 99 miata with bolt ons. My after market header has bung for #1 o2 about 4 inches from the flange on the head. Would that be too close for the aem wb sensor. I currently have the wb sensor in a clamp on bung, after the non existent cat (resonator). The reading is very jumpy and my first thoughts are it's too far downstream. Any thoughts?:dunno:

"Jumpy" is a somewhat subjective description. The reading provided by a wideband O2 sensor should be expected to oscillate under steady cruise and exhibit sharp and dramatic excursions on throttle change.

In the absence of a catalytic converter, there's really no such thing as "too far" from the head, right up until you get to the point, just a few inches from the tailpipe, where fresh air starts to interfere with the reading. The only thing length costs you is response time, and this is measure in milliseconds at reasonable engine speeds.

By most reasonable interpretations of the instructions provided by the manufacturers of wideband O2 sensor kits, 4" from the head is too close. AEM, for instance, suggests that the sensor be located "at least 18 inches downstream from the exhaust port."


How's the weather where you are, incidentally?

flooding was pretty crazy here, i got lucky my neighberhood was fine.. I guess i will just leave it be.. Thanks joe

You should be fine with 4 inches of clearance.

+1
joes right, it'll be fine there. Think about it, when you go to the dyno they shove it in the end of the exhaust. If you go that close and it's too hot then you can use a spark plug extender to get out of the high heat. Just choose a place that's easy to change. :2cents:

That's a good idea with the extender bung, but will it still be accurate? My buddy has one like that but his has some sort of scoop inside...

So long as the bung does not permit outside air to enter the exhaust pipe, the accuracy of the sensor will be unchanged.

I see too close a lot. For instance a lot of the subie DPs have bungs on them. Usually they are not at the suggested angle or distance from the tubro. However, people are usually not smart enough to read the manual and they use said bung for their sensors. AEM has a very nice manual showing you the angle and distance you should be using.

http://www.schnitzracing.com/manuals/AEMWBK.pdf
This is the 30-4100 manual. Not sure if it changed with the 30-4110 and the new 4.9 sensor... but I will be able to tell you when my ships this month :)

On an NA car you typically want to as close after the collector as you can get it without effecting flow in the collector. The further from the engine the more lag between the sensor reading and the actual events in the engine. Like on a subaru the sensor the recommended length from the turbo is over 3 feet of pipe away from the exhaust port and theres very noticeable delay in the readings (which is why mine is right after the bellmouth since its still like 6 inches further than the innovative recommended spec away from the head).

You dont want to use the bung extender, if its too hot you need to make the copper heat sink that is detailed in the innovative manual, no idea if its in the AEM manual.

I run mine about 6" past the turbo. It's fine.

Same with my Innovate.

Heatsinks are the key. as Leafy says, make a copper heatsink and it's fine even when really close to the head/turbo.

So I wanted to research Leafy's theory here. It made sense but I wanted to know how heat could affect the readout, and what other unknowns may be out there. So i went directly to the source, AEM. We sell/suggest AEM over any other wideband gauge we have access to at the shop. I have owned a 30-4100 and i currently have paid for (pending delivery) the new 30-4110 with the 4.9 sensor. So i reached out to their tech team for their thoughts and possible testing on this...

- This is a copy of part of what i sent them for reference, not my whole email -
1. Do you have any data to support the claim that a sensor 36 inches from the turbo creates a noticable delay in data? Would moving the sensor closer provide a more real time reading? Is 36" too far from the engine for reliable realtime data?

2. Being closer will heat negativly affect the readings? Does the copper heat sink theory hold any weight to allow accurate readings?

Anything else that would be helpful to support the 36" from turbo suggest location.



AEM Tech Support-
There is a delay, noticeable is relative but yes... There is a delay. Moving the sensor closer would reduce the life of the sensor, but it would be worth the increase in accuracy. I have seen these copper heat sinks before, but we have not done direct testing for them. My personal thinking on them is they are a bit excessive and don't return any benefit. They will provide small amounts of heat dissipation/insulation, but from what my basic understanding of thermodynamics tells me is there is little to no benefit.

Basically; Moving it closer WILL lower the life, but will provide more accuracy.



I thought I would share.