0.5 volt dc battery to fool O2 sensor?
 

I have been driving around watching a digital multimeter which is tied in to my one wire oxygen sensor wire (87 civic with feedback carb.)

The feedback system seems to be doing its job (readings jump around but 90% of the readings are below .45 volts when cruising on the highway). When I floor it, readings jump up to 0.8 and when decelerating in gear, readings go down to near zero volts.

I want to fool the system into leaning the mix out a little, so if I could just feed 0.5 volts into the oxygen sensor wire, then the computer would think it was running rich and tell the carb to lean it out.

Its the same theory behind the "fever buster" mod that is listed here on the intake temp thread.

So, I figure that all I need is a good low amp DC 0.5 volt source that will mimic a rich mixture oxygen sensor signal, but won't have enough amps to fry the computer.

Any ideas from the electronics whizzes here? How about if I use a 1.5 volt battery and a ton of resistors in series, might that be safe?

I am not looking to do this long term, just for a 30 mile test run.

Thanks in advance for any help.

You could do it with three equal sized resistors in series. Tap off of the positive of the battery and at the connection between the first and second resistor of the three and you would have exactly 0.5 volts. Then, connect the circuit in series with the wire going to the O2 sensor with the positive going to the ECU side and the negative to the O2 side. You would want to use fairly high value resistors to limit current flow. If you wanted to have some fine adjustment, you could add a variable resistor in the above circuit.

Thanks diamondlarry! But let me double check that I understand how to do this.

So, I have a simple loop with the positive end of a 1.5 volt battery (example AA) connected to three strong resistors in series and then back to the negative end of the same battery.

Then, I run another wire off the positive end of the 1.5volt battery and connect it to the wire that goes to the ECU and I run a wire from the connection between the first and second resistors and run it to the oxygen sensor. So I am forcing the oxygen sensor output in series through this circuit.

Its the resistor junction that I want to double check to be sure I know which one I should use.

The wire that goes to the oxygen sensor, should that be connected between the first and second resistors as viewed from the positive end of the 1.5 volt battery or as viewed from the negative end of the 1.5 volt battery?

Thanks again

It would be the one as viewed from the positive end of the battery. You would want to double-check everything with a multi-meter before making connections to the vehicle.

So you want to fool your computer into running lean... :confused:

Just be careful and I hope you know what your doing... get too lean and you'll be buying a new engine.

I really envy the VX lean burn system and thought I could tinker with one of my own :D

Seriously, I do appreciate the word of caution, yes I'll be listening carefully for spark knock during this limited time test. A few months ago found and patched a leaky vacuum hose (manifold vacuum). It was likely leaking for at least 500 miles.

This is a bad plan. Yes, you can fool the motor, but you will in all likelyhood get the motor to think the o2 sensor is bad. Then it will go into limp mode... and have bad fuel economy.

"lean" burn motors, don't just inject less fuel. they use special valve and combustion chamber setups to make sure that the air fuel ratio is correct near the spark plug.

By running lean, you get MUCH less power for a given amount of fuel. You want to get the greatest amount of power you can from each drop of fuel you burn.

You will get the best fuel economy by burning less gas.. with the correct amount of air. That means using smaller throttle openings. However, by using smaller throttle openings you're reducing the effective compression ratio. That too reduces the amount of power you can extract from a given amount of fuel.

You might consider a very thin copper headgasket and taking great care to avoid using large throttle openings. That would increase the efficancy of your motor at part throttle.

That said, to fool the motor, I would not use straight resistors. I'd use a 100k pot, and a 1mohm resistor, fed off of the battery. Use the 100k pot to get the output voltage you want.

I've been monitoring an '86 one wire sensor tied to a EFI throttle body system (4x4) and it acts just like yours. I was also using a vac gauge and found that unless I kept the vac above 10-12 " HG...the cycling stopped and it went into open loop. With this truck I had to be very attentive to keep it in closed loop during cruise...headwinds and hills were issues.

I doubt if the .5 V added will do the job because as I understand it...the ECU looks for the cycling AND the avg voltage level to determine the mix. You would need to add extra voltage ON TOP of the cycling voltage...which is what an EFIE does.

O2 sensors are tested for being in spec by watching for the number of cycles in say 10 seconds...if they don't cycle enough they are considered as LAZY and need to be replaced.

You could do damage to both your O2 and your ECU if not careful.

It would be nice to find an easy way to add some voltage TO the O2s cycling voltage...but not so sure it can be done.

Running a motor lean can also produce worse emissions. Also, one of the main functions of gasoline is actually to cool the combustion chamber and piston head.

If you use water injection correctly, you should be able to remedy the problems listed above.

Gotta run, I'll post more later.

Cvcc
 

I already have the special combustion chamber/valve setup that you mentioned on my stock CVCC engine. Its a 3 barrel carb that feeds a rich mixture to a mini combustion chamber directly surrounding the spark plug through a tiny extra intake valve and special ports in the intake manifold. The other two barrels work like a normal carb (secondary kicks in as needed) and feed the rest of the cylinder with a lean A/F mix via 2 intake valves.

I am assuming that the feedback carb further leans out the "lean" carb mixture, but I haven't been able to get enough info about this system to know for sure.

The head was shaved when I replaced my head gasket 5 months ago, but the machinist just milled off enough to get it truly flat.

In layman's terms, what is a 100K pot? A 100K resistor? Or a variable resistor like a volume knob on a radio?

I'm not sure if what I have is technically a computer. Its called a "control box". There is no check engine light with this system.

Water injection might be worth a try if it would help me here, but I'm worried that I might end up with too many variables to test at the same time.