O2 sensor

[hwholder]

Got questions about O2 sensor for XFI 2.010.
It seems to me that the O2 correction response has slowed down and it has become harder to smooth out the fuel curve. Can O2 sensors or the ECU circuit get lazy?

I went to the XFI 2.0 ECU op manual in the Q&A section for instructions and did the following:

The Lambda symbol appeared indicating it was connected to the ECU. I started the vehicle and looked at a sensor called UEGOS (V) in the main dashboard display. The manual said that this voltage should stabilize between 0.43 and 0.49 volts within 30 seconds of operation. I read 0.41 for three different tests.

The manual went on to say that if the UEGOS voltage is out of range, the sensor will become inoperative and that this could be the result of wiring harness damage, physical damage to the sensor, contamination of the sensor, or damage to the ECU. Visually everything appeared good, but the UEGOS (V) was slightly out of range (0.41 vs 0.43 to 0.49) - does this mean I need to replace the O2 sensor?

Next I took the O2 sensor out and turned on the ignition to see if the element would glow orange in the end of the sensor. It was and the body of the sensor was getting very warm as I expected.

Next I followed the steps to verify that the oxygen sensor circuitry in the ECU was working properly. I disconnect the oxygen sensor to observe the following sensors in the main dashboard display while online. The manual said that UEGO (V) should read approximately 3.51 volts, UEGOS (V) should read approximately 1.02 volts, and UEGOR (V) should read approximately 4.98 volts. My readings were UEGO (V) 3.51, UEGOS (V) 1.00, and UEGOR (V) 4.98 - Is UEGOS (V) at 1.00 vs 1.02 close enough?

Any advice would be appreciated.
Cal - If you think I need an new sensor have Holly call me.

 

[norbs]

If everything looks good as it shows, it still could be the sensor going bad. As the sensor gets worn the response will be lazy as you say. You can use brake cleaner in a rag on the heated sensor and data log the response time from placing the rag into the sensor, and maybe compare the test to a new sensor. However don't leave the rag into the sensor too long or it may catch fire.......... I just tested a sensor for you. A/F went from 7.25 to 15.94 in 7.1 seconds. You only have to hold the rag on the sensor 3 seconds the most then pull it away. The Rag must be soaked wet with brake cleaner, just give it a squirt enough to cover the sensor. Just becareful its very flammable, so use common sense....

 

[hwholder]

Hey Norbs, thx for the reply.
Think I'll call Holly and order a new one.

FYI - My J&S was giving me a slight indication (LED flickers) just before the shifts (air shifter) and I wanted to add some fuel to compensate.
Looked to me that added fuel was occurring way late. First time I noticed this. Now that I'm paying more attention, the O2 correction seems to be slower that I remember it being.

By the way, I revisited a thread about cam sync last night and wanted to throw something out about my set-up.
MSD crank trigger set at 50 deg BTDC. MSD distributor with the tabs cut off of all but one. Mag pick-up in distributor for cam sync. Tab triggers the mag pick-up at the #6 wire position with the distributor rotor phased at 25 deg BTDC.
I did the timing light test taking timing out of #1 and changing cam sync precedes #? and came up with cam sync precedes #6.
On the dashboard the cam sync number is 111 BTDC. I saw where you had some discussions in the past about cam sync numbers. How does my number sound to you?
The cam sync number is displayed on the dashboard in red and I have noticed that red numbers appear sometimes.
Do you know what determines when a number is displayed green, yellow, or red on the dashboard and what that might mean?

I see you have a J&S. I set mine up using an AEM data logger to get the speed I needed to log the J&S output along with my crank trigger and cam sensor. I had to put a 5 volt clamp on the crank and cam sensors going into the AEM. I knew that if I could see the crank and the cam I could determine which cylinder might be acting up. Interestingly enough, the close proximity of the cam and crank signals going into the AEM connector superimposed the cam signal onto the crank signal. Therefore I can just look at the crank signal and see both.

I got tired of downloading and studying the AEM files at the track (the XFI is enough to do without more data to sort through) so I put a Gopro camera in the car to visually look at the knock LEDs. When I get back to the pits, I preview the Gopro and if I get a light then I download the AEM. Otherwise I wait till I get home to download the AEM so I can study the data.

I can't say enough good about John and the J&S. Being able to sort out individual cylinder behavior is huge.
Just wish the XFI data logging was fast enough and did not need the AEM.

 

[norbs]

I have some more test data for you to compare, interesting results the lZA sensor reads down lower to 7.-7.25 and the L1H1 sensors read only down to about 8-8.1 ish. Look at the logs of each sensor.

 

[norbs]

Wow I am amazed that, you have gone through all the efforts to do the logging, however I don;t think the frame rate is fast enough, you will need 300 samples per second for crank signal at 6000 rpm and 100 frames per second for cam sync. The only way to get this speed is buy the fast dash and spend the extra cost to 1000fps so you can input raw data into the speed channels. The ECU needs 10 degrees difference between cam and crank signals so it has time to calculate things. Cal was mentioning that cam sync can be between 60-120 and it even could be 120, optimum is 70-90 I would estimate as it puts you in the middle of range. As the timing chain stretchs you will see the cam sync start to retard a bit. The cam sync is a calculated number the xfi comes up with, but I have no idea how it comes up with the info.

 

[norbs]

For those who are to lazy to look at the files........here is a pictorial view of the testing, you want to see the response time from fuel rich to lean and vs versa, results can show between a new and used sensor very easy with this method. The used sensor took 5,66 seconds to go from rich to lean and the new sensor 4.66 seconds. From lean to rich the new sensor abnout 1.14 seconds and the used sensor about 2.25 seconds

 

[norbs]

Now the NEW LZA sensor took 5.95 seconds to go from rich to lean and the USED sensor 7.61 seconds to move the same amount. I would say if your into the 8 seconds plus its time to get a new sensor...These results are pretty much repeatable if you do the same method each time, within .2 seconds.

 

[hwholder]

Wow I am amazed that, you have gone through all the efforts to do the logging, however I don;t think the frame rate is fast enough, you will need 300 samples per second for crank signal at 6000 rpm and 100 frames per second for cam sync. The only way to get this speed is buy the fast dash and spend the extra cost to 1000fps so you can input raw data into the speed channels. The ECU needs 10 degrees difference between cam and crank signals so it has time to calculate things. Cal was mentioning that cam sync can be between 60-120 and it even could be 120, optimum is 70-90 I would estimate as it puts you in the middle of range. As the timing chain stretchs you will see the cam sync start to retard a bit. The cam sync is a calculated number the xfi comes up with, but I have no idea how it comes up with the info.

The AEM logger I am using to log knock, crank, cam, rpm, and MAP can sample up to 1000 samples/sec and has 2 GB memory. Not too bad for <$400.
I paralleled the crank, cam, rpm, and MAP from the XFI and the knock from the J&S. Triggered logging "on" with the clutch flag.

Thanks for the data files. Interesting info.

 

[norbs]

So your saying your not getting good data logs from the crank and cam signals? why not?
?

 

[hwholder]

No, I am getting good logs on the AEM data logger and the XFI logger.
The FAST XFI logger is just not fast enough to get the crank at 7000 rpm where I shift.
The AEM logger is capable of logging 1000 samples/sec/channel.
I log the J&S knock output, crank, cam, rpm, and boost on the AEM.
I do all the regular stuff on the XFI including rpm and boost that I am also logging on the AEM.
Kinda a reality check comparing the two.
The logs are very telling when there is knock.
With close examination and a little time I can tell exactly which cylinder is knocking and at what rpm and boost on the AEM log.
I trigger it and the XFI logger with the clutch flag in the XFI, therefore the logs are in sync (start at the same moment).
Comparing the XFI logs and the data from the AEM is priceless when you want to know which cylinder(s) might be knocking.
#6 was my problem child. A little retard and a little more fuel worked wonders on the knocking cylinder.
I've been sneaking up on my tune for a couple of years. The effort is never ending but getting there without blowing up the motor frequently is a blessing.

Strange thing happened to me last Wednesday night at T&T. The roll pin holding the distributor gear on sheared during a burnout.
Haven't figured out why this happened yet.
Looks like the pin shifted and moved to one side and when that side cleared the shaft the pin couldn't take it any more and broke.
Have you heard of anyone else ever having that problem?

 

[John at J&S]

Bill:

Can you quantify the gains made since you started tuning for the knocking cylinder?