Low Fuel Trim's Don't Make Sense, Front/Rear Polar Opposite

[Blurr]

Situation: I'm doing the initial tuning setup on my 91 VR4 an a Jester/Chrome ECU.

I'm only focused on getting the trims where they need to be. The latencies are dialed in (acquired from PTE) and I've been adjusting the injector size so far to get within that +-5% range. I haven't touched any other settings at all. The mid's are getting closer after adjusting the injector size a few times, but I'm getting very concerning front & rear low trim readings via Evoscan.

Front Low Trim = -12.1
Rear Low Trim = +10.1

Just for contrast, my mids, while not perfect yet are Front = -8.8, Rear = -4.7

I've been adjusting the injector size a tick or two at a time from an initial setting of 550cc. Like I said, I've only been adjusting the injector size. While the mids are getting better after flashing a new change (and resetting the ecu), the lows have always stayed around -12/+10ish area.

What am I looking at here; what am I missing? What are some of the causes of this? The lows do adjust themselves after resetting the ECU at each 4 minute interval, it's just that they eventually just go back to where they are now.

I did check to see if the primary O2's need to be switched as I've read this can be a symptom for. Besides not having jumpers that control switching on the ECU, the front and rear O2's when disconnected properly stopped reading in Evoscan for their respective fields. Under normal conditions the O2's cycle as expected (plus they're new). Oddly though, while idling at operating temperature with the O2's disconnected, the car didn't respond any differently; I don't know if that means anything.

Log: https://docs.google.com/spreadsheets/d/19n-QduzYejNGQuYKL0DaHb0htc0QZvO4TqxDsrzizik/edit

Setup: 3.1L, 15g's, Walbro 255, PTE 550cc's, Chrome v2, stock fuel filter/rail/fpr

Considerations: Compression test showed equal good compression over all cylinders, injectors cleaned and flow tested by PTE, new O2 sensors, new plugs, good wires, new oem upper & lower intake gaskets, intake bolts torqued to spec.

[Greg E]

Any cracks in your exhaust manifolds or leaks in the gaskets before the o2 sensors?

[Blurr]

I don't know about the manifolds (yet) but the turbo to pre-cat gaskets were replaced when the turbos were installed.

I don't have any specific reason to believe that there is a problem with the exhaust before the O2s at the moment.

[RealMcCoy]

Have you done a pressure test? Without have a starting baseline, all you really know right now is that your rear bank is 22% leaner than the front. Could be a big vacuum leak.

[Blurr]

I'm doing a pressure test on Sunday just to double check but I'm struggling to see how I could have a vacuum leak that isolates one bank over the other. If I wasn't so diligent with correctly reinstalling the upper and lower plenums, I might suspect a leak at one of the gaskets. That's all I can think of which *might* allow me to have a leak for one bank and not the other. Correct me if I'm wrong, please.

I do have a stock fuel loop and FPR. Is it possible that the rear bank at idle is experiencing fuel pressure / delivery problems that wouldn't be there in normal or aggressive driving conditions?

Edit: I should really get a fuel pressure gauge...

[Greg E]

I could see FPR over run causing this issue. Not from the rail loop but just from too much flow from the pump at idle.

Is your pump hotwired?

[Blurr]

Did some pressure testing this past weekend, we found some interesting things.

Before that though, the fuel pump is hotwired. I'm not sure what you mean by over run exactly. I don't want to guess.

We found 3 intake leaks and patched them up. We were able to change the injector size down and get the mid trims more in line but I'm still having the same problem with the low trims. Things looked promising at first but eventually the trims settled on -12/+10 again. The low trims on the new log look the same as the old.

I called PTE directly and got the exact latency voltages/ms values for their 550cc injectors. It turns out that the values that their injector specialist gave me are much different than the those mentioned here (http://www.3sgto.org/tuning-engine-electronics-ecu-discussions/13782-pte-550-tuning-issues.html), which is what I was using.

I'm going to plug the new latency values in and see how it affects things tonight.

[murphys_law]

Did some pressure testing this past weekend, we found some interesting things.

Before that though, the fuel pump is hotwired. I'm not sure what you mean by over run exactly. I don't want to guess.

We found 3 intake leaks and patched them up. We were able to change the injector size down and get the mid trims more in line but I'm still having the same problem with the low trims. Things looked promising at first but eventually the trims settled on -12/+10 again. The low trims on the new log look the same as the old.

I called PTE directly and got the exact latency voltages/ms values for their 550cc injectors. It turns out that the values that their injector specialist gave me are much different than the those mentioned here (http://www.3sgto.org/tuning-engine-electronics-ecu-discussions/13782-pte-550-tuning-issues.html), which is what I was using.

I'm going to plug the new latency values in and see how it affects things tonight.
Over run as referred to here is the fuel pressure going past(higher) the factory 43.5psi due to the regulator not being able to lower the pressure enough.

Sent from my A0001 using Tapatalk

[Chris@Rvengeperformance]

You can't use the manufacturers latencies and expect it to work. The conditions they come up with these values are way too different to be useful data.

[Greg E]

Before that though, the fuel pump is hotwired. I'm not sure what you mean by over run exactly. I don't want to guess.

Bet you a taco if you went back to stock wiring, the trims will behave normally.

[murphys_law]

You can't use the manufacturers latencies and expect it to work. The conditions they come up with these values are way too different to be useful data.
I understand you're using chrome so I'm sure this is experience speaking, but this statement doesn't make sense to me.

I would think the latencies stated by manufacturer should get you damn close.

Latency is simply the time it takes for an injector to open @ x battery voltage.

All tests I've seen are done @ 3 bar pressure with a calibrated test fluid.

Does this have to do with an intricacy of the Mitsubishi control strategy?

Just trying to learn, so excuse my ignorance.





Sent from my A0001 using Tapatalk

[Chris@Rvengeperformance]

I understand you're using chrome so I'm sure this is experience speaking, but this statement doesn't make sense to me.

I would think the latencies stated by manufacturer should get you damn close.

Latency is simply the time it takes for an injector to open @ x battery voltage.

All tests I've seen are done @ 3 bar pressure with a calibrated test fluid.

Does this have to do with an intricacy of the Mitsubishi control strategy?

Just trying to learn, so excuse my ignorance.





Sent from my A0001 using Tapatalk


For one thing you have to consider the mitsu ecu latencies are based on battery voltages. The injector companies rate on injector voltage, which due to the resistor pack is not what they see.

For another thing the mitsu ecu has a hidden constant for closing time that would have to be subtracted from the injector companies test results.

For another thing our injector drivers will behave differently than a test bed.

[CoopKill]

And if you contact the manufacturer they will tell you that they can be quite a bit off depending on the environment they are being installed in. (voltage/ecu/wiring/...)

With 800cc injectors I had to use close to what the manufacture recommended for thier 1000cc. They are tested in a perfect environment, not on 20y/o cars.

[bboyalan]

Is your ECU set to California or Federal specification?

Front and Rear Low Trims exactly opposite (http://www.3si.org/forum/f103/front-rear-low-trims-exactly-opposite-715018/) <<

[KeithMac]

As above, I've read a few times on here about opposite trims and it always cam down to front and rear bank O2s swapped.

Saying that they would normall max/min out?.

[Chris@Rvengeperformance]

As above, I've read a few times on here about opposite trims and it always cam down to front and rear bank O2s swapped.

Saying that they would normall max/min out?.

You don't hit the mids as often, maybe he just hasn't driven enough to max them.

[CoopKill]

I had to swap my o2 pins as well. I would definately start there.

[Greg E]

I did check to see if the primary O2's need to be switched as I've read this can be a symptom for. Besides not having jumpers that control switching on the ECU, the front and rear O2's when disconnected properly stopped reading in Evoscan for their respective fields.

Guys....

[CoopKill]

BooMM! :thumbup:

[KeithMac]

I'll get my coat...

[Blurr]

Is your ECU set to California or Federal specification?

Front and Rear Low Trims exactly opposite (http://www.3si.org/forum/f103/front-rear-low-trims-exactly-opposite-715018/) <<

That was the forum post that initially lead me to be believe that my O2's do need to be switched.

The ECU is for sure a Version 3 Fed spec 91-93 VR4 etc ECU. To be overly clear, it's specifically this ECU here, Chromedecu.org | 91-93 US VR4/91-98GTO/91-95 SL Fed Spec Version 3 Plug and Play Flash Ecu Installation (http://chromedecu.org/?page_id=1573)

This is the only board that has that particular jumper layout, the Version 3 board. I spoke with Jester via email and he confirmed that the ability to switch the O2's was removed.

Based on the VIN and the Information plate in the engine bay (unless the wiring harness was switched out at some point), I definitely have a Fed spec car.

I'm still willing to believe that the O2's might need to be switched. I'm also willing to believe that the Version 3 ECU by default has the O2 plugs switched (incorrectly) and EvoScan is simply representing them as being correct. Hell, I wouldn't even be surprised if some idiot swapped in a cali-spec wiring harness or something weird like that; this car has a seriously messed up history.

Can someone tell me which would be easier to do, swapping the signal pins on the wiring harness (at the ECU clip), or extending the O2 plug wiring to reach from the back to the front and vice versa?

[Blurr]

Over run as referred to here is the fuel pressure going past(higher) the factory 43.5psi due to the regulator not being able to lower the pressure enough.

Sent from my A0001 using Tapatalk

I've considered this too. I can't say that I know what the true working condition of the stock FPR is but even if it isn't the cause of my low trim issue, I still want to replace it.

[Blurr]

For one thing you have to consider the mitsu ecu latencies are based on battery voltages. The injector companies rate on injector voltage, which due to the resistor pack is not what they see.

For another thing the mitsu ecu has a hidden constant for closing time that would have to be subtracted from the injector companies test results.

For another thing our injector drivers will behave differently than a test bed.

Speaking of the latencies, after plugging in the manufacturer recommendation, it ran insanely rich. No amount of time letting the trims cycle mattered. I reduced the Injector size value by several ticks and it still didn't matter.

I'm going to go back to the latencies that I had plugged in before, the ones that other forum posts have cited. Even if they're not perfectly correct, they aren't nearly as off as what PTE gave me. I can't at least work with them.

[Blurr]

As a consideration, perhaps the O2's really do need to be switched despite the Version 3 board not having the ability to via the jumpers or software and all other evidence to the contrary.

Since I have a Version 3 board, which appears to be mentioned very little so far on the forums, it might be a good test to see what happens when they are switched. There are questions like "Will EvoScan still show the appropriate O2 disconnected regardless of which clip they're connected to?" that perhaps we need answers to.

If it solves my problem then we'll at least know and have a record of the issue. Maybe it will be something that Jester needs to address in future versions.

I have time this weekend so switching them is what I'll be doing. I'll report back either way.

[Chris@Rvengeperformance]

Once you get the plastic safety tab off on the ecu plugs it isn't too hard to move some pins around, but its a bit of a pain. I don't enjoy it, especially if you end up putting it back. I'd be tempted to take the lazy way and cut the wires for bullet connectors you can swap at will, but that's easier for me to say since I have an adapter harness.

I don't know why your evoscan values would be swapped though.

[RealMcCoy]

Just pull a front injector wire, and see if the front bank goes lean....

[CoopKill]

Beat me to it!

Also be sure that the setting in chrome is using both O2's

[RealMcCoy]

Just pull a front injector wire, and see if the front bank goes lean....

Just to clarify what I mean by "goes lean", confirm that both the front O2 sensor responds to leaning out the front bank, and the front short term trim responds to the front O2 showing lean.

[Blurr]

Just to clarify what I mean by "goes lean", confirm that both the front O2 sensor responds to leaning out the front bank, and the front short term trim responds to the front O2 showing lean.

This is a good test, I will definitely do that. Thanks, hadn't thought of that. All things considered, I strongly believe that it will but I'll test it anyway.

[KeithMac]

Have you got a spare injector you can plug in and lay on top of the engine (to avoid a cell) ecu may freeze O2 correction if an injector fault is detected?.

[tom93vr4]

Any further results on this issue?

I'm having essentially the same problem, except my Front Low Trim is high (+8) and Rear Low Trim is low (-9). Tried all sorts of things; looked & listened for exhaust leaks, swapped injectors front to rear, pressure tested the intake with a smoke machine looking for leaks, new O2 sensor. As with the OP, unplugging an O2 sensor causes the corresponding EvoScan reading to change drastically so it does not look like they are swapped.