VE at max value but I need more fuel (suspect bad injectors)

[servicesoon]

This is a 2.0L 4G63. The first time tuning I ran out of fuel. The VE was maxed but it was still lean. This lean issue started to rear its ugly head starting around 120KPA.

The attached cal file and log is setup as 2.5L in the settings so I could add more fuel. With these settings at 180kpa I'm already at 133% VE, CL Correction is at 20% and it's still lean. Checked spark plugs many times. They all look similar. I can tell that some are slightly more leaner/richer. I suspect bad injectors but...FIC1650 injectors were cleaned & flow tested about 12 months ago. They flowed 1554cc @ 300kpa. 3-D battery offset latency setup based on FIC data when injectors were purchased 15 years ago.

Individual Haltech IGN1A coils. Trigger scope looks good to me.  Strike on spark plug grounding electrode are all in the same area. I've swapped plugs. I've swapped coil packs and spark plug wires.

During boost, coolant temp is steady, IAT temp is steady, differential FP steady, battery volts steady.

Fuel is E11. It does have flex fuel capability, but I turned all of that off to eliminate as a possible culprit to this issue.

I cannot figure out why I keep running out of VE adjustment in the main fuel table?

Please help, my next move is to spend $1,200 on new 1700x Injector Dynamic injectors, I don't know what else to do.

 

[K4Tuning]

Hey taking a look at your log, Trying to understand exactly what you have going on. I see your fuel table at 133%VE, about 12psi and your probably sitting around 12.0 afr or so, which isnt too far off. When you say your VE is maxed out, you mean if you enter a value of say 150? the fuel does not richen up? Im seeing your injector duty cycle at around 50%which means as long as they are good they have plenty of head room still. The VE/ Modeled fuel mode is super dependent on correct information. I would go back to the correct engine size displacement and double check the basics and get them as accurate as possible. Just looking at your fuel table, it needs some work. Idle area your VE values should be anywhere from 45-65% depending on how everything is working together. Your lambda target table looks slightly lean also so maybe tweak that and see if the correction wants to follow. Hope any of that info helps and hope you get it sorted out.

[servicesoon]

@K4Tuning I appreciate you taking the time to respond. While the afr during boost is around 12, that is only achieved after CL Lambda Correction adds 15-20% additional fuel. When I tune a vehicle, I want to get the afr within plus/minus 5%. That means I need to add at minimum 15% to my VE table. My VE values are already 133% and adding 15% results in 152.95% VE, but the highest amount that I can enter is 150%.

At 15 psi the issue is even worse. Fuel values across the 15 psi row is 150%. AFR hovers around 13 with 20 CL Lambda Correction (That AFR is dangerously low). I need to be able to add more fuel, but VE is already maxed.

If I enter engine displacement 2.0L then this problem will be worse. The issue occurs around 4 psi of boost. By increasing engine displacement to 2.5L the issue occurs around 12 psi. I could further increase engine displacement to 3L but something is wrong because I should not have to do that.

My best guess for the cause of this debacle is that the injectors are not flowing what they should. They are defective. In vacuum areas everything functions as it should. The issue is only prevalent during boost, when injector flow becomes more important.

[K4Tuning]

41 minutes ago, servicesoon said:

@K4Tuning I appreciate you taking the time to respond. While the afr during boost is around 12, that is only achieved after CL Lambda Correction adds 15-20% additional fuel. When I tune a vehicle, I want to get the afr within plus/minus 5%. That means I need to add at minimum 15% to my VE table. My VE values are already 133% and adding 15% results in 152.95% VE, but the highest amount that I can enter is 150%.

At 15 psi the issue is even worse. Fuel values across the 15 psi row is 150%. AFR hovers around 13 with 20 CL Lambda Correction (That AFR is dangerously low). I need to be able to add more fuel, but VE is already maxed.

If I enter engine displacement 2.0L then this problem will be worse. The issue occurs around 4 psi of boost. By increasing engine displacement to 2.5L the issue occurs around 12 psi. I could further increase engine displacement to 3L but something is wrong because I should not have to do that.

My best guess for the cause of this debacle is that the injectors are not flowing what they should. They are defective. In vacuum areas everything functions as it should. The issue is only prevalent during boost, when injector flow becomes more important.

Its no problem, id like to help out if I can. Im pretty familiar with 4g63 stuff. It looks like fuel pressure increases as should, so I wouldnt expect the regulator to be the issue. Can you try just going back to a simple 2d injector dead time table? just to rule out some of that stuff. I love FIC injectors, the usually perform very well. I would still go back to correct engine data entry before trying to solve the issue with incorrect information. Durring the pull I only see your TPS reaching like 95% idk if this i related but I would recal your TPS sensor, MAP/BAP all the basics. I would probably set your lambda target table to like .750 Lambda or like 11.0 AFR in the boost region. Everything else in the setup looks decent. 

[Laminar]

Where are you measuring fuel pressure? Any chance that fuel pressure is restricted somewhere between your measurement point and the fuel rail?

It would be awfully coincidental for all injectors to fail in the same way at the same time, so it's not like you have one injector not running right and making that one cylinder extra lean and the AFR just reading the average of 3 good cylinders and 1 lean one.

Was this engine running with these same parts on a different ECU?

Have you tried turning Asynchronous Injection off?

[servicesoon]

@Laminar The Fuel pressure sensor is installed on the feed input side of the fuel rail, 4 inches from injector number 4.

I agree that it would be coincidental for all injectors to fail. I have explored that possibility by reading the spark plugs. The variance between the plugs is present but minor.

This engine was previously running (albeit not very well) on AEM EMS-4 with some gremlins. That ECU didn't have sensor ground. That ECU would randomly not power on. AEM was ditched and Link Extreme w/ onboard Lambda controller to get past those issues.

I have not tried Asynchronous Injection OFF. Doesn't that feature increase injector pulse width if the previous injection event could not disperse the entire amount of fuel that was optimal? I'm willing to try anything but fail to see how that would contribute to this VE issue.

[Laminar]

If fuel requirements increase between squirts, Asynchronous injection adds an extra little pulse. It's probably not at play here but I'm looking for anything that can mess with the fuelling calculation. 

[servicesoon]

@K4Tuning I have verified that the fuel pressure regulator matches what the Link says. The MAP sensor has already been replaced because I did suspect that was the issue. Correct MAP cal table was entered that came with the sensor. The TPS probably does need re-calibrated or the throttle stop plate needs to be adjusted. I would think that a 2D injector dead time table would make things worse but I am willing to try anything.

[K4Tuning]

3 minutes ago, servicesoon said:

@K4Tuning I have verified that the fuel pressure regulator matches what the Link says. The MAP sensor has already been replaced because I did suspect that was the issue. Correct MAP cal table was entered that came with the sensor. The TPS probably does need re-calibrated or the throttle stop plate needs to be adjusted. I would think that a 2D injector dead time table would make things worse but I am willing to try anything.

Just go back through and do all your "dummy" checks. Make sure the TPS is called proper, MAP/BAP match. I would go back to as simple as possibly before making things more complicated. Go back to the correct engine settings even if it maxes out the VE table. I would change to a 2d injector dead time table, and probable adjust the dead times. I did a quick search on data for the FIC 1650s and looks like your deadtimes are slightly off from what im seeing. Turn Asynch injection off for now, see if that does anything different. Give those a try ad see what happens. Let us know.

[servicesoon]

I have been over dummy checks multiple times. The MAP & BAP has already been executed. I don't believe the vehicle will start without performing this function.

Base ignition timing has been synced. I reviewed the Trigger Scope and it looked acceptable. I even changed the ignition sync value to make sure that the trigger 2 (cam) was not rotated 360 degrees out.

The TPS range is close enough that it would not contribute to the VE fueling issue I have presented.

I've already done everything you have suggested more than once.

I have had two professional tuners review the cal file and they both gave approval.

Something mechanically has to be the culprit.

Engine compression was 160 to 165 psi for all cylinders.

[K4Tuning]

Hmmmm, Sounds like a tricky one. Let us know if you come up with anything.

[servicesoon]

I will try the suggestions in this thread and report back. I appreciate everybody's input!

[Adamw]

Small deadtime errors arent going to make much difference to VE numbers at full load when PW's are relatively long, it is more a mass flow rate issue when VE is high at high loads.    It sounds like you have already verified most of the variables that will affect mass flow related calculations - the main ones being MAP sensor, fuel press, fuel density and stoich settings. 

Assuming you have confirmed there is full battery voltage reaching the injectors - no ballast resistor in the loom etc, then I would be suspicious the injector flow rate isn't what you believe it to be.  If you wanted to verify, one no-cost option would be to remove an injector and rig it up to the FPR temporarily, you can often just clamp an injector in an old length of 3/8 hose.  Obviously be aware of the risks of testing with real fuel.  Use the "Adv Test Inj 1" to test fire the injector for a specific pulse count into a measuring cup to determine flow rate.  If you dont have a suitable measuring jug/cylinder, you can really use any container and a kitchen scale to determine the volume. Typically I'd use 100Hz, 3000 counts (30sec), do a test at 3ms and another at 8ms so you have two data points within the linear operating range, plot the calculated flowrate (cc/min) from these two tests Vs PW on a graph, the gradient of your line (x 10) is flowrate, the intercept is deadtime. 

This is more accurate than doing a "100%" static flow test for a fixed time.  

[servicesoon]

 @Adamw I have not verified how many volts are reaching the injectors. I will perform that test tonight. 
 

I’ve tested various injector battery latency values. One of the many techniques I used was to target a specific estimated airflow at idle and then adjust the entire latency table to achieve two different afr ratio targets (say 15 and 13) that tracked with minimal Cl correction. I did the same procedure targeting a specific VE range at idle. Good to know latency is not that important at the engine loads I am having trouble with. 
 

I attached picture of injector battery latency from mfger 15 years ago when injectors were purchased and flow test done by mfger about 12 months ago. 

[servicesoon]

I entered 1554cc @ 300kpa into link cal but base fuel pressure on physical vehicle is 350kpa. I assume the link will increase the flow of the injectors (In the links calculations) due to the higher fuel pressure. I do have the fuel pressure sensor as an input into the Link.

[Adamw]

4 hours ago, servicesoon said:

I entered 1554cc @ 300kpa into link cal but base fuel pressure on physical vehicle is 350kpa. I assume the link will increase the flow of the injectors (In the links calculations) due to the higher fuel pressure.

Yes, the ecu will correct for the rated pressure vs the measured pressure.  

[servicesoon]

The battery is in the rear of the vehicle. Voltage across negative & positive terminal was 12.7

With key ON, all four injectors measured 12.4 with the negative probe grounded to strut tower.

Injectors are getting the appropriate amount of voltage.

I'm going to measure the injector flow this weekend as instructed by Adamw.

[servicesoon]

@Adamw I really appreciate your contributions! I will be conducting the injector test this Sunday. Are you willing to share the document (I assume excel format) you posted on Monday at 05:41 PM?

[Adamw]

Try this link: Injector flow and deadtime.xlsx

[servicesoon]

Picture of rig to test injector flow and results. Tested the delinquent injector twice. Verified plenty of fuel and pressure stayed at 300 kpa during entire test. 

 

[Adamw]

Good work.  Mystery solved I guess.  It's a good validation tool to have up your sleeve in future. 

Even the good ones flow and deadtimes are a fair bit different to the supplier's data.  Deadtime variation is understandable as driver hardware has quite an influence, but I dont trust a lot of the flow numbers I see from many suppliers.  

[servicesoon]

FIC received the injectors and said that all four were majorly plugged with black specs.

This vehicle sat for many years (5 to 10) with E85 in the lines, tank, etc. My best guess is that the ethanol degraded the SS rubber lines and that is the source of the black specs.

Every fuel hose is going to be replaced. The gas tank will be rinsed. The fuel filter cleaned (metal mesh filter). Then the newly cleaned and flow tested injectors installed.

[Adamw]

Once cleaned up I would still do your own flow test again to confirm as even the longest deadtime and highest flow in your testing is very different to theirs.