VE "dip"

[Calan]

Not a problem, just change those two cells to 134.5 and 157 in your VE map. Lean spike eliminated.

I know right?


What could injectors do to cause a spike like that, and a dip at 3500? The latency would have to be off the charts to make that much of a difference. Notice that fuel pressure, voltage, airflow, and everything else is rock solid. The car just decides to spike ultra lean at about 2800 rpms for some reason. (And it truly is going that lean; my ass tells me, along with the torque curve).

The only thing I can think of is maybe the injectors have some really bizarre looking flow ramp or something...but I still don't buy it, because this anomaly is too localized to a narrow area of the table, and it's RPM-based... not load based like I would expect an injector issue to be.

Besides that, there are just too many people running them with good results.

Someone send me a set of loaners and I'll swap them out and see what happens.

 

[Sam@RRE]

Looks like injector to me. The other people with "great" results could just be happy with even though its not working right. I get people who come in with dead knock sensor and boost leaks and they think their cars are running "strong". People with vented blow of valves saying the car runs fine and smooth.

Take this hypothetical perfect scenario:
If your RPM increases at a steady linear rate, boost stays constant at 0psi, load stays constant. Airflow will increase at a steady linear rate corresponding to RPM. Thus fueling should increase at that same linear rate corresponding to RPM if you want a constant Air to Fuel ratio of 14.7:1 right? Airflow increase, thus fuel needs to increase at that same rate.

A lot of injectors have a dead spot in which an increase in injector open time will have no extra increase in fuel delivered. (ie. 2.2ms = 122cc fuel, 2.5ms = still 122cc fuel) due to a design issues.

 

[Calan]

I see your point Sam...but this amount of error almost looks like the injectors would have to be shutting completely off for a split second around 2800 rpms. The error in AFR is around 50% or so.

(The people I was referring to know what they are doing...not some kids that don't understand what a BOV does.

You guys have an extra set of injectors you want to loan me for an experiment? I've spent enough $$$ with you over the years to be trusted.

 

[gofer]

Shoot me a PM Craig, my set of 950's can be loaned if you honestly want to experiment with this issue you're getting.

 

[Sam@RRE]

There can be a spots you can actually have less fuel with more duty cycle. Everything in your provided log proved this, every value that goes into fueling remained linear, but actual output value was different from the the ecu would expect. Unless the engine actually consume more air due to some weird venturi effect in the runners, or if the injector's output flow isn't linear corresponding to open time, output wideband value should be linear.

Take into account the injector stops flowing more fuel, and the engine requiring more fuel. Your airflow increases 18% from 2400RPM to 2800RPM, your IDCs increase 21% in that same RPM range, why is AFR error 52ish%? The values the ECU spits out to the injector seems correct, but what actually happens in real world is not. Sounds like the injector isn't doing what its told. BAD injector! you naughty naughty injector you!

 

[Calan]

Well you have me convinced to try some different injectors Sam. Everything you've said makes perfect sense, and I sure as hell can't think of a better (or any other) explanation for it.

"If you eliminate the impossible, whatever remains–however improbable–must be the truth."

 

[Calan]

Isn't Scott (Liquidx) running almost the exact same setup and injectors that I am?

Come on Scott... get in here and post up your VE table.

 

[Calan]

I just had another thought (it was bound to happen eventually).

What if this is ignition related? Since even drastic changes to the VE don't have much effect (and since both cells on either side of 2800 rpms aren't nearly as far off), I'm thinking I may be getting a momentary misfire or weak spark across at least two or more cylinders at that specific RPM, causing it to go lean.

Sound feasible? If so, what in the hell could be causing it? A glitch in the tach signal maybe?

 

[Ravenous]

I would say that is a possibility but our ignition systems are completely different yet we have the same "dips." I do have cheap wires and a corroded coil connector on #1 so if there were an issue with your one of your COP coils it could be, I would imagine with your setup though that spark energy should not be an issue.

 

[Sam@RRE]

You would feel a miss. It either fires or doesn't and why would it only be at that specific RPM. Ignition is more likely to fault under more load and high rpm. More pressure in the cylinder requires more energy to create an arc. And high rpm doesn't allow for much charge time before the next spark event, especially with DSM wasted spark, giving the coil only half a revolution of dwell time, roughly 4.5ms at 7K RPM I think if I calculated it right..
I've had a car miss specificly at 3000RPM on AEM and that was due to injector phase timing.
I've had 2G's with swapped ecu's/harnesses/6bolt/different cam sensors misfire due to not swapping the injector pins/checking or unchecking the "use non 95-96 CAS" option.
Also, a lot of cars have that constant miss every now and then at idle (low load and rpm) due to injector dead times.
I don't know man...everything just sort of points back to injector relative things.

 

[Calan]

Yeah... I didn't really believe it was ignition-related while I was typing that; I'm just grasping at other possibilities (straws).

I'm working on a set of loaner injectors to try out.

 

[BoostedBeaver]

Any luck on the loaners??? I am very interested as these are on the list of upgrades.

Let us know how the "dip" pans out.

On another note, I received the catch can that I ordered and it is a gorgeous piece. The welds are magnificent and quality is hands down the best I have seen. I could not have asked for better service and just want to thank you for the business.

 

[Sam@RRE]

I've been on vacation but....Any updates on this? Get a chance to borrow Corey's injectors?
Or just throw in some stockers and stay out of boost. Just to see if that dip is still there.

 

[Calan]

Not yet. I've spent a lot of time scouring logs and looking for anything else that might explain it, but so far injectors are the only possible cause that I can see. It's always between 2500 and 3000 rpms at varying low loads (usually really close to 2800), and seems to be centered around a fairly consistent range of pulse widths (about 2.5 to 3.8ms)... which further backs up your theory of the injectors just flowing less at that particular point. There are some variations, but it could be logging anomalies or latency. I don't know how wide the affected pulse-width range would typically be for injectors with this issue...do you?

The lean spike is a PITA though as it's right at the shift point for DD cruising...and I can definitely feel the torque dip at the beginning of each gear unless I shift higher and keep the RPMs up over 3000rpm all the time. I switched to E85 and it's still there, although it doesn't appear to be quite as extreme as it was on pure 93 pump. I'm guessing this is due to the operating range (IDC's) of the injectors shifting upward a little because of the additional fuel requirement with E85.

I've talked with Corey, and he is going to send me the loaner 950s when he gets back from vacation, if I can't find some locally before then. Unfortunately I no longer have any of the other injectors that I used to; in fact I can't even find anyone with stock ones around here.

@Robert - glad you like the can. Please let me know via email how it works out for you.

 

[Calan]

I met with Jens from FIC at the shootout, and he is kindly sending me a set of Bluemax 1350s to swap in.

I hope this resolves it...otherwise I'm all out of ideas.

 

[LiquidX]

Isn't Scott (Liquidx) running almost the exact same setup and injectors that I am?

Come on Scott... get in here and post up your VE table.

My VE table is U G L Y! Remember, I'm a 2.0 not stroked like you Craig.

Enjoy...

You must be logged in to view this image or video.

I'll take any advice on how to severly clean it up. I'm about ready to steal either Corey's table or Sam's table and work with that.

 

[Calan]

Remember, I'm a 2.0 not stroked like you Craig.

Yeah... I always forget that for some reason.

 

[Calan]

Well it appears that for some reason, my lean spike issue might have been a failing CAS.

While waiting on the injectors, I parked the car for some minor stuff a couple of weeks ago, and when I went to start it back up it ran like ass. After a few days of troubleshooting, it turned out that the CAS had died, even though the symptoms didn't immediately point to that.

After swapping the CAS and getting the timing dialed back in, the lean spike at 2800 rpms appears to be gone. I still have some more tuning to do and still plan on swapping the injectors, but so far it looks like the CAS was the main culprit...somehow.

 

[gofer]

I was going to mention Craig, I had a failing alternator that created a similar effect to my VE table when I was trying to maintain STOICH locked in open loop in that RPM range. I was pissed at first because I thought you cursed me since it was right after you created this thread.

Scott if you'd like I can attach the .eda file so you can have a nice smooth VE map to work with, then just make minor adjustments and smooth it out as you go. Just let me know.

 

[Calan]

I was going to mention Craig, I had a failing alternator that created a similar effect to my VE table when I was trying to maintain STOICH locked in open loop in that RPM range. I was pissed at first because I thought you cursed me since it was right after you created this thread.

Scott if you'd like I can attach the .eda file so you can have a nice smooth VE map to work with, then just make minor adjustments and smooth it out as you go. Just let me know.

Post yours up Corey... and someone with a 2.3l also just for comparison.

My VE table is U G L Y! ... I'll take any advice on how to severly clean it up.

Scott, you should be able to extend the open loop thresholds in Link, and then use the SD CombinedFT tool to smooth most of that out below 16.5 apsi.

After that... turn your boost to minimum, do a pull or two, and work on the next few rows. Up the boost a couple psi, rinse, repeat.

 

[LiquidX]

I uploaded Sam's table and have been working with that one. Everything is going well now on mine. I need to do some pulls to get that spectrum taken care of however.

 

[gofer]

Mine is basically Sams VE table tweaked for my car since he's the one that tuned it on the dyno.

 

[MY1GDSM]

My VE doesn't have any Dip @ 3.5k (or spike at 3k) Also I'm not a 2.3l Either

But, the VE in this one of Sams posts Looks a lot like mine too...

Except, I built my from scratch for the Talon

 

[turboglenn]

I have looked back on almost every old map i've felt worth keeping from both AEM and haltech systems and all the daily driven cars with mild cams seem to want a extra gulp of gas at 2500 then lean back off by 3k at the light load areas.. is this similar to what you're seeing? I've kinda lost track f this thread

 

[Sam@RRE]

a 2.0 vs a 2.3L shouldnt matter at all.
Speed density works off given constants and measured variables.

Take this for instance. A 2.0L engine at 14.7PSIa at 3000RPM can theoreticly consume 6000cc's of air at 100% VE per given moment. Lets throw in a poor design head and crap cams and say that at that same RPM and pressure, the engine has only a 40% VE, the volume of air consumed will be 2400cc's. Now lets boost this thing with 14.7psi of boost. So a 2.0L at 29.4PSIa @3000RPM and a VE of 40% will consume 4800cc's worth of air. Now lets take heat into consideration. When air heats up it expands, no more actual air, it just takes up more room (think of your tire pressure hot vs cold, no more actual air in there but pressure went up). So lets say that the air under boost is hot, say around 150*. Some of that pressure is not compressed air, but just pressure from hot expanded air. Lets say that the temp increase makes up for 2psi, roughly 14%. So if 14% of the boost pressure is just heat, then only 4464cc's worth of air is being consumed.

Now after all that, lets change the displacement to a 2.3L. You just change one constant variable, VE on the two engine are still the same for the most part because thats denoted by cams, valve size, port design, intake/exhaust manifolds, turbine wheel. Anything that effects the way the engine breathes.

Yes the 2.3L will consume more air at the same given RPM X boost X tempurature and the ECU will already take that into account when you select 2.3 in the drop down menu. But if nothing else changed in the engine other than the displacement, then VE generally would be the same. Only thing I would see effecting it is if the flow characterisics were really hindering the 2.3 and only allowig it to consume 2.0L values of airflow, the VE would drop.

When you select a value in the VE table you are denoting how much air actually makes it into the cylinder. The ECU already has the Displacement value, its measuring the air density through the MAP and AIT, and speed of the engine from the CAS. It just needs one other value to determine how much air in theory is in the cylinder, VE, then look up the target air fuel ratio, and determine an injector pulse width to make it happen. If it didnt come out right, then you have to determine if the the wrong amount of air was calculated due to wrong VE value, or if the the wrong pulse width was determine due to improper injector values.

And for Craig, I can see how a faulty CAS would effect fueling, if the ECU was seeing a lower RPM value than actual RPM, then the Speed part of the Speed-Density calculation would be off. (granted you are using a 6bolt with a 1G CAS that sends the signal for the crank and cam sensor)