ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
Jun 20, 2016 22:10:22 GMT
|
I really should go out and get a lambda sensor and not be so tight but I also don't fancy changing things when I don't need to and at the same time get a bit of insight : AFAIK the sensors are around £30-40 for the single wire sensor. ATM my Mazda MX-5 will not pass an MOT on emissions Over the years the emissions have been as follows: | MOT Standard | 2016 | 2015 | 2014 | 2013 | | CO=3.5% | 4.73% | 3.71% | 3.10% | 0.33% | | HC=1200ppm | 503ppm | 283ppm | 317ppm | 256ppm |
So it seems the car *somehow* passed last year but it did indeed fail the first time around. The MPG of the car seems fine @low to mid 30s MPG, but it does smell rich, like an old 70s car if I am honest. In that time the car has covered around 6,000 miles. When I got it, the oil and plugs looked like they needed a service, so I treated the car to a full service with new spark plugs (NGK BKR6E-11s), an air filter, oil & oil filter and fuel filter, which was also quite new looking. It's running a pattern 82 degree thermostat (the genuine ones are 88) but I cannot see a few degrees making all of the difference to the emissions. From what I can tell the car appears to have had a decat fitted since 2008. Also, the car is running a pattern O2 sensor which is joined to the original (and now gone) Mazda sensor connector with a crimp joiner, which is not the tightest of things. I plan to solder the connector one of the days and see what happens in addition to doing the LED 'diagnostic' tester to see what that brings up, as well as checking out some of the other sensor values. Am I correct in saying that if the car is held at a cetain certain RPM (say 2.5kRPM) the sensor's voltage will oscillate between 0.2 & 0.8V as the car's ECU continuously attempts to keep the mixture stoichiometric at the magical 14.7:1 ratio? I am trying to get hold of an oscilloscope to see if I can check the sensor out, but doing a basic check with a Fluke Multimeter the voltage seemed to stay at around 0.1V around idle and not really move from there. With it staying at 2.5kRPM for some time the value shifted to 0.25-0.3V, but did not flicker any higher. It did seem to change if I blipped the throttle, but again, below around 0.25, which if I am correct would "tell" the ECU that the car is running a little on the lean side. As said, I never saw it even approach to 0.8V at all, since IIRC Lamdba sensors generate a voltage depending on the difference between the exhaust oxygen content at the outside world? While I hunt around for an oscilloscope is there something else it could be and that I am missing or is it looking like a likely culprit? The TPS was a little off and the car used to hunt slowly for around 5 seconds if I brought the car to a halt; this has now been sorted after I adjusted the TPS (after the emissions check). |
| |
Last Edit: Jun 21, 2016 6:07:41 GMT by ChasR
|
|
|
|
rbs
Part of things
Posts: 64 
|
|
|
|
|
Tight for time for a proper answer but if the CAT has been removed it will fail the MOT no matter what he emissions are.
|
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
Jun 21, 2016 17:05:56 GMT
|
Tight for time for a proper answer but if the CAT has been removed it will fail the MOT no matter what he emissions are. Cheers for the response. I did consider this, and I am not ruling out buying a car, partially to get the exhaust quieter as well, but I am not sure. This car has a pre-cat emissions limit of the following being a pre-92 car (1990 MY is the car): CO<=3.5% HC<=1200ppm The same limit for cars running K-Jetronic, carbs and many other cars made in the 80s. My Porsche 944 Turbo which was running a decat and with 300BHP managed to pass under the same limits and was a year newer (1991). Besides, surely with a cat it would still be very close to the threshold of 3.5% I'd have thought ; it's at 4.7% now. If I am missing the point I'm happy to listen  . |
| |
Last Edit: Jun 21, 2016 17:08:02 GMT by ChasR
|
|
froggy
Posted a lot
Posts: 1,099
|
|
Jun 21, 2016 17:34:03 GMT
|
|
At a steady state 2.5k a single wire lambda will drift up to .7 or 8 so yours is shot by the sound of it as long as there is no leak between the mass air Meter and inlet . With no cat it should sit at around .5 co in closed loop
|
| |
|
|
mk2cossie
Club Retro Rides Member
Posts: 2,961
Club RR Member Number: 77
|
|
|
|
From past experience, the lambda sensors that you splice in to the loom are utter garbage. I'd try and get a ntk sensor for it and it should all be ok. The voltage should flucuate between .1 and .9 of a volt if it's working well, maybe a bit slow if you can pick it up on a multimeter though  |
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
Jun 22, 2016 18:39:51 GMT
|
I think you peeps might be right mk2cossie and froggy. However, I may be getting my hands on an AFR gauge so will check that with the changes made. Today I did the diagnostic check on the car with the LED which flashed two codes: 15 – Oxygen Sensor (output low) 17 – Oxygen Sensor (output not changing) It is certainly pointing to the Lambda but of course I don't know if the sensor has previously been disconnected. Before I go out and splash out on one I did a few things: -Briefly checked around for leaks - the PCV I could suck on but blow in at a slow rate but I could not blow into the engine). It's hose was a little loose as was the clip so I replaced this with a jubilee clip. -Soldered the connection for the 02 sensor -Cleared the codes. When I came to starting it the car initially ran on 3, something it has never done, even if left for a few days. I shall check the codes again once I have driven the car about and see what happens. |
| |
|
|
|
|
|
|
|
|
Their not supposed to be soldered, I found this out when searching a similar problem with my E36, apparently the lambda uses the oxygen gap in the cable as the comparator. I thought this was utter tripe but when I did my compex course one of the regs is to create a barrier gland to prevent the oxygen in the cable air gap fuelling the explosion. Worth a good Google as this guy had pictures of a Bosch course he's been on to back it up, in fairness it does make sense as well as otherwise where can a sealed unit that sits directly in an exhaust get the oxygen to compare it too?
Believe me I also soldered them in on my beemer as I thought it looked a lot neater than the original, I then found that the 328 uses different types of lambda to the 325 which are really expensive, I forget the exact names but it was something like tabia and fabia but the 325 is a £70 replacement and the 328 something like £130 (x2).
Hope this helps.
Another quick thought is now you've lowered the engine temp could the system be thinking its a bit colder than it is and therefore injecting more fuel? Just some thoughts.
|
| |
|
|
Ryannn
Posted a lot
Posts: 2,421
|
|
|
|
|
I've had a couple of troublesome fuel injected cars. From experience, some will happily run with an aftermarket sensor but others, my mini and MR2 for example, either won't, or need a specific brand.
Have you asked any other MX owners? The MR2 club has members who know which will work and which won't.
The only other thing I could suggest would be to clean the air flow sensor. You can get electrical contact cleaner from Halfords for £5 and it made a massive difference to the running of my car!
|
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
Their not supposed to be soldered, I found this out when searching a similar problem with my E36, apparently the lambda uses the oxygen gap in the cable as the comparator. I thought this was utter tripe but when I did my compex course one of the regs is to create a barrier gland to prevent the oxygen in the cable air gap fuelling the explosion. Worth a good Google as this guy had pictures of a Bosch course he's been on to back it up, in fairness it does make sense as well as otherwise where can a sealed unit that sits directly in an exhaust get the oxygen to compare it too? Believe me I also soldered them in on my beemer as I thought it looked a lot neater than the original, I then found that the 328 uses different types of lambda to the 325 which are really expensive, I forget the exact names but it was something like tabia and fabia but the 325 is a £70 replacement and the 328 something like £130 (x2). Hope this helps. Another quick thought is now you've lowered the engine temp could the system be thinking its a bit colder than it is and therefore injecting more fuel? Just some thoughts. Thoughts are always good . It's always nice to cover the angles and see what I or someone else has missed. After all, it is the point of the threads. Anyway, I did read the same thing about the air gaps in the wiring. My lambda however is a single wire unheated unit, which uses the exhaust to earth it I assume. I wonder if there should be an earth strap attached to the exhaust... I did solder it in the end but it seems you may have been right. It's now bringing up one code: 17- Oxygen Sensor (output not changing) A kind RRer also lent me an AFR gauge with a wideband sensor. It shows 0.88 (around 13:1 AFR) out of one tailpipe and 1.04 (shy of 15 I think?) out of the other (it's a dual exit system), so it is clearly running rich. As for the cooler running I've since fitted the two stage genuine thermostat and cured the cam cover leak which seems to have done almost nothing for the AFR readings. The only red herring is at times the car will only run on 3 cylinders on start up sometimes, but it always goes to running on 4 after a few minutes. I have a spare set of injectors which I may give a go, but I'm wondering if I should get them in an ultrasonic bath first or try some injector cleaner/V-Power in the fuel tank - I've only ever run it on supermarket fuel but again I cannot see it being an issue. :/. I've attempted to obtain another sensor, which has been interesting, as detailed in my build thread. MX-5 Parts do them but they are not mentioning the brand of their aftermarket units, and I really don't want to be changing a bad part again (this sensor I took off has barely done 5,000 miles...). Their genuine item however is £108 which is a bit too much for me to stomach! I've now ordered an NTK sensor off Mister-Auto on eBay. Fingers crossed it improves things. |
| |
|
|
|
|
|
|
|
|
Sounds like your on the right track then, I thought I'd done such a lovely job on the beemer and wondered why bmw never did the same themselves 😂😂😂
Obviously not as clever as I thought I was!!!!
|
| |
|
|
|
|
Ryannn
Posted a lot
Posts: 2,421
|
|
|
|
The only red herring is at times the car will only run on 3 cylinders on start up sometimes, but it always goes to running on 4 after a few minutes. I have a spare set of injectors which I may give a go, but I'm wondering if I should get them in an ultrasonic bath first or try some injector cleaner/V-Power in the fuel tank - I've only ever run it on supermarket fuel but again I cannot see it being an issue. :/. My MR2 developed an intermittent misfire not long after I bought it. It would run rough but be cured sometimes by a blast on the motorway. Anyway, I changed the plugs, changed the injectors and even changed a coilpack, nothing fixed it. Since I cleaned the MAF sensor a few months ago, it hasn't uttered a splutter since! |
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
It's about time I updated this Sounds like your on the right track then, I thought I'd done such a lovely job on the beemer and wondered why bmw never did the same themselves 😂😂😂 Obviously not as clever as I thought I was!!!! You'd think so, but it seems this isn't the case. The new NTK sensor arrived today, well NOS. It fits the connector but the wiring is too short. I have since found out that my 4-2-1 manifold requires the harness extending (or the wire in my case). My last O2 sensor with a grand total of 4,000 miles on it has been extended previously with a soldered joint before. Clearly, this isn't a new problem I have come across. Like you, I have also read that O2 sensors should not be soldered, but crimped. I am looking into making a harness for the sensor so as to extend it. Here is question though. Can you buy the connector on the left from elsewhere? I know Bosch make a few but it would nice to get one like below:  I do have a couple of the O2 crimps but my old wire has been butchered a couple of times before with solder and iffy repairs etc. My other choice which is OTT is to get a wideband and see if it can emulate a narrowband signal. |
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
|
It's about time I updated this thread.
In the end I got an NTK sensor. However, I still had to extend the loom due to the manifold - it is further downstream due to the branches of the manifold (it's a 4-2-1 MX-5 Parts jobby).
I was considering making a sub loom since you can buy such a loom for the cars with a 4 wire sensor, but in the end I got some good quality crimps and some heatshrink. With that done I put the new sensor it. It still smells a little rich, but there is no error code as of yet after around 40ish miles.
Prior to changing it I did do an AFR check on the exhaust with the old sensor. The nearside tailpipe gave me an AFR of around 12.3:1 at idle but the offside was more like 15.1. Are the tests meant to take a combined value for a car with two tailpipes, as by anyone's standards that is quite a discrepancy; My M3 has no such difference across the tailpipes.
|
| |
|
|
Nathan
Club Retro Rides Member
Posts: 5,626
Club RR Member Number: 1
|
|
|
|
|
Normal issue for MX owners.
Even on the GT they mention every year it's a little high (cant fail though), pre-SC I bumped the timing up to 14 Degrees (common on MX's) and it made a small difference and enough to pass. Your's is a little higher than mine ever was but worth checking over the AFM/Timing etc chap.
|
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
Thankfully I do still have a spare AFM, so I could try testing both out potentially. We'll see what happens come the next MOT . I would have though advancing the timing would cause it go richer on the emissions, but it is worth a shot . |
| |
|
|
froggy
Posted a lot
Posts: 1,099
|
|
|
|
|
Single wire sensors are pretty slow when they're close to the exhaust port . The further away the worse they perform as the gases are cooler . When I test these early cat types I put the pedal pressor on and let it sit at 2800-3200 for a couple of minutes to get everything up to temp before putting the probe up . If it's all in good order it will be in closed loop and if not a bit more rpm will get the co and lambda down enough to pass .
|
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
Well, it's time for an update Single wire sensors are pretty slow when they're close to the exhaust port . The further away the worse they perform as the gases are cooler . When I test these early cat types I put the pedal pressor on and let it sit at 2800-3200 for a couple of minutes to get everything up to temp before putting the probe up . If it's all in good order it will be in closed loop and if not a bit more rpm will get the co and lambda down enough to pass . Well, it's time for another update. Cheers for that. I did wonder whether it being further downstream and unheated would affect things, and potentially the life of the sensor as well. I have some news however. The car seems to have done more than 100 miles on the new sensor ; well, I mean a few trips from Gaydown to Coventry. So far the error codes have yet to come back on diagnostic procedure . Maybe the latter is a placebo but it does seem to be slightly perkier if I am honest. The AFR reading however is a strange one now from the exhaust ports and hopefully the sensor is not broken, but the oxygen content reading in free air is the same as before and recalibrating it still gives the same reading! Previously one tailpipe was miles too rich! Now: -The Offside tailpipe can barely give a reading at all. To be fair, there are very little exhaust gasses present on this tailpipe compared to the nearside. -The nearside tailpipe as said has plenty of pressure there. With the engine warmed up that is now giving out a value of 1.17 on the lambda value; pretty lean! It could be that the other tailpipe is influencing this reading The engine doesn't appear to be pinking etc. and the MOT is on Wednesday. We shall see what happens. |
| |
|
|
froggy
Posted a lot
Posts: 1,099
|
|
|
|
|
If the rear pipes are divided around the back end one will pull air in unless they're perfectly balanced , I end up stuffing wet rags on one side to stop it affecting the emmisions results
|
| |
|
|
ChasR
RR Helper
motivation
Posts: 10,194
Club RR Member Number: 170
|
|
|
|
If the rear pipes are divided around the back end one will pull air in unless they're perfectly balanced , I end up stuffing wet rags on one side to stop it affecting the emmisions results That's what I figured . My BMW didn't seem to have the 4 pipes influence the readings massively but that could also have a strange design within the 'box . The MX-5 are split on the rear as shown:  |
| |
|
|
luckyseven
Posted a lot
Owning sneering dismissive pedantry since 1970
Posts: 3,839
Club RR Member Number: 45
|
|
|
|
|
Can I just pollute this thread with a (slightly) related question? If you can't solder the Lambda sensor wires, is there any way of shortening them? Mine has about three feet of cable coiled up in the engine bay when it only needs to be about five inches long to reach the wideband computer bit, and I'd love to shorten the wires. But it appears this will screw it? Could I just use crimp butt connectors? Any opinions?
|
| |
|
|
|
|
.
