toytrain13

When I start up my 2002 ES300 I get a really nasty exhaust smell- really stinks up the garage. In a minute it's gone. I've put up with this for the year I've owned the car, but now I'm requesting comments! Thanks!

BDSL

Define "nasty"....

toytrain13

Smells like incomplete combustion. Hydroicarbons!

BDSL

Gasoline smell is normal on cold engine.
Cold engine runs rich (i.e. in closed loop and more fuel).

PuReChaos

I think you mean open loop, closed loop is what you want to achieve when the engine is warmed up. But I would also wagger that just sitting in the garage wouldn't be the beset option for smell unless the garage door was open and the tailpipe was facing out.

BDSL

Yes. Open loop. Sorry, my brain and eyes aren't working 100% as I have been working long hours for the last 5 days consecutively....

Hayk

iTrader: (1)

I have the same problem, so I would like to chime in.

I have heard people comment on how bad my car smells on cold starts, even when it's outside. I just recently hooked up my OBD2 scanner and monitored the Long Term/Short Term Fuel Trim data for both banks, while the car was warming up. Problem is - I have no idea how to interpret the data. What values should I be seeing on a perfectly tuned 1MZ-FE?

Side note: One of my Upstream sensors has been replaced recently by a Denso unit, the other appears to be the original sensor, as I pulled it out of Bank 1(Firewall) and it had Toyota markings on it. I decided to put it in Bank 2 for easier access in the future.

If I remember correctly, the car had higher Short Term Fuel Trim percentages in Bank 2(original sensor), when it got to the Closed Loop. I will have to double check the data, one more time.

PuReChaos

Fuel trim measurements are only really accurate when in closed loop. In a perfect 1MZ-FE, you want long term fuel trim to be 0 on both banks and short term to be 0 as well. However, thats in a perfect world. Realistically you want to see long term fuel trim as close to zero percent as possible as its stored in memory. It can be either positive or negative. Short term fuel trim is a direct result of oxygen sensors so it will bounce back and forth and is not stored in memory. It is normal to see it bounce around as long as its lower than 10% its "okay" and functioning within reason for a well maintained car. The main goal is to have long term fuel trim be close to zero. I like to say my car is in good shape and currently its running 1% long term. So technically my car is making up for a slight lean condition. If you would like to know more about fuel trim don't hesitate to ask.

Hayk

iTrader: (1)

Thank you for the response. That is exactly what I gathered from watching an educational video from ScannerDanner on YouTube, about Fuel Trim data. I also recorded about 14 minutes of my own car - starting with a cold engine and getting it up to temp by driving around.

From what I understand, Long Term Fuel Trim on startup is a memory of the last position that the computer was in, at that particular load (RPM). The computer relies on LTFT, until the O2 sensors get up to temp. Once the engine and sensors warm up, the computer goes into a Closed Loop, and starts controlling the fuel injectors based on the voltage readings that are coming from each bank. This is known as Short Term Fuel Trim. As you said, STFT will jump from + to -, in order to keep the catalytic converter happy, by supplying Carbon Monoxide and Oxygen.

With that said, here is what I gathered from my little experiment.
Side note: Bank 1 has a brand new O2 sensor, while Bank 2 has the original sensor.
Vehicle is at 210,000+ miles.
Data points are recorded every 3 seconds

1. The initial Long Term Fuel Trim was set to Bank 1 (-0.8%) and Bank 2 (6.3%), before I even cranked the engine.

2. Thirty seconds after startup. Coolant temp is at 50 degrees Fahrenheit. Engine RPM = 1500.
As soon as STFT kicks in, both banks start to almost equally subtract fuel. Both dipping as low as -17%, at slightly different times. (about 20 sec apart).
While this is happening, LTFT Bank 1 and 2, is at 0% and 4%, respectively. This lasts for the next seven minutes.

3. Seven minutes after start up, I begin to move. Coolant temp is at 150F.
If you add up Long and Short Term numbers for each bank, they are jumping from -4% to 4%. Most of the time, each bank adds up to zero.

4. Thirteen minutes after start up. I come to a stop and begin to idle. Coolant temp is at 180F. Engine RPM = 700.
Short Term Bank 1 and 2 never go above zero. Bank 1 dips to -0.8% and Bank 2 is at -1.6%
Long Term Bank 1 is at 0%, Bank 2 is at 6.3%


I do not notice a bad smell after driving. It's only on cold starts, after idling for a minute or so.
From what you and ScannerDanner have said, LTFT +/- 10% is nothing to worry about, unless there are symptoms of something wrong. A positive Fuel Trim number means that there is a Lean Condition, so the computer is adding fuel.

Also, the computer seems to report that both banks are running rich, about a minute after startup. Strangely, it appears to me that there is a constant lean condition in Bank 2.

I would like to replace the original sensor and rerun the test. What do you think?

PuReChaos

When you first start the car, even though the scanner may be showing you STFT/LTFT neither of them are truly accurate because in open loop the ECM ignores signals sent from the O2/AF sensor as they only are accurate once they are heated up. So the scanner will show what the O2/AF sensors are telling it, but only LTFT is being used from your previous drive, in your case Bank 1 (-0.8%) and Bank 2 (6.3%), not the one that is currently being displayed by your scanner.


So just clarifying, your car still isn't at operating temperature yet correct? Closed loop only occurs when the engine is at operating temperature meaning ECT reading at or above 180F. I also don't know if its the scanner you're using or not, but typically when I hook up a techstream to a car, the LTFT won't move until after a load has been placed on the car for a couple of minutes when in closed loop (ECT reading of 180 or higher).

The final readings are good in my opinion, the short term fuel trims are in the 'great' range (less than 5%). Your scanner may not be able to display it properly since STFT literally changes almost every second and three second intervals might not show it properly but I would wager it goes between positive and negative.

Honestly, I'd keep the old sensor as it seems to be reporting more accurately in my book than the other. The fact that prior to turning your car on, your scanner read Bank 2 as 6.3% and later on after your car finally went into closed loop it continued to read 6.3% is good. An engine that is maintained shouldn't have the LTFT change day to day unless something were to change such as a vacuum leak, restricted fuel injectors, dirty MAF, etc. Yes it will change over time but only slightly as the engine ages and things being to wear. Does this mean I think your other sensor isn't functioning right? No, it very well could be a sensor "issue" and that it is tending to want to bias one way over another (referring to Bank 2). This does happen with sensors sometimes and more common with old sensors.

If you truly want to diagnose your sensors and STFT/LTFT, have your car parked in your driveway or on the street, let your car get to operating temperature, both by your scanner and cluster, meaning ECT reading at least 180F. Then record the STFT/LTFT numbers at idle, 1000 RPM, 2000 RPM, and 3000 RPM for at least a couple seconds to see what sort of numbers you are getting since your scanner is in increments of 3 seconds. If your scanner is able to read the Total fuel trim (Total FT) all the better and that number should be recorded too. Not necessary, just helps further with diagnosis as Total FT is an ECM interpretation of the air/fuel mixture. I will again be more than happy to help with diagnosis if possible.

Hayk

iTrader: (1)

My scanner can tell me the state of the fuel system - open or closed, but I didn't have that function turned on at the time of the test. I believe my car goes into a closed loop before it gets to 180F. I will double check and report back.

Also, if I remember correctly, the scanner's live data refresh rate is 1.4sec, but it records every 3 seconds. So I should be able to watch the values change a little quicker than what's in my Excel spreadsheet.

About the test, should I let the car get to operating temp without moving it? It's usually sitting outside, so that should be easy to do.

Lastly, you mentioned a possible scanner issue. I know my car uses A/F sensors, as it is a CA spec vehicle. The scanner kept saying that they were both around .66V the whole time, but I heard that A/F sensors behave differently to O2 sensors, in the fact that they hover around 3Volts instead of switching from .1 to .9. Is this true?

Edit: .66 is one fifth of 3.3V. My values just have to be multiplied by 5.

PuReChaos

Actually I do believe you are right about that, typically we'll start the car, leave it and work on something else while the car is reaching operating temperature and then come back, and when I return it generally is around 180F. However I don't know the temperature range, but if your scanner is able to tell the status, that could be used also to determine when the fuel trim numbers are actually being used.

But yes, I just wasn't sure of your exact refresh rate but that should be fine, just hold for like 3-5 seconds at each and record the numbers. And yes, let the car just sit and reach operating temperature.

PuReChaos

Just noticed your updates/edits.

If I did I apologize, I meant sensor issue. Sensors as they get older can sometimes get "stuck" or bias one direction over another. It may be lean or it may be rich, it is easy enough to detect if you're using an oscilloscope for diagnosis.

A/F sensors are just more accurate O2 sensors. They are more accurate on the lean side than rich (someone correct me if I stated that wrong). They operate similar to O2 sensors in that they have a middle voltage at 3.3v and lower than 3.3v is rich and above 3.3v is lean.

Not sure what to make of your .66v reading as my first thought is extremely rich condition or incorrectly functioning sensor. So your scanner was reading the A/F sensors at .66v? I honestly can't believe that to be right.. but i'll wait to hear back from you.

Hayk

iTrader: (1)

I found a video by ScannerDanner where he states, as written in his Snap On Verus Troubleshooter, that regular OBD2 scanners will only display 1/5th of the Voltage reading, probably because O2 sensors work on a 1 Volt scale, as opposed to a 5V reference for WideBand O2 (Air/Fuel) sensors. So in short, .66V is 3.3V.

Updated my post, after reviewing the logged data.
In regards to your proposed test run. I did it yesterday and here is what I saw.

1. My fuel system went into closed loop within a minute of startup. S1 closed after 30sec, and S2 closed about 30sec after that. The coolant temp was still around 50-60F degrees. STFT on both banks immediately started subtracting fuel, just like they did in my original test.

2. I waited till I got to 176-180F and stayed there for a few minutes, before measuring my Fuel Trim at different RPM ranges.

3. At 2000 and 3000RPM, the fuel trim is 0 on both banks (Short and Long)

4. When I drop it down to idle (700RPM), the Fuel Trim seems to be a bit slow to respond. It starts cutting fuel (negative STFT numbers), possibly due to Deceleration, but the values stay in the Negative (Rich Condition) for 20-30 seconds.

5. After those 30 seconds, Fuel Trim on Bank 1 will come back up to zero, but Bank 2 will go up above it.

6. It eventually settles where Bank 1 will be zero and Bank 2 LTFT will be at 3%, and STFT will be zero on both banks.



I did a few Snap Throttle tests by observing the sensors on a graph (very basic feature on my OBD scanner). It appeared that both sensors were in sync. Maybe the old sensor was a little different, but not by much.

I'm thinking maybe this "rich condition" at Idle and a slow response to change is due to carbon buildup in the throttle body. I will try to thoroughly clean out my Idle Air Control Valve, as I have never taken it apart, only sprayed some Carb Cleaner inside. I've been cleaning my MAF sensor on a regular basis, so I don't think that's an issue, but I will check it as well.