O2 sensor question
 

I have been in the ongoing process of enhancing the engine control by adding various sensors and feedback systems to the DME/KLR. Posted in the KLR and WG threads. So now I am at the O2 sensor part. I have been using a WB from Bosch a LSU-4.2 for a while, the issue is that I have been using the existing sensor mounting bung upstream from the turbo. I realize with the pressure and heat prior to the turbo the life span of the sensor would be low, as it turns out after 10 years or so of sensors, it works out to about a sensor a year. Which while not ideal is acceptable, the problem is that the sensors fail in unexpected manners with the wrong readings/ slow response being the most troublesome.
So the question is this:
If I were to close of the original bung and put a new one past the turbo, just before the cat how much latency would this generate ?
I believe that the gas passing over the sensor should have current enough O2 info so that the DME is not out of step when it applies fuel trim. Is this assumption correct?
Thanks Sox.
P.s my test station

http://forums.pelicanparts.com/uploa...1718808711.jpg

I have the narrow band sensor in the stock location, controlling the DME as per stock.

I also have a WB sensor located on the downpipe, which I use to monitor A/F mixture while on the track.

It is kind of interesting to watch the mixture go from rich to lean when you let off the gas, and some other situations.

I know Vic Timpauer used this same location to control his ViPEC standalone setup, so I expect that you should be fine with that.

Cheers, Keith

https://i.postimg.cc/TYScyDjR/100_3940.jpg

https://i.postimg.cc/xTqRTRzx/100_3946.jpg

Thanks Keith,
just added info, I used a bung cap to plug the old sensor location, it leaked till I put a copper crush washer on the cap.
Sox

Keith I think, I am going to do what you did. Put the narrow band back in it's original spot. The WB's electronic don't quite seem to be up to the task of simulating a narrow band's output and the switching speed. I have been putting up with an uneven idle, which I have tracked to the WB not being able to simulate the narrow band sensor.
I do have another question, do you know how Vic routed his sensor cable? I have my WB electronics in the old cassette holder.
Thanks Sox

He has it like I do, with the dual pod on the A-pillar. Routed through behind the glove box.

Thanks, where did you route the cable into the cabin ?
Sox

BTW my install, I used a step lap joint clamp as I would like to be able to disassemble the exhaust at this junction. Also I always use exhaust seal compound inside the clamp to pipe seal.

http://forums.pelicanparts.com/uploa...1719675401.jpg

I ran this up the firewall, and into the grommet behind the glovebox, then behind the dash over to the DS A-pillar.

The grommet that the dme/klr cables come thru ?
Thanks Sox

No, there is an unused grommet on my car that I was able to use, I think. Look around the vacuum canister near the battery and you should be able to find it.

This was 15 years ago, hard to remember the details...

EDIT:
I took a look at what I did, and I used an unused grommet just below where the battery cables come through the firewall to get to the battery area.
https://i.postimg.cc/TYpFSF4Y/IMG-20...-200343241.jpg

https://i.postimg.cc/m2JJGjv3/IMG-20...-200353741.jpg

I then used the large grommet just behind the battery to get into the cabin area.
https://i.postimg.cc/59PDsbpR/IMG-20...426286-HDR.jpg

So in testing after my Cali jaunt, I have been testing the WB sensor response. It seems to be a huge amount of latency as the injectors react in about 22ms after the afs plate closes, but the afr shows cut off at about 825ms ??
See the two pics below. Am I missing some thing? at 3200 or so rpm it should not take the exhaust gases that long to get to the sensor in the turbo down pipe.

http://forums.pelicanparts.com/uploa...1743183426.JPG


http://forums.pelicanparts.com/uploa...1743183450.jpg

Upon more pondering, with a little assist from an adult beverage....
The fuel cut off occurs after the throttle idle switch engages and the the rmp is above 1900 or so.
The time it takes for the cut off to occur seems to vary quite a bit, from about 50ms to 330 ms. still not as much as I am seeing in the wb sensor's out put.
Sox

http://forums.pelicanparts.com/uploa...1743197076.JPG

Testing the WB O2 in the the position of the original O2 sensor (just before the turbo) versus where I have it now few feet downstream of the turbo, I can see a definite lag in the readings because of it being further away from the engine.
Sox

Yup conclusive,
Car systems lag time (processing, injectors etc.) is about 140 ms maximum.
O2 sensor response (Set to instantaneous when gas pocket hits the sensor) in original pre turbo spot is about 395 ms including the above system lag time.
O2 sensor response at post turbo spot where I have it is about 850 ms including the system lag.
So in my opinion 0.85 of second is to long a lag for it to be used in real time closed loop control.
The reading were all done at 3k rpm, the time measured from WOT to AFR dropping suddenly as the car went in enrichment mode. The error in measurement is +/- 33 ms.
If the above does not look right let me know.
Thanks Sox