MAF voltages - reading low and detonation
#31
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
At the beginning of the thread I said:
"Resistance in the loom from MAF sensor to ECU?
Dodgy ECU?"
But we did not get time to check them. I never detted when I saw it, but it was said to be detting on the rolling road BEFORE mapping. However, before doing anything to it it (ie on standard ECU) was running on the lean side for a standard car and hitting zones for ignition advance which were more suited to lower load zones - as such there was a lot of knock correction so that on a rolling road it could easily have detted.
After remapping we had it hitting more sensible fuelling and load zones, the owner is getting a knocklink for day to day monitoring and running it again on the rolling road. On his return we will see if we can find the cause, but an ECU fault seems unlikely. The voltages I gave were the output from DD. Another possibility is a leak that is causing air to be taken into the engine but not read by the MAF - candidates could be between the MAF sensor and the compressor wheel I suppose.
"Resistance in the loom from MAF sensor to ECU?
Dodgy ECU?"
But we did not get time to check them. I never detted when I saw it, but it was said to be detting on the rolling road BEFORE mapping. However, before doing anything to it it (ie on standard ECU) was running on the lean side for a standard car and hitting zones for ignition advance which were more suited to lower load zones - as such there was a lot of knock correction so that on a rolling road it could easily have detted.
After remapping we had it hitting more sensible fuelling and load zones, the owner is getting a knocklink for day to day monitoring and running it again on the rolling road. On his return we will see if we can find the cause, but an ECU fault seems unlikely. The voltages I gave were the output from DD. Another possibility is a leak that is causing air to be taken into the engine but not read by the MAF - candidates could be between the MAF sensor and the compressor wheel I suppose.
#32
yes, leak on intake, why didn't I think of that (don't answer please!)
The standard plastic hoses seem to need the clips tightening one hell of a lot, as the plastic isn't very grippy, and doesn't compress much at all.
Paul
The standard plastic hoses seem to need the clips tightening one hell of a lot, as the plastic isn't very grippy, and doesn't compress much at all.
Paul
#33
not very clued up on MAFs but how can a leak, cause det? it will cause loss of power or **** boost control.
think about it, the maf reads the amount of air going in, and provides fueling accordinly. so if the leak is after the maf then the car will be runing less power and rich. unless the leak is before the maf. then it will be the induction
was the car runing lean on the rollers ? you know i don't believe in standard lambda sensor. mapping a car with it is own standard map sensor can be dangerous but that is my opinion. has any one actually checked it against a wide band. Pavlo please back me up on this one! let me say it again, the standard lambda senor reads rubbish after 3-4%co that is 13:1 afr anyone that will actually argues with this is actually disputing facts!
why does the problem has to be so technical why could not be a good old fashioned mechanical problem!
is there any faults showing up on the ecu?
have you put det cans on teh car to actually differentait if it actually knock or a broken valve? or spring?
any one doen a compression test?
have you established which cyclinder is knocking?
injectors
fuel pump
etc..
sam
think about it, the maf reads the amount of air going in, and provides fueling accordinly. so if the leak is after the maf then the car will be runing less power and rich. unless the leak is before the maf. then it will be the induction
was the car runing lean on the rollers ? you know i don't believe in standard lambda sensor. mapping a car with it is own standard map sensor can be dangerous but that is my opinion. has any one actually checked it against a wide band. Pavlo please back me up on this one! let me say it again, the standard lambda senor reads rubbish after 3-4%co that is 13:1 afr anyone that will actually argues with this is actually disputing facts!
why does the problem has to be so technical why could not be a good old fashioned mechanical problem!
is there any faults showing up on the ecu?
have you put det cans on teh car to actually differentait if it actually knock or a broken valve? or spring?
any one doen a compression test?
have you established which cyclinder is knocking?
injectors
fuel pump
etc..
sam
#34
moon phase?
I've yet to get the standard AND wb lambda in the car at the same time. Going to get an extra boss put on the down pipe.
but anyway, Sam, yes you are correct if the engine sucks air in between turbo and maf the ECU doesn't know about it.
John has a fair bit of experience of using the standard Lambda, but I wonder if there is some variance between sensors, cars, running temps that makes even half accurate readings difficult.
Paul
I've yet to get the standard AND wb lambda in the car at the same time. Going to get an extra boss put on the down pipe.
but anyway, Sam, yes you are correct if the engine sucks air in between turbo and maf the ECU doesn't know about it.
John has a fair bit of experience of using the standard Lambda, but I wonder if there is some variance between sensors, cars, running temps that makes even half accurate readings difficult.
Paul
#35
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
Sam, if your engine is taking in air that is unmetered by the MAF - ie a leak between the MAF and the compressor then the MAF will under-read leading to lean fuelling and advanced timing. On full boost the car was hitting load points mapped for very little boost. The ECU was pulling timing appropriately for the amount of boost but it doesn't keep pulling it forever. The amount the injectors were opening was inappropriate for the boost level. For whatever reason the MAF reading in the ECU is under-reading. This is the problem to persue IMHO.
When the lambda and MAF sensors are working correctly there is quite a good correlation between the targets you put in the map and what you get. It doesn't matter if you can't see the lambda voltage in this case - the fact that the load point is only 2/3 of the way across the map on full load shows the problem. Lambda voltages being lowish (but not startling the ECU is so rich to start with) just back it up.
Will try an AE801 Matt, thanks for the info.
[Edited by john banks - 8/23/2002 4:50:11 PM]
When the lambda and MAF sensors are working correctly there is quite a good correlation between the targets you put in the map and what you get. It doesn't matter if you can't see the lambda voltage in this case - the fact that the load point is only 2/3 of the way across the map on full load shows the problem. Lambda voltages being lowish (but not startling the ECU is so rich to start with) just back it up.
Will try an AE801 Matt, thanks for the info.
[Edited by john banks - 8/23/2002 4:50:11 PM]
#36
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
Matt, since I reflashed the ECU and the problem persists, if the AE800 ECU is the problem it is a hardware not a software/bug problem. The relash of 128K includes all the maps AND the code, and I use a later version of code - invariably AE802 since there are fewer known bugs compared with AE800/1.
Mapping using the lambda sensor dangerous, that is an opinion, but how many cars have blown up as a result of this? Since the mapping is based on AFR targets from MAF inputs and you know roughly what points should be hit on the map for a given spec of car, then I am hardly going to keep leaning it out chasing 12.6:1 if the lambda over-reads. However, on full load, lambda sensors tend to under-read and if they are failing under-read dramatically and do not respond. If their failure mode was to show 890 mV when they should be showing 850 mV (e.g 9% rather than true 5% CO) then there might be an issue. But since I aim for usually 9%+ for safety and cylinder cooling especially on these crap turbos then once I start getting below "B0" on the fuel maps caution sets in.
[Edited by john banks - 8/23/2002 4:59:15 PM]
Mapping using the lambda sensor dangerous, that is an opinion, but how many cars have blown up as a result of this? Since the mapping is based on AFR targets from MAF inputs and you know roughly what points should be hit on the map for a given spec of car, then I am hardly going to keep leaning it out chasing 12.6:1 if the lambda over-reads. However, on full load, lambda sensors tend to under-read and if they are failing under-read dramatically and do not respond. If their failure mode was to show 890 mV when they should be showing 850 mV (e.g 9% rather than true 5% CO) then there might be an issue. But since I aim for usually 9%+ for safety and cylinder cooling especially on these crap turbos then once I start getting below "B0" on the fuel maps caution sets in.
[Edited by john banks - 8/23/2002 4:59:15 PM]
#38
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
Too little. Pretty much as WREXY's car with the APS induction - low MAF readings for the boost, giving low injector duty cycles, advanced ignition, negative knock correction and a low lambda. Feels good to a point, until it pops Maybe on a track or a rolling road, or on a hot day with a boost controller is the sort of situation where this could occur?
#39
Scooby Regular
Join Date: Jul 2000
Location: Glasgow
Posts: 3,040
Likes: 0
Received 0 Likes
on
0 Posts
But, does the ECU, having been told by the lambda that the mixture is lean, not increase the fuelling, thus compensate to some extent the problem of any possible air leak after MAF and subsequently increase IDC's?
A
[Edited by AlanG - 8/23/2002 5:36:40 PM]
A
[Edited by AlanG - 8/23/2002 5:36:40 PM]
#40
Subaru Tuning Specialist
Join Date: Jun 2002
Location: 7.74 @179 mph 1/4 mile - road legal
Posts: 6,654
Likes: 0
Received 1 Like
on
1 Post
think about it, the maf reads the amount of air going in, and provides fueling accordinly. so if the leak is after the maf then the car will be runing less power and rich. unless the leak is before the maf. then it will be the induction
#44
Subaru Tuning Specialist
Join Date: Jun 2002
Location: 7.74 @179 mph 1/4 mile - road legal
Posts: 6,654
Likes: 0
Received 1 Like
on
1 Post
Yes but considering that the turbo is sucking along the induction system then the whole track needs to be at a pressure lower than atmospheric in order to create flow
Therefore any leaks will be into not out of the system
[Edited by Andy.F - 8/23/2002 5:46:22 PM]
Therefore any leaks will be into not out of the system
[Edited by Andy.F - 8/23/2002 5:46:22 PM]
#45
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
That, or an ECU or loom fault are the only possibilities I can think of.
Bob reckons there may be some input from the lambda on full load, but the effect is not massive. Not enough to sort a dodgy MAF reading.
Bob reckons there may be some input from the lambda on full load, but the effect is not massive. Not enough to sort a dodgy MAF reading.
#47
Scooby Regular
Join Date: Jul 2000
Location: Glasgow
Posts: 3,040
Likes: 0
Received 0 Likes
on
0 Posts
Going back to thread a bit.
Did the car det only on full throttle at certain rpm? or was it a specific rpm any throttle?
[Edited by AlanG - 8/23/2002 5:51:52 PM]
Did the car det only on full throttle at certain rpm? or was it a specific rpm any throttle?
[Edited by AlanG - 8/23/2002 5:51:52 PM]
#50
Scooby Regular
Thread Starter
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
The car was pulling out timing on full load clustered around 5000 RPM which got worse as the heatsoak worsened. These engines seem susceptible at the 4400, 4800 and 5200 RPM points on full load. This all fits with det on full load at 5000 RPM on the dyno.
For a leak to cause a MAF under-read on full boost it has to be in the area of the system that is pre compressor wheel - ie below atmospheric pressure I reckon. A leak after the turbo will cause a MAF over-read and run rich and idle funny, turbo overspeed etc.
For a leak to cause a MAF under-read on full boost it has to be in the area of the system that is pre compressor wheel - ie below atmospheric pressure I reckon. A leak after the turbo will cause a MAF over-read and run rich and idle funny, turbo overspeed etc.
#51
Subaru Tuning Specialist
Join Date: Jun 2002
Location: 7.74 @179 mph 1/4 mile - road legal
Posts: 6,654
Likes: 0
Received 1 Like
on
1 Post
lol @ Alan
These MAF's, do they have temperature compensation ? I can see how they measure flow by the changing resistance of a hot wire or film but depending on the air temp the actual mass of air will change for a similar cooling effect on the wire/film.......i think
These MAF's, do they have temperature compensation ? I can see how they measure flow by the changing resistance of a hot wire or film but depending on the air temp the actual mass of air will change for a similar cooling effect on the wire/film.......i think
#52
John has a fair bit of experience of using the standard Lambda, but I wonder if there is some variance between sensors, cars, running temps that makes even half accurate readings difficult.
there is no wondering about it here. lambda sensor are a consumable! they all vary from on car to another and they all vary in time. there is also no such a thing as expenerience using a standard lambda sensor to map a car. it is a cheap why of doing it. this is also a fact. if you check my posts we all did it, before JB even bought his scoob!
very common thing to use over in the scooby community for a strange reason. probably the scoobs are underpowered anyway. but once you are running high boost and high power, it is dangerous. reagardless who you are. no disrespect to all the gurus over this board.
could it be that the injectors are faulty. very common on subarus!i know of two that had this problem, one of which ended up in engine failure. that was runing maf. teh other one was runing mass and you did get det from on time to another. the problem was irratic and inconsistent.
CC
sam
party time, later guys
all the above is obviously my opinion, as i don't pretend to be an expert in any way or shape of form.
#58
you know what i meant andy yours excluded, bitch
i meant standard comes running low boost like yours . around 1 bar and even less for the earlier ones. i am runing my evo at 1.6-1.7bar not going to take chances with this.
actually i did at the begining. that is why i know what i know. i have done back to back comprasion and please don't give me that subaru lambda sensor is good and the evo is not. they are all made by the same fricken company
i meant standard comes running low boost like yours . around 1 bar and even less for the earlier ones. i am runing my evo at 1.6-1.7bar not going to take chances with this.
actually i did at the begining. that is why i know what i know. i have done back to back comprasion and please don't give me that subaru lambda sensor is good and the evo is not. they are all made by the same fricken company
#59
Subaru Tuning Specialist
Join Date: Jun 2002
Location: 7.74 @179 mph 1/4 mile - road legal
Posts: 6,654
Likes: 0
Received 1 Like
on
1 Post
Alan
I've got the boost cranked up to 1.7 bar for Star RR day to see what it could have done at Knockhill if I had needed to try
Are you going over for a look-see ?
I've got the boost cranked up to 1.7 bar for Star RR day to see what it could have done at Knockhill if I had needed to try
Are you going over for a look-see ?
#60
Subaru Tuning Specialist
Join Date: Jun 2002
Location: 7.74 @179 mph 1/4 mile - road legal
Posts: 6,654
Likes: 0
Received 1 Like
on
1 Post
Sorry to drag O/T but Sam, since my last run at Star, have things changed ? I seem to recall some difference in temperature compensation or something ? Are the current runs comparable with the runs from way back in Feb or whenever ?
PS Sam, I'm a bit dissapointed you don't bite anything like you used too
[Edited by Andy.F - 8/23/2002 6:36:06 PM]
PS Sam, I'm a bit dissapointed you don't bite anything like you used too
[Edited by Andy.F - 8/23/2002 6:36:06 PM]