Addressing fuel trim issues

After a suggestion from @lnengineering on another thread, I decided to check the fuel trims on my '99 C2. I already had a wifi OBD adapter, so I picked up the OBD Fusion app to gather some data (BTW I don't have a lot of experience with scan tools, but this seems like a ton of functionality for $10). I took a 30 minute drive at various speeds, and then took the data series starting from after the engine had come up to full operating temperature. As you can see from the charts below, Short Term Fuel Trims on both banks are running in mostly +/- 3% but Long Term Fuel Trim is mostly at 8% or more. I think this means my engine is running lean (that's why it's adding more fuel). I've read that you should keep the fuel trims within +/- 5%, even though the CEL won't trigger until 10%. I want to address this before I put too many miles on my rebuilt engine. My questions for more knowledgeable folks: 1. Am I gathering and reading this data correctly? 2. Assuming there is an issue to be addressed here, what are the next steps for trouble shooting? I gather the primary suspects are the MAF, O2 sensors, and fuel injectors. How do I test these systems with a scan tool?

This is a new engine, right? I talk to so many people and read so many threads/emails, that I lose track. :-) Although I am no authority when it comes to fuel trims and vehicle diagnostics, with the fuel trims adding fuel, it thinks it's lean, when in fact, it's probably running rich and has a vacuum leak or bad injector. IMHO, first step would be to smoke test the engine to look for vacuum leaks. Usual suspects are oil fill tube and AOS vent hose if they are original.

Quotes source post: 16561346

@lnengineering As always, your insight is very much appreciated. This is a new short block with refinished heads and a new AOS. The oil tube is original, but I'm not sure if the AOS vent hose was replaced. Of course, I will address this with my mechanic who did the engine rebuild, and I'm sure he'll test for vacuum leaks. Since this isn't an emergency and we're under shelter-at-home here in Chicago, I'm using this opportunity to learn about my car and see if there are any things I can do to diagnose from home. Does anyone have suggestions on how to tell if the injectors are bad? The fact that fuel trims are off for both banks would lead me to believe it's an issue with multiple injectors, or something upstream of the injectors, right?

Quotes source post: 16561420

If it was me, I would replace all the injectors. Don't risk it. We mandate it for all rebuilds. I for one wouldn't install a fresh engine without a new MAF or set of primary o2 sensors either. They all are critical to the tune and performance of the engine.

If you want to track fuel trims a Durametric is what I would use. The posted graph is tough to decipher, at least for me. ​​​​Your RKAT value or adaptation at idle should be 0. This value is taken at idle, engine at operating temp. RKAT values are most sensitive to intake leaks and will be positive if adding and negative if leaning fuel input. If numbers are below 0, you will need to check MAF numbers. At idle MAF should read around 17 KG/HR. FRA is adaptation under load and should be 1 or nearly so. Your RKAT value can deviate by about 4 and a half percent before setting a fault. Pretty sure when fuel mixture is balanced O2 sensor will read around 4.5 volts. This info may be of little or no use without a Durametric. . edit:. Decimal point misplaced. Should read .45 volts equals balanced mixture....oops.

@lnengineering I considered replacing the injectors during the rebuild, but last summer the injectors for the 3.4 were $240 each. Luckily, it looks like they've come down from that astronomical price to a merely exorbitant $95 each. MAF looks pretty inexpensive and easy to DIY. I'll probably go ahead and do that next weekend. Which would you consider to be the primary O2 sensors - the ones pre or post catalytic converter? @jumpy chunky Thanks for the detailed info. My charts aren't pretty because I dumped the data into Excel and didn't spend much time on formatting. I should have left RPM out and taken shorter time window. In any case, I'm not looking to spend Durametric money on a scanner, but I wonder if I can back into the RKAT values from other readings? I can definitely get the MAF and O2 sensor numbers from my current setup and will compare those against the values you suggested.

Looking at the data logged from that drive, my MAF numbers (warmed up, at idle) are around .8 lb/min or 22 kg/hr. Is that a significant amount of deviation? The O2 sensor values for sensor 1 are reading .75v and sensor 2 around .12v and .16v. I wonder if I'm not reading those correctly.

Quotes source post: 16561995

Your MAF value is higher than it should be for a 3.4. Shouldn't deviate more than 1 or 2 KG. You may want to measure again or clean or replace MAF. Your O2 sensor value shows a lean condition. I bought new injectors last Summer for $76.00 each, from Sonnen Porsche. A sound investment for the health of your 996. I think following Charles' ecommendation would be wise, injectors, O2 sensors and clean MAF or replace. Used Durametrics can be found for $200.00 on Market place, every now and then.

Considering the cost of a rebuild vs the savings on new injectors is small money, Protect your investment!

I had some experience with fuel trim issues and they are very hard to troubleshoot without some sophisticated tools, such as gas analyzers. As you want to learn about your engine, I suggest a slower methodical approach is better than a "change everything at once" approach. The slower approach will let you learn about the working of the individual components. I did not read your other thread so I cannot tell why you want to do this. Is there a running issue? Is there a CEL code? Is there a hot glowing Cat? What leads to this point? The computer uses the front O2 sensor signal to control the fuel mixture. You should start there by observing the live reads of the O2 sensor signals. It should swing in the range of 0 to 1000 mV during the course of normal operation. 1000mV is a lean condition (a lot of Oxygen) and 100mV is a rich condition (not much Oxygen). Although, if you have an air leak in the exhaust near the O2 sensor, the O2 sensor will read more Oxygen and crank up the fuel which results in a long term +% fuel trim issue. Start there before you change O2 sensors, injectors, etc

Quotes source post: 16561966

Primary are pre-cat.

Check Pelican for Bosch injectors: https://www.pelicanparts.com/catalog..._pg1.htm#item1 I got the last Bosch MAF from Amazon at about $120. There are others there going for about $200. Hope you get it sorted.

Quotes source post: 16562749

Just make sure you buy any of these components from a trusted source. There are counterfeits on the market...

Quotes source post: 16562749

The injectors you linked are designed for the 3.6, OP has the 3.4. While it is possible to use the 3.6 injectors, some modification is necessary, just a heads up.

Quotes source posts: 16562775, 16562777

Yes. I've been careful to try and trace the sources. I'll need to check this. I recall reading about this in the past. More research is needed. Thanks for the heads-up.

Just a bit of info on injectors , currently got my engine out and injectors is one of my while your in there jobs!! I have a 99 C2 and the original style injectors are an older design and quite expensive , I was steered into some 3.6 injectors by someone on rennlist for sale on autodoc , original Bosch and like �25 each or something insanely cheap so I ordered them , they came in original Bosch boxes with the Q scan codes on them for authenticity and everything , there's a slight Mod you need to do though as the old injector has two grooves at the top , the top one being for the injector securing clip to connect it on the rail, the new injectors don't have this so you have to carefully cut one into the plastic, I did one yesterday and it worked ok, I could start a thread showing this mod if there's any interest, for me it seems a good way to get one of the bog worry's out the way on the older cars without spending a fortune , you hear so much now about bad injectors and fuel trims etc probably being the underlying cause for failures it's worth the price of mind and although easier with engine out is a job that can be done with the engine still in the car!

Quotes source post: 16563378

Yes, please. I know I've seen this mentioned previously. Did not note that significant details were provided.

Quotes source post: 16563378

Definitely interested as I'll be doing this sometime in the next week.

As stated, 3.6 injectors will work with 3.4 fuel rail. I purchased from https://www.germanautosupply.com/cat...-Fuel.Injector . They arrived with Bosch package security tape, and a way to validate the code on Bosche's website. You'll need to cut another groove in the 3.6 injectors. I used a "micro" dremel, with a cutting tip about the size of a sharpened pencil. Here's a link to a photo: left to right (all Bosch injectors): LEFT: 3.6 (link above) MIDDLE: aftermarket for VF Engineering Supercharger RIGHT: original 3.4 3 Fuel Injectors for the M96 If out of stock, try Warehouse33 (use code RENNLIST for discount) . The web apps' front-ends indicate their suppliers are the same (SSF)

Quotes source posts: 16562503, 16562675, 16562717, 16562777

Thanks for the details on the sensor values. Since the cost of the rebuild is so high, everything else seems small by comparison. All these little "insurance" items can add up. Charles' recommendation is $800 in parts, plus I'll need to pay my mechanic to replace the injectors - not a job I'm confident doing myself. $1k here, $1k there, and soon we're talking about real money I have a new short block with refinished heads and a bunch of other new parts including AOS and coolant tank. I don't have any CELs or problems with the engine (except it seems to idle a little low at times). This journey was sparked by the fact that my last block died of bore scoring, and Charles and others have pointed out that the fuel system could have contributed to that. I want to make sure it doesn't happen again. FYI - All you Mk2 owners may not realize that a few of the Mk1 parts are way more expensive than what you spend on your cars. Coolant tank is $600 (vs $200) and injectors are $95 ea (vs $35). Last year the best deal on genuine or Bosch injectors was over $200. It looks like FCP Euro has a close out on Bosch MAF sensors for $120. I may get that and the O2 Sensors as something I can do at home that isn't too big of an investment. Bosch O2 sensors are $50 each.

Ordered the sensors. We�ll see how long they take to arrive under the current circumstances. I also find this post on O2 sensors that I found very informative. It�s Boxster-centric and the parts info is a bit out of date, but I thought the description of how the sensors work in the first section was very helpful for understanding what�s happening. https://sites.google.com/site/mikefocke2/oxygensensors

I get it, I couldn't afford to have someone else do all my work, so I have to learn, but...If I can do it, anyone can!

Quotes source post: 16565672

I�m learning and taking care of the stuff I can. I did my own brakes, fixed the sunroof, and plan to do more. My issue is not having enough space and time for more complicated projects.

Quotes source post: 16566779

Most of us have all the time in the world right now. Glad your still working!

Quotes source post: 16566997

I count myself as extremely fortunate that I'm still working during the current circumstances and that I can do it safely from home. I know what it's like to be out of of job, so I empathize with everyone out there who can't work right now. That's why I'm glad to be working on a project that helps auto body shops accept payments online, so they can continue to work and maintain a safe social distance. As nice as it might be to have time for garage projects, I hope you and everyone else on here are back to work ASAP.

I installed the new MAF sensor this afternoon, along with a new air filter. There was no visible gunk on the old sensor, in fact it was hard to tell it apart from the new one. I let the engine come up to temp and looked at the live readings on my OBD scan app. It looks like my air flow readings are lower now (.6-.7lbs/min or ~17-19kg/hr) and my STFT is bouncing slightly negative to slightly positive, as expected. So maybe there was something off with the MAF sensor? I still see higher readings for LTFT (+6-8%). In fact they didn�t budge even a .1% over several minutes. Does LTFT only change once you�re driving, or is there potentially a bug with my scanner?

LT, is long term. It takes, as the term might imply, time. More driving and continued monitoring are advised. Not sure if there is a reset. Maybe the disconnection of the battery and touching the two leads together for twenty(?) seconds will reset the DME. Just going off the cuff on this last one...

Quotes source post: 16577845

�Long term� is relative. I�ve read that short term is sampled at less than a second. Long term changes take place in something like 29-30 seconds. Not sure what correct answer is for these engines or if it matters whether you are driving or not.

Quotes source post: 16577733

Ideally, you need to perform a vehicle handover to "reset" the learned fuel trims. It will adjust, but it will take time to correct itself if you don't do a vehicle handover. I do not think you can do a VH with a Durametric either - that's something you need a PIWIS for.

Quotes source post: 16578888

Can you explain this in more detail when you have a chance?

Quotes source post: 16579222

There's nothing more I can really explain. A vehicle handover resets the car to like new settings, erasing any learned adaptations. We mandate this when installing a new engine in a car. Learned adaptations on an engine with prior failure can cause damage to the new engine. Likewise, if sensors are bad and are replaced, the adaptations are based off readings from the bad sensors. YMMV, but a trick an old dealer tech once showed me was to disconnect both leads of the battery and using a 12 v test lamp, to connect the positive and negative leads together (remember, battery disconnected and no trickle chargers installed, etc). This drains the residual voltage from all the vehicle systems and will in effect, carry out a vehicle handover.

Quotes source post: 16579355

This last bit, is similar to what I suggested earlier. Is there a known minimum period of time the two leads should be in contact? Or is the test lamp there to show the job is done when it goes out?

Quotes source post: 16579375

When I saw it done, the light lit up for not even a second. Doesn't take long to draw the voltage down, but it does it in a safe and controlled manner. I would still prefer to do it with a PIWIS or similar tester that has that functionality.

Found another cool feature in OBD Fusion that runs an O2 sensor health check. According to some things I�ve read, the O2 sensor transition time (high to low) should be 200-300ms. Anything less than a 1s won�t throw a CEL, but higher transition times are indicative of an old sensor wearing out. That can lead to distortions in LTFT. Mine are at 480ms and 600ms, so definitely due for replacement. I just need my O2 wrench to be delivered so I can do it. Here is the OBD Fusion O2 sensor test output for anyone who is curious.

Resetting fuel trims:: Clearing fault codes sometimes resets the fuel trims depending on if a fault code is set and DME version. The Vehicle Handover does reset fuel trims along with a host of other things, but that is not what it was designed for. The cars are built and put in "transport mode" at the factory for the long haul across the pond and being trucked on car carriers and dealer lot. Transport mode has limit function off accessories or electrical consumers/electronic suspensions, speed limit reduced to 30mph, ect. During PDI (Pre-Delivery Inspection) when the car arrives at the dealership, or sometimes they remain in transport mode untill they are sold and the customer is ready to pick up, the car is "handed over" to the customer by running the program with PIWIS on the car that restores all comfort functions/accessories and speed limits and sets everything to "as new".. Disconnecting the battery for 30 min will reset all learned functions, or a "hard reset" as Scott and Charles mentioned will reset them. I have seen the "hard reset" be necessary to clear some really bad learned info that would not get cleared any other way. Doing as hard reset either way is safe as long as no high powered audio system is installed with huge amps/capacitors.

Quotes source post: 16579426

Is there any trick to invoking that function? I bought OBD Fusion for another vehicle (Land Cruiser) and never thought to try it on my 996es. I just fired both of them up yesterday after a long winter's nap and the 2000 had a check engine light. I seem to remember chasing fuel trim issues last fall, but honestly forgot all about it. I also just did the battery unhooked reset Porschetech3 suggested. Would love to eliminate that light before summer driving season starts (my wife enjoys driving the 996.1 on occasion)...

Quotes source post: 16580119

No special trick. Once you are connected to the vehicle just go to Monitors on the home page and then select the O2 Sensors tab at the bottom. It takes a while to gather the data. I thought it had hit an error, but if you are patient it will come back with results in a minute.

So I tried to replace the O2 sensors today and I can�t get the *#%^ things loose. I bought a special wrench for the job. Tapped on the sensor to loosen it, applied PB Blaster. Tapped on it some more. Went for a drive to heat up the exhaust. No luck. Any tips for getting the sensors out short of destroying them? You know I�m desperate if I�m using PB Blaster. I hate that stuff.

This is much more effective than PB and it doesn't stink as much. Unless you leave the acetylene on too long... You'll want to drop the exhaust system first.

Quotes source post: 16580206

Patience, the right wrench and penetrating lubricant. It is easier with the engine dropped, but doable in place. I prefer Aerokroil over PBBlaster these days, but both work, given time.

Spray it with the PB's a day or two in advance. Acetone and transmission fluid works wonders too...

Quotes source post: 16579355

Very helpful, thanks!

I've applied PB Blaster 3 times over the last few days and haven't been able to get the sensors to budge. Without a lift it's really difficult to swing a hammer or get more leverage. I'm going to give-in and take it to my mechanic. On bright side, I've driven a few more miles and I'm definitely seeing LTFT values coming back into line. So I think replacing the MAF sensor had a positive effect. I've gone from Bank 1/2 at 8.6/6.3% on the old sensor to 7.0/3.9% on the new sensor. I'd like to see both those values below 5% so I'll keep an eye on them.

Have you tried a 1/2" breaker bar with a 3' cheater pipe? I've broken loose a lot of bolts with just that combination before. It can generate more torque than my Snap-on MG725.(1200ft/lbs) For O2 sensors I normally use a socket similar to this: https://www.autozone.com/test-scan-a...B&gclsrc=aw.ds

Success! @TexSquirrel Your suggestion to try the crowfoot socket was a good one. Thanks! I also bought a 30" cheater pipe, but didn't have enough room under the car to use it. Luckily, with the crowfoot socket I could use my 18" breaker bar and that was enough to get the job done (at least, after 3 applications of PB blaster in the last week). My son wants to help with the O2 sensor replacement, so I left the old ones in place after I broke them loose. I'll post an update once the new ones go in. For now, a few more miles driven on the new MAF sensor has resulted in the LTFT getting closer to where I want it. Today it measured 6.3/3.1% (down from 8.6/6.3% before the MAF sensor replacement). Still looking for sub 5% on both banks.

That’s awesome.

Got the new o2 sensors installed today. Not too difficult once the old ones were broken free. Here�s the old one on the left and a new one on the right. I warmed it up and ran the o2 sensor diagnostic on OBD Fusion again, but didn�t really see any difference in the data gathered. However, everything checked out, the car runs and I didn�t cause a CEL. So I�ll call that a victory.

Quotes source post: 16606272

How long/far did you drive it?

Quotes source post: 16607506

just long enough to get it fully up to temperature.

Quotes source post: 16608818

So about 5 minutes/5ish miles? I think you should drive it longer.

Got some more miles on the new sensors. LTFT on Bank 1 continues to come down and is now at 5.5%. So close to sub 5! Hopefully a few more miles will see it come down a bit more. Bank 2 seems to have stabilized around 3%, which I can live with.