What name or letters would i look for that are the mixture readings in the ECU?
What name or letters would i look for that are the mixture readings in the ECU?
What name or letters would i look for that are the mixture readings in the ECU?
It's usually termed lambda equivalence ratio or equivalence ratio in most OBD scan tools.
There may also be a command equivalence ratio which is the mixture that the engine ECU is aiming to achieve for the given engine load conditions, for most off-boost conditions this is usually very close to 1 (which is 14,7:1 the chemically correct 'stochiometric' ratio for petrol), for high engine load conditions with mixture enrichment this may fall to 0.8 (around 11.7:1).
I'm sorry I overlooked your post #20,
If lambda equivalence ratios are close to 1 that's a good sign there isn't much wrong with the wide band probes.. and there won't be any misfires if that reading is close to 1 either because a misfire reflects in a lean reading from lambda probes so that's another good sign.
Adding to what I said in my last post, similar concept to not letting the engine idle for too long before going for a drive.. when driving don't keep to the same engine load for too long either. All the time you're at a particular engine load the long term fuel trims are steered by short term trims but if short term trims reach an extreme (say -25% or +25%) the steering of long term trims stops and then if mixture is lean or rich you get a fuel trim error code. So one way of allowing long term fuel trims to be learned whilst not causing an error code is to drive only briefly at a certain engine load if short term trims are at an extreme at that engine load. E.g. Say you're doing 60mph at (I forget the gearing) 2500rpm in a certain gear, lambda equivalence seems good (close to 1 or command lambda) and short term trims aren't at an extreme then you could drive forever at that engine load and it won't bring a code on. But if you're at that engine load and equivalence ratio isn't close to 1 or close to command lambda and the short term trim reaches an extreme you'll get the error code... But before you got the error code the long term trim will have been steered a bit and if you'd changed engine loads before the error code was triggered you could go back to the same engine load and the fuel trim for that would be steered in the right direction a bit more. With enough little steers of the long term fuel trim for that particular engine load eventually the short term trims wouldn't need to reach an extreme to correct mixture and then you could drive at that engine load without triggering a MIL code.
I have 3 options
commanded equivalence ratio
equivalence ratio (lambda), bank 1 sensor 1 (wide range oxygen sensor)
equivalence ratio (lambda), bank 2 sensor 1 (wide range oxygen sensor)
lambda bank 1
lambda bank 2
I have graphed all of these for a 4 min period at cruise
nothing below .988 or above .1.002 unless accelerating or decelerating
nothing jumps out as out of spec or serously delayed, nice sawtooth patterns, some variation in length probably due to load
More info found in the manual
Since this is a North American vehicle it has a DTML pump on the evap system. It checks the evap system from the fuel tank to the purge valve. Any leak in that system would set off an evap code before/with a lean code.
The equivalence ratio readings are the mixture readings to use - bank1 S1 and bank2 S1. The voltage/current readings aren't really useful.
You've said your lambda equivalence readings at cruise remained close to 1, which is a good sign.
If you drive only during those (cruise) conditions do you get any error codes / MIL come on?
If you drive at other engine loads (with constant throttle position) does equivalence ration stay close to 1? Do you get any error codes / MIL come on under any of those conditions (if so what conditions e.g. constant 30mph very little throttle, 80mph constant throttle)?
Even if the equivalence ratio stays close to 1 you can still get fuelling error codes if fuel trims are at an extreme to achieve equivalence ratio close to 1. It would be a good idea to check LTFT B1, LTFT 2, STFT B1 and STFT B2 in various driving conditions. I would expect STFT's to swing by around +-12% during accelerator pedal movements (that's normal) but what are all those readings at constant throttle at various engine loads?
You previously mentioned fuel pressure readings (which I said seemed normal). The fuel systems on 4.2 supercharged Rangerovers reacts to fuel pressure readings - Say the fuel pressure at a certain engine load is supposed to be 450kpa, if the actual fuel pressure reading at that engine load is only 400kpa the system will react by increasing injector pulse length to compensate for the slightly lower fuel pressure than expected. On a conventional / old-skool fuel injected engine the need to adjust injector pulse duration because fuel pressure is low would be reflected in fuel trims but on this model vehicle the low fuel pressure probably won't be reflected in fuel trims because instead it reacts to the low fuel pressure by compensating the pulse length.. But the fuel pressure 'fiddle factor' can only work properly within a certain range of deviation from correct fuel pressure.. So it's important that the fuel pressure isn't too far from the norm across the entire engine load range or incorrect fuel pressure could cause OBD error codes / MIL light. The following sentence may be a bit too much info but if you want to understand the way it compensates for under/over fuel pressure - Behind the scenes the system compensates by comparing the square root of fuel pressure, so let's say (hypothetically) that the fuel map (and injector pulsed duration) for a certain engine load is set up to be correct for a fuel pressure of 500kpa above expected manifold pressure (the amount of fuel an injector will flow is effected by it's pulse length, fuel pressure and manifold pressure... fuel pressure minus manifold pressure because we have fuel pressure on the input side of the injector and manifold pressure on the output side of the injector). The square root aspect is a physics aspect - If we double the pressure of a liquid through a nozzle (or injector) it will flow 1.402x the amount of the original pressure. So if the map dictates an injector pulse length of (say) 10ms at a certain engine load for an expected fuel pressure of 500kpa, if the actual fuel pressure reading is only 250kpa (half of 500kpa) then we might expect the system to respond by adjusting the pulse length from 10ms to 14.02ms and for the fuel trims to be the same as they'd be if fuel pressure were 500kpa and injectors were being pulsed for 10ms.
Summing up, if you identify a certain condition / engine load that cause the MIL to come on - what are the equivalence ratio readings, fuel trim readings and fuel pressure reading at that engine load?
I don't have much to suggest that could be the underlying cause of the error codes at this point but assuming you haven't just let it idle after resetting an error code and/or fuel trims and assuming you're correct about no vacuum/boost leaks and the advanced evap system... if I had to hazard a guess it would be the MAF. But if it were mine before buying a MAF I'd be gathering the extra data I outlined above.
Sorry I have been busy logging data and dealing with glitches on my IID gap tool. I have been logging and correlating lambdas and fuel trim, o2 and lambda, fuel pressure and injector pulse width, etc.
I live rural. Its 10 min to the first stop, turn , or town in any direction, So I spend most of my time at cruise 100-110 kmh. I only go thru town on my way home and it has 3 stop lights and a timmies (coffee shop for non Canadians). 5 cars in a row is considered heavy traffic.
The mil code takes days to show up after the adaptive values are reset to zero. The p0171 lean bank1 is the only code I have. There are no other fault codes for the engine. When it does show up it is usually shortly after restarting once the truck is up to temp. I dont recall it ever showing up at cruise speed. It always seems to be 1-3 min after a restart. And oddly enough during cold weather. Last week it warmed up quickly to 27'C No code all week. Sunday night it dropped to 1'C, and we have been 5-7'C since. Tues aft on my way home thru town. Stop for coffee mil comes on about 2-3min after restart. (Last summer no code, Fall single digit weather mil every 3 or 4 start. Related? Probably not just hilarious.) I reset the mil but I dont remember if I cleared the adaptive values
Logging lambdas, very even only changes with throttle and load. Range 0.88 to 1.1 at extremes 0.88 under fairly hard accel and 1.1 on a downhill decel? Mostly .95 to 1.04
Kind of what we expect
Logging O2 sensors Bank1 reads about 0.5V higher than bank2 on both sensors, Bank1 pattern is saw tooth and bank2 is more of a wave. Other than pattern the trends are the same. Lambdas seem to follow this but slower wave motion
Logging fuel pressure and injector 1 & 2 pulse width Fuel pressure lags injector pulse changes (response time?) but does increase with longer pulse width and drops with shorter pulse width. At cruise a bit of a saw tooth pattern fluctuating about 5psi with the occasional larger fluctuation, Seems to change with load and throttle as it should.
Logging Fuel Trims they generally follow the same trends but there is a difference at cruise. The difference grows larger at higher RPMs but seems to return to very similar values at idle.
Im thinking I need to revisit the smoke test. My method. I placed a sheet of plastic film over the air filter in the breather and then put the cover back on. Smoke line was inserted into the intake breather via the rocker cover inlet. I had 15 psi on the smoke system and nothing came out. But this does not check the crankcase since the pressure thru the intake closes the pcv valve. I should have had smoke pouring out of the rocker cover inlet but there was none. There was also none coming out of the exhaust, I must have stuck it lucky and had no exhaust valves open. Or did i get something wrong?
Is it actually possible to be in any position without any exhaust valves at least partially open?
Yea i wondered that too but exhaust and intake on the same cyl have to be open. Since there is only a few degrees of valve overlap in 720 degrees I think that is possible.
Also i think there are some check valves in some of the vacuum lines that would prevent smoke from getting to certain areas ie the vacuum pump and the booster itself.
I think i goofed while doing my homework, I havent dealt with this stuff in so many years Ive forgotten a lot.
So I guess i get to do it again.
p73990: On that smoke test there are you quite sure you injected it on the 'right' side of the butterfly ? Similarly if you crank it very briefly (HT leads off) does any smoke exit out the tail ?
You say the O2 sensors are 0.5 volt different between banks: What are their actual voltages ?
Yes, you could have a leak on the intake manifold and/or an O2 sensor that is slightly 'off' (and its output is maybe temperature dependent too ?), although as a general point I am sure with some ECUs you will know that the related MIL Codes (especially with off/on marginal values) do often take a while to show up - as they have to persist for a certain number of Drive Cycles to actually register on the ECU ? And it all definitely makes these 'odd' faults much harder to resolve...
although as a general point I am sure with the GEMS you will know that the related MIL Codes (especially with off/on marginal values) do often take a while to show up
But it isn't a GEMS, it's an L322 Supercharged.....
Richard, meant to write some ECUs and so will correct it now !
GEMS can be 'the worse' though... ie. Codes that can take ages to manifest themselves properly are its speciality..... ?
and maybe disappear spontaneously too... ?!
The smoke test was done over winter inside my shop. I placed a screwdriver in the throttle body to make sure it was open, But i did not turn over the engine, I plan on redoing the smoke test tomorrow to see if i missed something, I can check the what the o2 voltages are tomorrow. I also have new plutgs to put in it. I doubt they are the issue as there are no misfires registered, I just dont know how old they are. Im considering swapping the post cat o2 sensors to the other side to see if any readings change.
Gilbertd You mentioned maf sensor logging, I have occasionally looked at the maf, It did increase and decrease with engine speed . From what i have read they generally work or dont, If you have to clean it to clear a code then its time to replace it. Am i wrong on this or is there something specific to look for? ie spike or dropout I did clean it last year when i got the truck and have specifically not cleaned it since.
@p73990 I've got the same on my 4.2 supercharger it running lean on bank 1 and 2 done same as you did smoke test checked all pipes be interesting to see what your problem is causing the engine management light to come on I can turn mine off but comes back on after it covered a hundred Miles
@karlos01, don't forget that you are running on LPG and if the LPG calibration is slightly out that will also cause too lean (or too rich) codes so may be a totally different problem.
Im batting zero at this point, All i can say start at the beginning. Vacuum leaks first, then try unloading a can of WD 40 all over the motor when it is idling, Listen for changes in rpm while doing this, Then start logging data on all the stuff in this thread, fuel pressures, lambda and look for abnormal reading or patterns. Its cheap enough you could try some injector cleaner. Maybe log fuel pressure before and after cleaner. The problem we both have is that we dont know what the normal readings and patterns should be. In my case since I only get one code, I can reference the other bank,
Sorry Im not much help.
So I revisited the smoke test today. I used the same method as described earlier. Apparently my memory is not that accurate, Using that method does check the crank case but does not check the brake booster or the vacuum pump due to check valves. However if you put the smoke into the brake booster line instead you get the same results and it is easier to seal with the correct size hose.
My results were the same as the last test. The was no smoke from the exhaust, meaning no intake and exhaust valves open in the same cylinder. No vacuum leaks showed up. I did pulled the dipstick and open the oil fill cap to make sure I had smoke in the crankcase. I tested the booster and vacuum pump separately. I pulled the hose going to the fuel pressure sender which comes from the boost side of the supercharger. I found absolutely no vacuum leaks. I did find the rubber adapter used on the fuel pressure sender between the hose and sender, was somewhat soft and spongy. It did not show any leaks though. I installed two zip ties on this adapter as a precaution.
Interestingly enough though, I did discover that at roughly 5 psi the air filter housing stretches enough to let the smoke out where the halves meet. LOL
I then took it for a quick drive to warm it up and get a drive cycle before replacing the spark plugs. Here is where the fun really starts.
When I removed the plugs the three plugs that looked lean a few week ago, they are now quite a bit better. They were not perfect but most certainly not clean and white.
Im beginning to think that this actually could be related to something else like outside temp being at or below freezing causing shrinkage. Or maybe just poor gasoline.
I could be trying to figure this out in the cold weather this fall. (yah...NOT)
All I can do at this point is keep occasionally logging data to catch something, and/or wait for whatever causes the code to consistently show up again.
Ive got to say Im stunned with the response I have received on this issue. Especially since this is not a P38 or older model rover. Hats of every one that put in their 2 cents regardless if it helps or not. I am certainly not done with this but I will not throw parts at it in hopes of this going away. Unfortunately it looks like this is going to take a while.
Im still trying to make pics available, Imagur does not seem to like me for some reason
If anyone has any thoughts on something Ive missed, Im all ears
I will keep you all posted if I find something
Pics of plugs and screen shots of graphs
Let me know if this doesnt work
Just tried to edit the post so the pics appear but Google drive does something odd with the url so they don't show up.....
p73990: Still not sure why smoke is not coming out the tail... !
Have you tried pressure tests on the pots since the plugs look better now (ie. except for one) ?
Again any slight leaks could certainly be temperature-related yes
Looking forward to seeing the pictures..