Rians Ke70

[LittleRedSpirit]

Narrowbands don't tell you anything, I didn't even know you could get one with a gauge. If its a narrow band sensor like a regular factory 02 sensor it will only read between 14 and 15.5:1. This might give you a false ndication of leanness. You should achieve those afrs while cruising at light throttle or when you back off. Cars routinely run that lean, but only under low load conditions.

 

Do you have an 02 sensor in your wiring loom? It should allow your ecu to cruise quite lean and economically if its working right.

[rian]

The needle sits just above the bottom of the "lean" scale under all throttle and rpm positions, then when I back off the needle jumps to half way on the lean scale. At no point when I was driving it did it get anywhere near the stoich or rich part of the gauge, which I thought was a bit weird considering how much the car smells like it is over-fueling.

I know the gauge definitely works because it will pick up the potential difference between my skin and the negative battery terminal, it says I'm running stoich haha.

 

Yeah there is an 02 sensors wired in, but I'm not sure if it is actually working. I need to test it with a multimeter.

[ke70dave]

Didnt you say you had holes in your exhaust?

 

any holes will add oxygen. adding oxygen will mean the o2 sensor picks it up and you will read lean.

[LittleRedSpirit]

You need a wideband, simple as that.

 

Wanna borrow mine for a few weeks to test it and datalog it?

[rian]

^Would be super cool, but I don't see how that will help unless I had a bung welded in for the sensor.

I understand what you're saying about widebands being much more accurate but I figure that I don't need to know the exact AFR, I just want to see what my O2 sensor is doing, and the O2 sensor is a narrowband sensor so I see no problem with putting a narrowband gauge on it.

 

Here's what I'm actually trying to achieve:

 

 

Assumption:

Engine only has AFR problems during closed loop operation.

 

Hypothesis:

O2 sensor is faulty, and reading the air/fuel mixture as lean when it isn't. The ecu receives the "lean" O2 sensor signal and adds more fuel in the closed loop circuit - engine runs rich.

 

Test:

1. add narrowband gauge, allowing me to read the signal from the O2 sensor while the O2 sensor is still wired to the ecu.

2. check O2 sensor voltage with multimeter

 

Results:

1. dial on narrowband gauge doesn't move until engine is hot, then when it's hot it reads as "lean" at idle and under full acceleration through the entire rev range, then "very lean" when I back off.

2. at idle (when hot) the multimeter reads between 0.5 and 0.6 volts (stoich), as I accelerate the reading goes down to 0.2 to 0.3 volts (lean) [possibly in open loop]**.

 

Taken from Wikipedia:

0.2 volts = lean

0.5 volts = stoich

0.8 volts = rich

 

Conclusion:

- I suspect the O2 sensor is stuffed and the gauge is also shit, or I have a problem with my wiring somewhere.

 

What I'm going to do:

- Replace O2 sensor, they're only $40 anyway.

- Retest the O2 sensor voltage with multimeter, if it's still says I'm running lean then my assumption was wrong and I'll look at a wideband so I can track down the problem.

 

 

**This is why I think the O2 sensor is faulty, the engine is meant to run rich in open loop because of the fuel map, but the sensor reads as running lean.

Either the sensor is stuffed, the injectors/fuel pressure are/is not right or I have a massive vacuum leak (not likely as the idle is fine).

Edited October 29, 2013 by rian

[rian]

Ordered O2 sensor.

 

Also would anyone be interested in swapping something I can daily for my blue KE70?

 

[B.L.Z.BUB]

Ordered O2 sensor.

 

 

 

Two?

[rian]

Two?

 

The catalogue didn't specify whether you could unscrew the sensor from the bit that bolts to the manifold, so I'm waiting for mine to arrive so I know for sure. If it is the screw in type I'll order another one ;). I think they're about $50

[It's_AUDM_Yo]

Also would anyone be interested in swapping something I can daily for my blue KE70?

 

If you sell it you can buy my Camry for 3-4k ;)

[altezzaclub]

I picked up a pair of O2 sensore from a Commode at the wreckers for the Jaycar mixture display. Its narrowband and reads from 11 through to 20.

 

You're probably right that the sensor is stuffed, you should definately be reading rich under accceleration. I also find it takes a minute or two to warm up and start reading correctly. Then its lean for deceleration, 18-20 or just "L", cruise is about 16-18, slight acceleration is 14.7 and hard acceleration is 12 or so.

[B.L.Z.BUB]

Commode

 

Best typo ever. If it was planned I still tip my hat.

[LittleRedSpirit]

I think you're missing something, or I am. Assuming youre using the stock toyota ecu?

 

Correct me if I'm wrong, maybe I'm missing something but the sensor you speak of is just the stock type 02 sensor from the original toyota efi system?

 

If so, then It reads only between 14 and 15.5:1. Thats all its designed to do. Watching it on a gauge that promises a wider range of operation is never going to work. The sensor itself is only used to meter the fuel under very light loads or decel conditions where maximum fuel savings are possible/desired to reduce emissions.

 

The trend you describe, showing lean under load, is completely consistent with the correct operation of a stock o2 sensor, you're expecting that it will do what a $90 wideband probe will do, and that's read a wider range of afrs than 14-15.5:1. Plus with a wideband there is a whole extra circuit to calibrate and help interpret the data that comes from it. A stock ecu will just read the voltage from your 02 sensor and interpret its values from its tables. Its only going to look for its input when your are doing no more than gentle throttle application.

 

The thing will never be able to meter the 12.7-13:1 that the tune will be making happen under hard load, as it cannot read any richer than 14:1. So therefore I'd expect that that is why you have lean under higher loads, it just stops there and cannot read richer. Giving readings as lean or leaner is not really scientific either. The only way to truly test what you are doing here is to install a completely separate wideband and observe that in real time, or to go to a dyno and have it checked out there with a good o2 probe and monitor your gas levels and temps too.

 

Looking at it from another angle, what sort of mileage are you getting from a tank? What makes you sure its rich?

[ke70dave]

I know this is just repeating what matt just said, but i like graphs, and this one sums up the difference very nicely.

 

[rian]

I picked up a pair of O2 sensore from a Commode at the wreckers for the Jaycar mixture display. Its narrowband and reads from 11 through to 20.

 

You're probably right that the sensor is stuffed, you should definately be reading rich under accceleration. I also find it takes a minute or two to warm up and start reading correctly. Then its lean for deceleration, 18-20 or just "L", cruise is about 16-18, slight acceleration is 14.7 and hard acceleration is 12 or so.

 

I think you're missing something, or I am. Assuming youre using the stock toyota ecu?

 

Correct me if I'm wrong, maybe I'm missing something but the sensor you speak of is just the stock type 02 sensor from the original toyota efi system?

 

If so, then It reads only between 14 and 15.5:1. Thats all its designed to do. Watching it on a gauge that promises a wider range of operation is never going to work. The sensor itself is only used to meter the fuel under very light loads or decel conditions where maximum fuel savings are possible/desired to reduce emissions.

 

The trend you describe, showing lean under load, is completely consistent with the correct operation of a stock o2 sensor, you're expecting that it will do what a $90 wideband probe will do, and that's read a wider range of afrs than 14-15.5:1. Plus with a wideband there is a whole extra circuit to calibrate and help interpret the data that comes from it. A stock ecu will just read the voltage from your 02 sensor and interpret its values from its tables. Its only going to look for its input when your are doing no more than gentle throttle application.

 

The thing will never be able to meter the 12.7-13:1 that the tune will be making happen under hard load, as it cannot read any richer than 14:1. So therefore I'd expect that that is why you have lean under higher loads, it just stops there and cannot read richer. Giving readings as lean or leaner is not really scientific either. The only way to truly test what you are doing here is to install a completely separate wideband and observe that in real time, or to go to a dyno and have it checked out there with a good o2 probe and monitor your gas levels and temps too.

 

Looking at it from another angle, what sort of mileage are you getting from a tank? What makes you sure its rich?

I know this is just repeating what matt just said, but i like graphs, and this one sums up the difference very nicely.

 

 

 

Ok I fully understand how both sensors work, I think I'm just not explaining my problem very well.

 

If you look at the 2 plots that Dave posted, you'll see both sensors will have a voltage output for fuel ratios between 10:1 and 20:1. The wideband is linear, the narrowband is piecewise/shit.

However despite being shit, the narrowband still has a voltage output for any of the fuel ratios, and notice there is ~1 volt difference between rich and lean.

My problem is that my O2 sensor seems to always be reading ~0.2 volts, which is lean. If it were working properly it would at least be constantly jumping back and forth between 0.2v and 1v at idle (ECU doing it's trim thing or whatever). So I assume it's stuffed.

 

Anyway, there is a new o2 sensor waiting for me at work. I'll plug it in and see if it makes any difference and I'll post up the voltage outputs here.

[rian]

Took photo, upload photo.