Replacement Narrowband Lamda sensors.

[Wasson Will]

Has anyone else had issues with getting a replacement Narrowband Lamda sensor for their CA18DETS or other engine that may use the same sensor. Nissan no longer make them or seems to be able to supply.

Have had to replace mine for the first time recently and am having problems with replacements, the one fitted since I owned the car is an M18 thread shown on the left, no visible ID number just Japan written on it. Ordered the first replacement and the on on the right turned up which was an M12 thread a lucas LEB5004, so ordered and LEB123 which was a universal fit M18 thread that was supposed to be compatible. Had to cut the plug off the old one, happy that it was wired correctly. the engine seemed to run even worse with the new sensor. Spoke to SMPE technical who are now the distrubutors for lucas parts and they reckoned the universal Lamdas tended not to work well with jap cars, I assumed may the stioch voltage may be different. He came back saying the car should have been fitted originally with an M12 sensor and a spacer bush and came back to the first sensor I had purchased.

Have got a bush and installed this one instead, definetly better and you can see under certain conditions it does seem to react and you can see the voltage flicker either side of half a volt and the engine runs nicely on cruise at 14.7. Then randomly the vehicle will start to run incredibly lean, when you look at the log the lamda is still reading .3/.45/.55 volts whilst the fuel mixture is getting as lean as 19 to 1, the screen shot below is taken just after driving for several miles everything running seemingly well and then whilst cruising the it starts running incredibly lean. Cannot get my wideband to work with the log at the mo but that screenshot is taken whilst the AFR is around 18 to 1. Also ended buying another brand Fuelparts sensor as well which appears to be identical to the Lucas LEB5004 and behaves the same.

I can see from old logs with the original lamda it would swing almost instantly to 0v or 2/2.5v almost immediately either side of 14.7. These units seem very unresponsive below .5, have increased the Enrich voltage to .45 which seems to make it a little better.

My only other though is that the smaller probe in the bush may be slightly restricted in the down pipe body, the tip of the probe is about 15mm long I reckon with the bush and the thread on the manifold there is about 7mm protruding out into the elbow which is pretty much the entire slot on the probe exposed. Though this may be causing a lag or unpredicable reading but it is not just for a second or two that it will run lean it seems to sustain for prolonged periods on cruise when it occurs.

[Matt]

Had to use a universal narrowband sensor in my R34 (cut and replace plug) but not seen those issues, but I dont take it out enough to monitor that closely lately, but haven't seen any issue with below 0.5V. I'll see what PL says. Logs from your sensors you have tried would be interesting to see. The screenshot shows a blank log window?

[PL]

I usually recommend that people just get either genuine NTK or Bosch sensors. Never had an issue. Avoid the generic ones.

PL

[Wasson Will]

Yeah I haven't got the PLX wideband working with Nistune at the moment so the log would not show the afr reading anyway. Screenshot was for the reading in the Consult window. of .44 volts, during that period of running the afrs were fluctuating between 16/19 to 1 whilst the narrow band reading was around that voltage this was at light cruise after running fine under load and could feel the lack of response/power. Will try and get a log of the this bit is of fairly limited use without the wideband info tied in.

Yeah I wasn't specifically looking for a cheap sensor, were just the ones I was able to get hold of, will look for a Bosch sensorvanyway but seems unlikely that three of the units are all unfit for purpose. The two m12 ones although packaged as lucas and Fuelparts are both marked as made in Japan and look like they are the same item anyway so would have thought they would be Japanese OEM standard parts anyway. Hence why I was wandering if there may be a problem with the probe depth in the manifold though this seems incredibly unlikely especially as it appears to sustain for 10/15 seconds plus when it does happen. And the universal fit one that I used was physically size wise identical to the original.

[PL]

Was speaking to a mate who runs a workshop this week. He needed an O2 sensor desperately so just used a GM post-cat sensor on a Nissan engine. Reckons it worked great - if anything BETTER than Nissan OEM! Had to fit Nissan connector that's all.

Just sayin...

PL

[Wasson Will]

Still fighting the battle at the moment, just trimmed 3mm off the thread bushing to make sure the proximity of the tip is totally clear from obstruction as it appears to function okay some of the time then under very light cruise it seems to stop working. Wideband is still not working with the nistune software so is not included in the log but from 1m 17 in it is a short cruise where the burn is 16/17 to 1 peaking at 18 and the lambda barely reacts. Got any reference number or a car that the GM sensor was fitted to just in case?

Also make it extra strange on my computer the cursor was not tracking the reference cell on the fuel map whilst it ran, this has happened several time to me recently, does not appear at the time or in the log when you re run it. Am wandering if there is a compatability issue with my laptop and the Nistune software hence why I can't connect the wideband either.

[Matt]

Grab the latest version 1.3.29 released. There is a fix for intermittent map cursor not updating since 1.3.27 version

This is not related to your wideband problem. My understanding of that is that you are using the PLX iMFD device, but you have made your own adaptor up to connect to it. The Nistune software is not seeing correct data coming in from your device, so it is not the software itself (since I double checked this on my bench already, and also confirmed your data is not matching the iMFD protocol)

I would suggest getting the proper iMFD interface cable for the device, and also make sure you are selecting PLX imFD and nothing else in the wideband window, since two of your log files had a different wideband device selected (ALM)

[Wasson Will]

Well all round progress is good, still have Lambda issues but other than that all seems good, having to drive with the lambda switched off. Have Got the Wideband working again, seems the second com port being open was stopping it from connecting or saving my config settings. That crash log I sent you I went into the config meni and selected the IMFD options and then tried to connect it, when it failed and I went back into config the sensor shown was always aml wideband. Disabled the second com port and selected items in config and it connected straight away, which then crashed the connection to the ECU repeatedly... However uploaded the updated version of nistune and all good, been out doing some rough mapping for a couple of hours and barely an issue.

Had a second problem which I also thought was related to the lambda and that was it what occasionally, normally after some heavy throttle it would through fuel in to the point of flooding itself to the point where I would be looking to find some where to pull over. Sometimes it would cut out but then start and clear itself after a minute. Seems it was a dodgy pin on the MAF which I only discovered after dropping the bonnet to latch with the engine running, caused the engine to go from ticking over perfectly to pissing fuel in at such a rate it was ticking over so slowly I though it had stopped. Hasn't flagged a fault code a single time it has done it so must have been on of the hotwire connections maing the resistance reading way out.

The Running lean on cruise still continues, have tried both the M12 Lambdas again both the same , will try the M18 universal fit again for better manifold pentration now I seem to have all other issues ironed out.

The log shows the response of the Lambda seems a lot less reactive on the leans side than the original one I had fitted from old logs I have. This was the log from after I shortened the threaded bush for the sensor as I was worried it was getting shielded under low cruise causing the issue. Thought it was running a little better but from 7.25m in I maintain cruise conditions for the majority of the log and the narrowband is way off...

[Wasson Will]

Tried running with the Lucas LEB 123 universal fit lambda again and yes it is way off, ironically seems to react in the opposite way to the way it should... My assumption is that the polarity on the heater wires is irrelevant on these as they are both the same colour so that would not be a source for a reverse signal. Funny that I have done a rough MAP that is pretty close to spot on and it runs horribly as soon as you turn the sensor on. The reverse reading causes the short term fuel trims to max out straight away as it gets confused by the reverse readings. Anyone got a part number for any narrow band lambda that they have successfully fitted to a type 1?? Given up on these 3 replacements now...

[Matt]

uploaded the updated version of nistune and all good

Okay good to hear the wideband side of things is sorted. If you have one device grab the COM port then the other selection cannot use it

dodgy pin on the MAF which I only discovered after dropping the bonnet

That explains more of the dramas. Would trigger fault codes potentially if it went out of range

Around 5:30 we see the STFT follow the O2 sensor voltage so it is switching. Bit on the leaner side, but that might be a discrepency with the sensor or placement between WB/narrow band. It seems to be responsive

There was a bit amount of fuel (10's) on the wideband for no reason. What is going on there 5:27, 5:29? No change with injection time at these points, so seems to be external

STFT I see is running away. Wait for the release on friday with LTFT for CA18DET to see that that does. I think three people on here are waiting for that now, so glad I found it