Fooling the ECU to burn lean...
 

Is there a simple way to fool my ECU into reducing the amount of fuel it tells the injectors to pass? Sort of a flip of the switch 'lean burn' mode?

Here's how I would use this lean burn mode, the goal is use less fuel:

I'm in top gear, the road is flat, the engine is humming along at 2800rpm in closed loop mode and under a constant load while traveling at 65mph. I'd flip a switch, that puts the engine in this 'lean burn mode', maybe an LED on the dash lights up (to remind me to switch it off while climbing a grade).


I would likely need to experiment to find what my engine likes, how much fuel I could take away under 'lean burn' trim. Maybe I would read voltage from one or both existing o2 sensors, maybe I would add a stand alone cylinder head temp probe under #2 spark plug and gage on the dash (to keep from melting the thing down).

My '96 (3) Cylinder TB injected engine is basically stock with (2) 4-wire O2 sensors (maybe the ECU averages between the two?). The car retains it's stock cat, it needs to pass California emmisions every two years. I am no electronics guy, I only know a little about the way EFI works, and have only used O2 sensors to keep an eye on carb jetting on a race car... a single wire narrow band that gave only marginally useful feedback (just another tuning tool like reading spark plugs). I can wire, solder, etc, but switching over to a Megasqirt system is out of the question.

Just brain storming here... please chime in with ideas and concerns, I'm thick skinned (hopefully not thick headed):cool:

Welcome,
metromizer, there is this thread. Also i have not searched the site for keyword, leanburn yet.


http://ecomodder.com/forum/showthrea...ules-2237.html

I have in the past made a little 2 channel box that used two 3v batteries and a simple variable voltage divider(megaohms, batteries last longer) to add 0.1 or 0.2 volts to the signal from the o2 sensors to the ECU. i rigged it up so it was switchable to be in series with the 02 signal or out (02 direct to the ECU). then i monitored the actual voltage from the 02 to see what really was happening. The best improvement i ever got was about 10%. (2 mpg) The ECU likes to see the 02 voltage cross over the .4v to .5v range every so often or it will turn on check engine light, so there is a limit as to how far you can go. before a check light. Some engines get a little better mileage around 16.5 to 1.(ref.johnnygrey)

another test i did is ref. in the above thread. using a zero to 10 ohm rheostat to drop the voltage to the injectors until the ECU can no longer lengthen pulse at which time the engine leans out. with this method i run open loop and unplug the o2 sensor and discount the engine check light. i have continuous read out on the o2 to avoid going below .25volts. obviously this brute force manual mode, your mileage may vary. (poor man's rheostat Craft shops sell heavy Ni-Chrome wire)

You'd probably need to adjust the O2 sensor's output so that the ECU thinks its running stoich when its actually running lean.

Here's some interesting info:

http://www.civicvx.com/lean_burn_info.html

There is a circuit that can be utilised to adjust the O2 sensor output, but I can't find the link.

Edit - Found it http://www.alternative-energy-resour...tallation.html

And available from the designer here http://www.eagle-research.com/store/...products_id=16

Just because you can get it to burn lean doesn't mean you'll save any gas though, and that's the trick. Cars like the VX, HX, and Insight used vtec-e and a host of other things to make lean burn efficient...normal cars might not like it, :p

How well does/did it work John?

Voltage divider worked good on a 5.0 HO mustang.(batteries lasted for weeks) 10% improvement
The rheostat is still on my 5.0 F250 don't have any improvement data because my pulse width mileage computer gets very confused when the pulse width goes to 80% and the injector still wouldn't open. But i have range enough range to stall the engine so i am sure lean burn is in between there somewhere. It is difficult to hurt these old cast iron engines at low throttle settings. never had any problems with detonation.:D

That answers my question ;)

metromizer -

Welcome to EM! Here's mine (I flip a switch just like you said) :

My Eagle Research EFIE
http://ecomodder.com/forum/showthrea...efie-1308.html

I don't use it much except for testing. I don't want to make it part of my overall system until I can verify emissions compliance.

CarloSW2

Because all part throttle, never had any problems with detonation.:D

The dual-O2 sensors is part of the OBD-II scheme to diagnose the proper functioning of the catalytic converter. One of the O2 sensors is upstream to detect pre-cat exhaust oxygen content and should switch back/forth past stoichiometric to let the cat function properly. The downstream O2 sensor is to detect that the cat is functioning properly and should not switch, but read constant...cant' remember which voltage. I have heard that some folks "fool" this sensor by putting it in an empty coke bottle with just air inside it to stop a MIL lamp with a bad cat.

I don't know much about Metros, but if you don't have a knock sensor that will automatically pull back timing if your engine pings you might get some gains by doing what you propose. If any timing gets pulled out to fight ping you'll probably not see much benefit. You also don't want to run lean-burn for too long because you may cause damage to your catalytic converter. The cats require the switching between lean/rich to properly function as both an oxidation and reduction catalyst. The cat needs a lean mixture to oxidize HC and CO, but then needs a rich mixture to reduce NOx. Without running rich, your cat can't reduce NOx so not only do you emit more NOx, but the cat heats up and can become brittle and crumble from only oxidizing HC/CO which produces heat.

I built an O2 modifier circuit from a document floating around called D17.pdf. I tried it in both cars (the Celeb and an 89 6000 with 2.8) and no matter what I did, mileage went down a little when leaning a little, and went down a lot when leaning a lot. Perhaps it would be better suited for applications when part of the gas is being replaced with something else. It also seems that the inline battery method might be a better idea.

I wouldnt go changing the o2 voltage by force. Narrow band stock o2s are only accurate at .5 volts(some sources say .45) and that is because they act like on off switches. Higher than .5 means rich, lower means lean. When you change the voltage from .5 to .6 it just switches at .6 causing the car to constantly think it is running rich and constantly build negative fuel trims.

O2 sensors are used in a closed loop feedback system to generate these fuel trims. Most cars have short term trims which run directly off the sensor reading in real time, and long term trims which are based on historical data that your ecu stores and are pretty much an average of your short term trims. You can log these fuel trims using any obd2 logger and you will always get the best economy/performance/drivability when they all average out to zero for every load/rpm. This can only be achieved with some piggyback tuning as there are variances manufactured into the car.

The only way I would suggest altering your cars idea of Stoich or the target closed loop afr is to get an Innovate LC1 wideband controller which has 2 programmable analog outputs. You can set one to read .1 volt for anything leaner than say 15.5 and .9 volts for anything richer than say 14.9 this way your computer will see a switching voltage of .5 and you can accurately tune what the appropriate switching afr is. Also, you can mount a cheap 25$ afr guage and program the LC1s secondary output to be say 1 volt at 12:1 and 0 volts at say 16:1 and get an idea of where your car really runs and also monitor your open loop operation to make sure it doesnt go too lean.

Your ecu bases its open loop operation on the closed loop developed fuel trims. It is basically guessing what ammount of fuel will be necessary to achieve a safe afr under heavy loads. If you are at wide open throttle(WOT at 2000 rpm is like open 30% before someone says you will never save fuel at WOT, you also would never make it up the hill at 2000rpm without going to WOT) going up a steep hill your engine wont last very long running a 15:1 afr. Be careful how lean you run in closed loop because it will have and effect on your open loop operation.

BTW... Lean burn engines tend to have staggered intake cam profiles within the cylinder so if you have 4 valves per cylinder you might be able to get custom ground lean burn cams. Wiki search the MVV ( Mitsubishi Verticle Vortex ) engine for more info. BTW most cars do have a slightly staggered intake cam profile anyway so we still might get to burn leaner efficiently without a set of cams.

Holy cow! The wealth of information here <on this forum> is nothing short of incredible! Thanks to those who have welcomed me, and a huge thanks to everyone who spent the time putting together very informative and understandable replies.

As I said, I don't know much about EFI systems (I appreciate your patience), all of my past tuning has been with carburators and mechanical fuel injection (MFI) systems, but those were WOT drag race applications, that have little in common with with what I'm doing.

let me see if I have this right; Narrow band O2 sensors used in OEM's help the computer learn what fuel map to employ the 'next time', but that could vary from ECU brand to ECU. I fully understand that pulling fuel away and burning a little lean isn't going to change the basic design (limitions) of my 3 cylinder Suzuki, but employing a few complimenatry changes might get me to my goal 50mpg at 65mph.

I have a head and cam change in the works, used stuff I will have reconditioned (taken off a Metro XFi) I'm contracting out for a little port match, bowl, and valve work by a proffesional head porter I know, I might bump the compression ratio up a little and file in some Singh grooves while I'm at it, for those who fiddle with Suzuki's.

From my race tuning I know that typically most race cars are running way too rich, safe, but rich and lower on power than they otherwise could be. I am wondering if my little Metro isn't also running a little on the rich (safe) side by design, to accomodate poor/uneducated driving technique, hot weather and marginal fuel, as to keep warrenty repairs at a minimum. I wonder if I could manually trim back the amount of fuel the engine gets under ceratin conditions, which BTW is where this commuter spends most of it's operating time, 65mph for about 1 hour per day. If I could reduce fuel consumption by 10% by trimming fuel, the excersize would be a success.

Catalitic converted overheat? noted... those things aren't cheap to replace.

Carlos: Was your O2 circuit 'voltage adder' ever successful on your car?

metromizer -

I think it is because I can *see* the A/F ratio creep up when I drive, but you need to read my whacky explanation in my thread. It doesn't settle at a new A/F ratio. It climbs up above 14.7, and then jumps back. I think this is because the lean signal from the second 02 sensor triggers logic in the ECU/PCM to keep the cat healthy (that's a good thing).

Maybe I will make a new YouTube demonstrating the behavior I am describing.

CarloSW2

I have a 98 civic ex. Does anyone know if it's possible to convert it to a vx without swapping the whole motor? I'm wondering if it's possible to simply replace some of the current motors components and then reprogram the computer so that it has lean burn?

Bingo. Unless you guys are concerned only with your wallet, these lean burn systems aren't good in a holistic sense. (As explained in the post below this one)

The vx engine, d15z1, is Obd 1 so it may pose more of a problem than its worth.

The hx engine, d16y5, is what you want to get parts from. You should be able to just put the y5 head on your engine and wire in the 5 wire 02 and add a d16y5 ecu. I think you also may need the y5 intake manifold as it has the appropriate EGR stuff. Dont forget the LAF sensor also, I just priced a new one for my engine because i was curious and my cost was $550:eek:


So yeah, its much easier to just swap in a whole engine. Its really not as hard as you may think, a good weekend of work is all it takes. Also Vx and Hx swaps go pretty cheap nowadays, my coworker just picked up a vx engine and tranny for $200.

Does anyone know if the Honda engine K20a3 uses leanburn? I know it utilized the same vtec-e technology as vx and hx engine but I havent found any info about lean burn. Using K-pro tuning software I'm sure you could lean it out quite a bit pretty effectively.

The usa civic si models also came with a dual stage intake manifold, really neat stuff

Atmospheric/Barometric Sensor
 

It is so much easier if your car has a atmospheric or barometric pressure sensor.

This sensor, depending on what your particular manufacturer calls it, controls the fuel mixture by sensing the ambient pressure at what ever altitude you are driving. The higher the altitude the leaner the fuel mixture.

All you do is put a 25k potentiometer switch on it and turn it down till the idle speed changes. All you are doing is telling the ECU you are driving at a higher altitude than you really are. There are no problems with warning lights or ecu problems this way.

Just be cautious you don't do to far or you can burn pistons or valves due to higher cylinder temperatures.

I have an old 89 Civic Wagon 5 sp, I would like to modify it and run it slightly lean. I have 3 pin PA sensor, One end/centre pin measure about 3K, other end/centre pin measure about 5K, Would you tell me more about how that how I connect those potentiometer?

Thanks

I was thinking the same thing. Until you can go into the ecu and tell it what afr to run at, just lie to it about the load it's under. It will go to another *lighter* part of the fuel map and put in less fuel. Just be careful...

I think a guy could be plenty careful if he used stand alone Exhaust Gas Temp (EGT) monitoring and data logging. It's the most 'real time' combustion temp sensing I know about.

But if pepole have tried it and had no better FE as a result, why bother?

3 pin connector hook up...
 

All I know to tell you is that one of the wires is 12v one of the other wires is somewhere around 2.75-3v. You want the lower voltage one. The other wire is the ground. You will have to use a volt meter to check them out. Remember to turn the key on or you may be scratching your head for awhile.

Thanks justpassntime:

I will spend little more time to find those wires and measure and see that it will work or not.

I know nothing, I may be wrong, I think that If car at high atmospheric, Air contains less Oxygen, So ECU picks up this condition and reduce fuels and it causes slightly lean burn.

I will let you know It works for me or not.

Thanks anyway

For my car, 1.5L 89 Honda Civic wagon, I saw a sample of Fuel map of my car, It seems that we can fool ECU to think that " Lighter load" by offset MAP sensor resistor value by resistor. It will offset fuel map easily too.

Of course, If we can change ECU fuel map easily, It is a better choice, We can fine tune is for best fuel map for fuel economy VS power.

Both Map or Atmospheric sesor mod are dirt cheap, below $1, ECU fuel map modify cost more money and it needs time to learn this editing program, That fuel map.... better but a lots harder...

Well, I find that SuperMid, It looks like that it is hard to get it and not cheap, and my car only worth few hundred dollars.....

From my understanding, SuperMid is using a ATmel Micro controller to check car sensors for speed, Amp, enginee speed, Fuel injector duration Vs Speed for fuel consumtion, SuperMid is designed really for Hybrid car....

diesel john

i like your idea about the battery voltage divider circuit for adding series voltage to sensors. it has given me ideas. :)

according to wiki, lean burn is beneficial because it reduces losses due to throttling the air and instead replaces this speed control method by leaning out the fuel. so for ideal efficiency you want to operate at WOT and control engine speed buy changing the air/fuel mixture.

we could use a variable resistor on the MAF sensor. opperation would be normal under closed loop control where lambda sensor "is god". at WOT, open loop control, the MAF is used and a variable resistor here would allow us to control the mixture.

my MAF is broken. it has lost calibration and the voltage is way too low so i am getting lack of power at WOT, too lean and the engine stalls. i am going to put a battery with voltage divider circuit in series with my broken MAF to try this same principle.

Lean burn, lean burn, lean burn...

Why havent yall thought about controlling the EGR valve instead? Or placing a second one to increase the flow during low load operations... ?

It practically does the same thing as lean burn, but more safely, does not increase emissions (actually decreases it), acts as a WAI, etc etc etc.

Don't get me wrong, I have thought about doing lean burn, but its not that simple since the engine has so many variables into controlling AFR. Instead, just keep it at 14.7 but add less oxygen (exhaust), AND it also keeps the CAT happy.

Thoughts? I am about to go either route, but I am seriously thinking more and more about doing more EGR (mainly because of reduced emissions)

basslover911 -

This is a cool idea. Different/bigger/multiple EGRs? I hadn't heard this before. What are the issues in implementation?

CarloSW2

I really don't think any. Heck if you inject more than necessary exhaust gas I think the worst that could happen would be a slight overheating and the engine dying (since there is no oxygen to burn) This, however, would not destroy the engine- start it back up and your set.

I would think injecting it into the air intake tube instead of inside the intake manifold would be best. That way you also open up more the throttle due to the denser air (hotter exhaust gas temperature) and decrease throttle losses- another advantage that Lean burn doesn't necessarily give you.

if you want dynamics like that, natural carb is the only way. ALL fuel injection as an action waiting for an action, it will never ever in the fastest mhz ECU in the world catch up to a natural carbs ability. You want fuel mileage? stop the switches, go with realtime nature. The japanese pulled the biggest bartering bullcrap of the 21st century and I am looking down for more than physically being taller. Old school seems to be synonymous with dumb, and that just may start a freakin World war. Truth can't hide forever dammit.
you want out of injection for reality, time in a carb for your engine. All weather survived as well.

when the engine goes wide open throttle is that like the same as a carb?

i made a lier box for my MAF sensor today. it worked and was easy to make. it defiantly helped the problems i have with the MAF. i noticed that it wasn't just wide open throttle that open loop that was affected but the whole throttle range. i had to dial it up when i accelerated and down again when i wanted to coast or engine would die. i think the ECU tries to adapt to the new settings each time you change the lie box setting and its confusing. wideband is needed.