Strange STFT spikes

2014 Chevrolet Cruze Eco 1.4 Turbo with 6spd Manual.   155k miles

After a P0299 code, checked the infamous boost check valve in the intake manifold to find it gone.  Replaced valve cover and intake manifold.   Car starts and runs pretty good.  Clutch on the AC compressor cycles when at idle and engine rpm bounces.  That could be a separate issue, so ignoring that for now and leaving the AC off. 

No codes.   Keeping the AC off, took a test drive with data logging.   Using a Amazon bluetooth obd2 with the android Torque App to monitor and log.   Looking at fuel trims, I'm seeing a strange STFT spike.   Doesn't seem to change LTFT that I can tell.   

Attached is a chart of my test drive.   What could be causing the STFT to do that?

Thanks!
 

 

Without some more context, there's not much be gathered from fuel trim alone.
At minimum I would want to see loop status, rpm, vehicle speed, throttle angle and o2s, MAF calculated and actual to know when and why the trims go high.
It could be part of the cat monitor, or it could be meaningless if it in open loop fuel control.

Attached is the csv file in a zip format, has everything I can get w/o custom pids. 

I plugged in my actron cp9780a. Went through the State OBD check. All monitors completed, except the EVAP System. I replaced the purge valve last month because it had a pending code.

Thinking it is time to upgrade to a Kingbolen K7, so I get all the functionality in one unit.

Curious too, the Actron is saying I have a bad LR wheel sensor, yet the ABS light is not on. Weird.

fj5gtx, I had a look at your data, but couldn't find loop status in there. Did I miss it?

Wanted to find loop status because that generation of Cruze commonly spiked the short term trim when transitioning from Closed to Open loop, always on fuel cut deceleration. I never got a satisfactory answer on what the strategy is, exactly, but they all do it. Probably even more noticeable on yours as a manual transmission.

Aside from the outliers, your Cruze spends most of its time within 10% of zero, which is excellent. 

Thanks Tyler!
Unfortunately you did not miss it.  For the Torque app and the bluetooth obd2 interface I am using, loop status wasn't in the list of available pids. I would have to add a custom pid to get it.

I wasn't paying too much attention to decel. I was aware the injectors would cut off the fuel in that condition, but it didn't occur to me that it would spike the short term as a side affect.

I received a Kingbolen K7 scanner last night.   I'll set it up and see what if I can capture more data, including loop status.   I'm assuming I can datalog and dump to a file that I can transfer to my laptop, if not I'll be disappointed. 
 

fj5gtx wrote:

Thanks Tyler!
Unfortunately you did not miss it.  For the Torque app and the bluetooth obd2 interface I am using, loop status wasn't in the list of available pids. I would have to add a custom pid to get it.

...

I received a Kingbolen K7 scanner last night.   I'll set it up and see what if I can capture more data, including loop status.   I'm assuming I can datalog and dump to a file that I can transfer to my laptop, if not I'll be disappointed. 
 

Gotcha, no worries.

Let us know how it goes with the Kingbolen. I haven't used the K7, but it looks to be Launch based, which don't include a way to export data files. I'm happy to be wrong. That's one thing that Torque gets right.

Tried out the KingBolen K7 tonight. I didn't see the openloop pid in the list for live data. In fact I didn't see vac/boost either - disappointing. I'll need to dig around a little more to understand if its possible to get more pids than Torque provides. I had hope that I'd get more pids just showing up in the tool without doing anything special.... I'll look a little harder at that later.

The reports are easy to export, but user interface does not support exporting data logs.

So I figured out how to mount the K7 as a usb drive and retrieve the datalog. I found a sample data log I created. However it is binary (with a .TC extension). Seems like TC spells out to Think Car.

Google search found someone who talked about creating a python script to convert it to a csv text file, but they were reverse engineering the TC format for a transmission module.

One could likely create a generalized TC file to CSV convertor. But I'd want a TC published format before I'd spend time on it. I'll look into this more later.

I'm kinda spoiled with Torque's interface on a 11" tablet. I can design whatever graphs/gauges I want, multiple pages if I desire, and easily datalog into csv files. K7 is downgrade in comparison, but the K7 is bi-directional.

Tried the python code on a sample TC datafile from my k7, and it translated it to CSV format w/o issue.

So you can effectively export live data, just a couple more steps to get it.

Tried the K7 bi-directional feature on that cruze.  Was able to confirm relays closing and voltage change at the connector, but the fan is dead.  Likely that big green resistor package, but I ordered a new GM fan anyway.   Removing will be quite the job.   Thanks Chevy    

I'm not sure why I couldn't see a bunch of PIDs when I first tried the K7 scanner.   I was able to record a pile of them tonight, including Loop Status and boost.  The interface is a little non-intuitive for me, so maybe I was in the wrong screen last time.

I don't have the car together enough to test drive it.  Later this week I'll give that a go and post the data including closed loop.   It also has a pid called decleration fuel cutoff.

It also has a pid called decleration fuel cutoff.

Perfect. That's the one you'll see changing when the short terms spike. Loop status will tell you the same, but GM is nice enough to give additional PIDs that spell out what's happening.

I'm kinda spoiled with Torque's interface on a 11" tablet. I can design whatever graphs/gauges I want, multiple pages if I desire, and easily datalog into csv files. K7 is downgrade in comparison, but the K7 is bi-directional.

That's why I have a love/hate relationship with Torque. I want the data export capability, but can't do without the magic button pressing functions. 

Likely that big green resistor package, but I ordered a new GM fan anyway. Removing will be quite the job.   Thanks Chevy

On the bright side, you don't have any transmission cooler lines in the way.   

Got the car together enough to test drive it. Turned all the pids on for datalog. Graphed the STFT and Decel, added a column and wrote a function to give me numbers so I could add to the chart. I used -10 for inactive and -5 for active.

Attached data file (no formula) and graph...

 
 

So maybe I'm onto something.  EVAP Purge Valve was replaced due to a pending code and classic hard start (flooded) after gas tank fillup a couple months ago.   

But as I said at the start of this, STFT spikes positive is still a bit confusing.   I noticed that the EVAP monitor wasn't set after a couple days of driving 400 miles/day.  And I'm seeing some correlation to the Purge Valve opening and STFT spikes.  LTFT always low, but always negative.

Then I notice this other pid, EVAP Vent Solenoid valve command is always "Venting".   I don't know what the normal operation of that thing is... but maybe something is up.

I graphed STFT and EVAP Purge Valve... and I think I see correlation, see for yourself. 

 
  

After some further research... I think the answer is: when the purge valve opens, the engine is programmed to expect a certain amount of predicted fuel vapor.  If that fuel vapor isn't there, the STFT spikes to compensate for the temporary unexpected lean condition.

New data - the EVAP inspection monitor(i/m) won't set, nor will the Catalyst i/m. Can't pass PA inspection with 2 i/m incomplete.

I also noticed that the gas cap isn't clicking anymore today. Unfortunately there is also some corrosion on the lip of the filler tube. Ordered a new OEM gas cap.  Hopefully that takes care of the problem, as I am not looking forward to changing the filler neck.

Then I notice this other pid, EVAP Vent Solenoid valve command is always "Venting".   I don't know what the normal operation of that thing is... but maybe something is up.

That's normal operation. The only time you'll see the vent close with the engine running will be while the ECM is doing the .090"/cap off leak test.

New data - the EVAP inspection monitor(i/m) won't set, nor will the Catalyst i/m. Can't pass PA inspection with 2 i/m incomplete.

The EVAP may be a bit of work, but the Catalyst is typically pretty easy to set on GM's.

Your K7 should have an ECM data list called 'Exhaust Aftertreatment'. You'll want to watch:

Catalyst Monitor Not At Idle Conditions Met
Catalyst Monitor Test State
Catalyst Monitor Test Counter
Catalyst Monitor Test Result

For fun and education, you can also watch:

Loop Status
Upstream O2
Downstream O2

Drive around until 'Catalyst Monitor Not At Idle Conditions Met' says yes. Pull over (safely), shift to neutral, stay on the brake pedal and sit there until 'Catalyst Monitor Test State' goes active. Sit there some more until 'Catalyst Monitor Test Counter' goes up by one.

If 'Catalyst Monitor Test Result' says Passed, congrats, you're done. The Catalyst monitor is now set. If it says No Result, start driving again until 'Catalyst Monitor Not At Idle Conditions Met' says yes again. Rinse, repeat until 'Catalyst Monitor Test Result' says Passed or Failed. It may take up to seven tests to get a result.

Alternately, you can also try raising the engine speed off idle and watch for 'Catalyst Monitor Not At Idle Conditions Met' to say yes. Then test as above. This works with Automatic transmissions, never tried with a manual.

Link to Lemon Manuals for reference.

For the EVAP, it may be worth having a look at Mode $06 results to see which test hasn't completed yet. You can also try using the EVAP Purge/Seal test in the K7 to get an idea if there really is an EVAP problem. GoTech has an excellent video:



My rule of thumb for vacuum decay rate is .1 inH2O per second. Once you've got 6-7 inH2O of vacuum in the tank and you've sealed the system, you can watch the fuel tank pressure sensor and count. The tank pressure shouldn't lose more than .1 in H2O per second. Try to do this test relatively cold, as hot fuel can skew the results.

Yesterday I had a new spare front o2 sensor, so I put it on to see if the graph would look any different than the 12 year old 156k mile original sensor.  Graph looked identical, and I got the old one off without ruining the wire - so I'll put that original on the shelf for a rainy day.   No change in I/M status, still 2 incomplete.

I graphed the purge cmd and the evap vapor pressure on the K7 and took a pic with my phone.... 

 

The OEM gas cap showed up late this afternoon - nice that Amazon would deliver next day and its OEM.   Installed the cap and took a another ride.  I've take alot of rides over the previous three days.   Got it nice and hot, mixed driving.  Sat with it running in a parking lot for a couple minutes.   Got back and checked, Catalyst is now complete - WHEW!

EVAP is still incomplete, but I didn't really expect that one to finish on the first run with a new gas cap.   Pretty sure I can get a PA sticker with only one I/M incomplete, so I think I'll get that tomorrow morning so the car is legal.  

Once I get a sticker, I'm going to capture another datalog and see what the EVAP system vapor pressure looks like when the purge hits.... also see if those STFT spikes are still happening.    Goal is to get EVAP I/M to set.   I'd really like to see this car totally squared away again. 

 

Interesting in that the LTFT are no longer always negative.   I made two changes yesterday: I replaced a gas cap that no longer clicked, and I replaced the original upstream o2 sensor.   The STFT still spikes.  I thought the new o2 sensor waveform looked identical to the old one - I'll have to dig around and see if I have the LTFT data after the o2 sensor but before the gas cap.  

And on the second chart, I noticed that the tank pressure changed significantly - then I looked at the evap vent solenoid actually went "not venting" for a short period of time, right about the same time as the tank pressure anomaly.  May have witnessed an EVAP I/M attempt there?

   

Got back and checked, Catalyst is now complete - WHEW!

Nice!

May have witnessed an EVAP I/M attempt there?

It was running the .090"/cap off test right there, yeah.

Either way, that's just one or two parts of the EVAP monitor. To figure out which boxes still need to be ticked, you'll have to do some digging. Checking the Mode $06 data is one place to start. You may be able to see which tests are completed, and which aren't.

Using your K7 is another. Somewhere in the fault codes menu, there'll be an option called 'Diagnostic Test Status: Since DTC Clear'. It's a list of all the fault codes who's tests haven't run and passed yet. There'll always be a list of codes in there - don't sweat it. Just look for EVAP codes. The codes you see are the tests that need to be completed.

The Actron scanner I have says the EVAP I/M is still inc. But the K7 shows everything "OK" in mode #06... not sure what to believe.