Running "Lean" with a Resistor Divider on the Air Temp Sensor

[steensn]

I tried to search for this, so sorry if this has been discussed...

I hear a lot of talk about HAI and providing a fuel mixture that will give you the least amount of fuel with each revolution of the engine. We all don't care about the loss of HP, unless you car won't have enough power in the mountains of California... AndrewJ...

So my thoughts were, running the engine hoter would obviously help keep the temp up, but what about putting a small resistor divider circuit on the air temp sensor signal to force it to read maybe 5-10 degrees F hotter than it is. This way you could get a "'lean" condition at all times and with the fuel mixture having a higher oxygen content to have a more complete burn.

This should be a simple modification if you can chart the air temperature verses voltage of your car and find out the degree per volt ratio then just make a small circuit board to trick the ECU into a lean burn.

This could be combined with a better intake temperature to optimize your engine temp so that you don't run it to hot in a HAI setup.

You might then be able to go back to a CAI and have a larger voltage divider so that you can get a super lean burn with the most oxygen content available.

Any thoughts?

[Daox]

Sorry, won't work on modern OBD engines. The O2 sensor checks your air/fuel ratio and will correct it. Plus, 5-10 degree difference will do next to nothing as far as fuel trim. Thats also not to mention that a lean burning engine emits more NOx emissions which is nasty stuff. Factory equipped lean burn engines have a special catalytic converter to deal with this.

[steensn]

What you did 20-30 degrees F then? For a 1996 Ford Aspire which is just "new" enough to have a OBDII, would you still expect a problem? I know a lot guys one here have non OBDII cars, like a 93 Firefly where maybe this could work. You could always tie the Cat to the same test and modify both. I'm in MI and I don't have to go through a NOx emissions test at all.

I don't want to shoot the idea down if it can be modified to work, ie if the air temp isn't high enough, what about a bigger resistor divider.

As for more NOx, Honda does the lean burn and sels it in 49 states where it is legal.

[cfg83]

steensn -

Daox is on the money. I don't think the IAT resistor mod will work because the ECU/PCM will normally be in "closed loop" mode. In that mode, the IAT sensor value is *ignored* in favor of the oxygen sensor. The oxygen sensor is used to try to keep the engine at the stoichiometric ratio of 14.7/1 (14.7 parts air to 1 part fuel).

Instead of doing the resistor mod, you can build an HAI/WAI (Hot/Warm-Air-Intake), where you route the input air in from the exhaust manifold. In that way, the engine will be getting "real" hot air that has a real effect on the oxygen sensor.

There is an experiment I need to finish where I hope to prove this behavior, but I have been too lazy to do it.

PS - The corollary to this is the "performance" resistor mod, where you put a bias on the IAT signal to make the ECU/PCM think it is getting cold air. I think this works because under *performance* driving conditions, you are typically driving at WOT (wide open throttle). Under WOT, the ECU/PCM will switch to open loop mode and use the IAT sensor to determine how much fuel to put in.

CarloSW2

[steensn]

I've read the HAI threads several times, but how do we know that is the most efficient? At some point pumping in too hot of air will decrease efficiency given you less of a return for the fuel mixture. If what the ECU is trying to do is take advantage of colder air, why not trick it?

If its the O2 sensor that is the main sensor in the e nd, then why not do it on the O2 sensor instead? you could easily hook up a circuit with a pot to run ABA test with a Scangauge.

In the end, what I would want to accomplish it to have a fuel mixture "independent" of intake temp so that you cna then optimize the intake temp.

So my question would be, why is a HAI the optimal and not a WAI? you would say because the HAI, in a unamed number of cars, would give you the best fuel sipping drive possible. My thought is, why not optimize the fuel being injected with a simple circuit so that you can attempt to optimize the air coming into the engine. This way you can optimize both to get the max benifit and MAYBE even run a little leaner so that you use even less fuel than a HAI would get you at a given temp.

[tasdrouille]

You need to work on multiple sensors to have it working properly. The ECU will only allow the o2s to lie so much before calling it quits and going in open loop. If you hit the o2s, cts, mafs or maps, iats and tps the ECU will only see smoke.

With the iats mod (res in series to simulate colder air) alone you still get a small timing advantage though.

[steensn]

Originally Posted by tasdrouille

You need to work on multiple sensors to have it working properly. The ECU will only allow the o2s to lie so much before calling it quits and going in open loop. If you hit the o2s, cts, mafs or maps, iats and tps the ECU will only see smoke.

With the iats mod (res in series to simulate colder air) alone you still get a small timing advantage though.

I am well aware of failsafing in vehicle ECU's in cars, I have several while in working here. The question is what can you get away with? I have a 1996 Ford Aspire, not very new and it isn't going to have the level of failsafing as some of the newer cars here.

I'll prolly rig something out to test this once I get a Scangauge from the group buy to test it out. At some point it will set a fault, but I'll just reset it and increase the with a pot. I heard there is a performance Cat for my car I can put on that could help with the NOx if this works out.

[cfg83]

steensn -

Originally Posted by steensn

I am well aware of failsafing in vehicle ECU's in cars, I have several while in working here. The question is what can you get away with? I have a 1996 Ford Aspire, not very new and it isn't going to have the level of failsafing as some of the newer cars here.

I'll prolly rig something out to test this once I get a Scangauge from the group buy to test it out. At some point it will set a fault, but I'll just reset it and increase the with a pot. I heard there is a performance Cat for my car I can put on that could help with the NOx if this works out.

My theory is that the scangauge will be spoofed by the IAT resistor mod. It's algorithms (most likely) use the IAT sensor to extrapolate MPG.

I really gotta do that experiment. I am soooooooooooooooooooo lazy sometimes.

CarloSW2

[steensn]

Originally Posted by cfg83

steensn -



My theory is that the scangauge will be spoofed by the IAT resistor mod. It's algorithms (most likely) use the IAT sensor to extrapolate MPG.

I really gotta do that experiment. I am soooooooooooooooooooo lazy sometimes.

CarloSW2

Good point on the scangauge probably being fooled also. At minimum I can sue the pot and the scangauge to read out what the ECU thinks the new intake temp is or O2 level. I could then see how much of a gain it takes to set a fault or go into another mode where it thinks the O2 sensor is failed.

I would be concerned that not all cars are going to have the same failsafing criteria as that gets more and more sophisticated as time goes by. I can compare our failsafing from 1996 and now and I would laugh at the crude failsafing they did back then. Who knows how much you can fool it unless you test it?

[cfg83]

steensn -

Originally Posted by steensn

Good point on the scangauge probably being fooled also. At minimum I can sue the pot and the scangauge to read out what the ECU thinks the new intake temp is or O2 level. I could then see how much of a gain it takes to set a fault or go into another mode where it thinks the O2 sensor is failed.

I would be concerned that not all cars are going to have the same failsafing criteria as that gets more and more sophisticated as time goes by. I can compare our failsafing from 1996 and now and I would laugh at the crude failsafing they did back then. Who knows how much you can fool it unless you test it?

I agree. The newer cars are going to be "harder to fool". I would rather work with a dumb ECU/PCM. I *think* my 1999 Saturn OBD II protocol is pretty close to it's 1996 specs, so I see it as "the good dummy" I can work with.

CarloSW2