Brad92SE

iTrader: (9)

I'm hoping that someone might have some insight on a super-annoying problem that I have been having for a long time. I've searched online, and have read about people having similar issues but no definitive solutions.

Basically, every time I start my car it will run normally for a bit, but shortly the air/fuel will start to creep into being extremely rich. It will stay rich for a few minutes of driving, then eventually creep back to normal. Normal driving AFR will be around 13-15 on initial startup, then creep to 10-11 (and even get into the 9s if I give it any gas and stumble) for a bit. Seemingly randomly, it will eventually even out back to 13-15.

This problem happens regardless of engine temp or outside temp. I've been having this problem since before I wrecked my Max, so it isnt related to my engine swap in the Z. It SEEMS like the AFR will normalize faster if I keep the throttle position light and consistent, but that could just be in my head. It happens every single time I turn the car off and restart.

Possibly related changes:
550cc Deatchwerks Injectors
Emanage Blue
Disconnected factory O2 (plug to ECU just hangs out)
No emissions equipment
Z32 MAF
Factory VE ECU
Adjustable FPR

The injectors and CTS have been recently replaced, and I dont think it is the MAF. I have tuned the car with the Emanage to run in open loop without the O2. The only thing I can think of is the ECU is richening up the fuel mix for warmup, and for some reason doesnt realize the car is at operating temp (unless it is supposed to do this no matter the current coolant temp, and it is just dumping a ton of fuel due to the 550cc's). I guess it could also be related to not having an O2, but I'm not sure why it would start off ok, then get really rich, then normalize.

Does anyone have any info that might help? What do our ECU's try to do on startup, and is there any way to correct them?

vqmaxman

I saw a couple things that might be causing you your problems,the o2 sensor plug to the ECU, If that's what you stated can be what's causing your car trying to figure out what fuel/air ratio it's supposed to deiliver to the car. The normal factory settings is what your cars trying to do. If the o2 plug that's disconnected from the Ecu than the Ecu is trying to set a good balance. The open and closed loop that the Ecu goes thru before coming to a normal operating tempereture. The emanage blue can be interfering with the Ecu operation. I would lean towards the o2 plug for the Ecu.

nickdoof

iTrader: (2)

doesn't seem like a bad, or unplugged o2 sensor could cause you to drop from 13:1 to 9:1 afr.. that's huge

4signs

iTrader: (2)

I had similiar problem several years back. Check you Adjustable FPR and make sure there are no kinks in the fuel line.

DanNY

iTrader: (14)

why not? the ECU uses the O2 and MAF for controlling the mix ratio...if any of those two are not operating correctly then the ECU is going to take whatever data it gets (bad or not) and mix accordingly.

open loop and closed loop switching will change the AF.

Brad92SE

iTrader: (9)

It's the O2 plug that goes to the ECU. My understanding was that if the O2 is disconnected, the ECU would run at full rich in open loop, which is what I tuned to by leaning things out with the Emanage. When I had the O2 connected I know that the Emanage and ECU would fight each other during normal driving which is why I disconnected it in the first place.

It makes sense though that the ECU is getting confused though if it doesnt see the sensor there. Do you know if it is trying to do something specific (heat the O2, etc)?

nickdoof

iTrader: (2)

The o2 sensor doens't make that dramatic of a afr change... A bad maf could. That way if your o2 sensor goes bad, or or the wires get cut, or some other weird situation your car wont bog down and run like crap. You could go out and unplug your o2 and the car would still drive fine.

You should get a wideband o2 sensor though mate.

Maxpwer

iTrader: (7)

No offense, but do you have any experience tuning a custom setup? While your points might be valid for a stock car, we are talking about injectors almost 4 times as large, a MAF that reads almost twice as much air, and a piggy back ECU that can make changes never imagined by Nissan.

As far as the air/fuel ratios that he gives for various conditions, you think he is able to tell with his nose? Im willing to bet his nose isnt that sharp, and he has a wideband already.

Brad, on start up the stock ECU does two things to help the motor warm up. First, it runs in "fuel enrichment mode" because a "cold" motor (less than 180 coolant temp) needs extra fuel to run properly. Secondly, ignition timing is advanced.

For the O2 sensor you might be on the right track, the heater. Most widebands have an anolog output signal that can be hooked up to the ECUs 02 wires to simulate a narrow band. For O2 sensors like ours that have a heater, a resister is needed to keep the ECU thinking the stock O2 is still there. You just need a 20ohm, 10watt resister between the positive and negitive O2 heater wires. (If you can't find a 20ohm, I think radio shack sells 10ohm, 10watt resisters, just use 2 and wire them in series, not parallel. Either way these resister will get pretty warm, so be careful where you place them.) I would try this alone, and then in combo with the wideband's narrow output hooked up to the ECU and see what results you get.

One other thing you might want to try is taking the CTS out of the equation. Try various resisters in its place and see what results. With my JWT ecu, I found that the ECU would pull signifigant timing once the engine was warmed up, and it would run crazy sometimes. With a resister in place of the CTS, the ECU thought the car was cold and from the advanced timing, I was able to make almost 80 extra hp with no other changes.

Im sure there is more that I can't think of right now but hopefully this will get your started in the right direction.

vqmaxman

The o2 in a way is a device that helps the ECU to better understand the readings into telling ECU how to run the air fuel mixture lean/rich ect. The o2 is sending signals to the ECU. I am assuming it has some voltage that the ECU is reciving from the o2 and the MAF. It makes since disconnecting this o2 sensor is causing your car to run rich or low. The ECU is trying to find the right balance after the open loop cycle is complete when it"s coming into a closed loop it"s normal temp or normal running. The injectors that you have might be bigger than normal but the emanage blue and the ECU are controlling the pulses and the amount of the injectors pulses. The o2 wire being out the car is still feeling like it"s not running in normal conditions.

Brad92SE

iTrader: (9)

Thanks for everyone's advice. I dont have any kinked fuel lines that I can see, and I'm pretty sure my MAF is good since I've cleaned it really well, and it only does this enrichment right after I start the car and goes away and runs fine after. I do have a spare Z32 I could pop in there though.

I think I'm going to try to find a resistor this weekend to replace the O2 heater and see what that changes. That is key information I have been wondering about. Hopefully it will help, if not correct the problem. Part of me is wondering if this is just a result of a normal ECU process and the large injectors are just amplifying it.

nickdoof

iTrader: (2)

I have some experience with megasquirt but no where near the knowledge you have. I know I wouldn't be much help honestly. I thought I read that he had his o2 sensor unplugged completely and was re-tuning based on the enriched fuel map the ecu defaults to.

LvR

I think that's the answer.

While in theory you are running a closed loop fuel management system (or are you?) , the individual parameters making up the system were not selected to ensure fast and optimal tracking of the desired result. In engineering lingo there are terms like overshoot, undershoot, hysteresis, rate of attack and decay etc etc etc when designing control systems. You have a mix of components there that may enable you to reach the desired AF values under very specific conditions quickly, but under a different set of conditions it will be close to a crap-shoot.

I am not 100% clear on your exact situation though - are you

a) running with the std O2 on the ECU disconnected while running the blue on a wide-band O2? (what I would suspect the desired state would be) or
b) running with no O2 at all? from your comments I just cannot figure.

If that was my responsibility to start tuning and you have situation a) above, I would try and replace the blue wide-band with a narrow band O2 ................... it would prevent huge attack rates and limit (decrease) the response times since you will be preventing (to a certain extent anyway) the over sized injectors overwhelming the blue's undoubted response limitations by allowing only smaller changes per compensation cycle.

If you have b) above its gonna be a crap-shoot ito fuel management and nobody can tune that thing properly ever anyway.
IMO that's just not true. If you look at the EC+EF sections of the FSM detailing the O2 and the ECU connections, its indicated that the reference resistor the ECU uses for voltage divider sensing of the O2 element output is actually integrated inside the ECU and its connected to ground (0V iow) .................. if you now go remove the O2 the ECU can only interpret that only and single input from the O2 as 0V because of voltage divider sensing action inside the ECU - it can deduce absolutely nothing more from that single input from the O2 into the ECU. Buggering around with heater resistors will not change anything on that disconnected O2 input on the ECU because in the end the ECU is a simple analogue controller looking for a voltage signal, and the removed O2 results in a valid 0V signal that still falls within its expected range..................... apart from that the O2 heater is driven directly from the battery and not from the ECU at all.
Sorry but that's plain nonsense - if the O2 is either missing/disconnected or its output signal is such that the ECU sees 0V on the O2 input, then the ECU is running in full rich mode - its not trying to do anything because its got no idea what is going on with the mixture - as far as it knows its causing a way too lean AF and will as a result run in full rich mode - its not trying to set a balance at all.
That is a classic symptom of a completely mismatched set of process parameters, and the answer is not to remove the O2 - the only answer is to correct the mismatch.
Erm ............... if you start buggering around with fooling the ECU into thinking the motor is cold why even bother with AFs then - one cannot talk max power/torque or whatever it is you are chasing while at the same time bemoaning a funky AF due to your own disconnecting/bypassing of items influencing/controlling AF.

Bottom line is this:

If you want a AF ratio that makes sense on an injection based engine, you need a working O2 as feedback mechanism somewhere on the system. If you cannot get the blue to work properly in unison with the std ECU and O2 you may just as well go get a Rochester Quadrajet carb and slap that on the motor - its way simpler, way cheaper and don't need tuning of parameters - I bet it will produce as much if not more power/torque than the current blue assisted injection system with a missing O2 and will most likely also be better on fuel

Brad92SE

iTrader: (9)

(a) is how I'm running my setup. I have the OEM O2 disconnected from the ECU (actually didnt put a place for it in my new exhaust) and am tuning based on my wideband O2 output. There is no closed loop feedback in my setup... definitely a shortcoming and pretty much makes it impossible to fine-tune cruising AFR, but works well enough for WOT. I dont think there is a way for me to put in a closed-loop system without fully replacing the ECU or going Nistune or something similar (which is on my to-do list).

This is what I assumed and heard would happen (full-rich) when I disconnected the O2. I have tuned to compensate for it, but for some reason shortly after I start the car (warm, cold, doesnt matter) the AFR will slowly creep richer. After a few minutes of driving, it will seemingly randomly creep back to normal.

It was fighting me because I wanted to lean the car out a bit during highway driving. It kept wanting to pull back to stoich (expected), so I took it out of the equation. I dont know if there is another way to lean out during cruise while still keeping the factory O2, unless I either fool it into thinking its always running stoich or reprogram the ECU.

I'm leaning towards getting Nistune, which may hopefully solve most of my issues. Having a piggyback is a bandade for tuning I think, but it works reasonably well for the cost and effort needed to tune (at least at WOT).

LvR

So the wide-band is only there to give you as a human a reading/idea what is happening and you then do your thing with simple entries in a table/map and hope the blue can make it happen repeatably and reliably? .................... yep that's open loop control - and I guess its actually not control at all in your case due to the factors I mentioned. (NOT a) mentioned - definitely b))

Where exactly is the wide-band located physically?

My guess your miseries with the running rich after start-up thing is the following:

The injection system enters simultaneous injection mode on startup - ie it dumps way more than normal steady state running fuel to ensure reliable starting ..................... the O2 then requires a certain time to get rid of the "poisoning" effect (heating up and burning off the excess) of too much fuel before its readings start reflecting what is really happening in the exhaust flow............................ iow you are simply reading too much too soon into the O2s output.

The blue with open-loop and no O2 is undoubtedly OK for WOT - for tighter control over the whole operating range there is just no quick-fix band-aid answer.

Brad92SE

iTrader: (9)

My wideband sensor is about 18" downstream from the turbine outlet; the gauge is in the car on my A-pillar. You are right about how I'm making adjustments to the AFR. I agree that for part throttle driving conditions it's pretty much impossible to get everything perfect, since all values are in huge maps, plus I have to adjust the Emanage to change the MAF voltage (except for when in-boost... I'm using a MAP sensor), which will change timing, etc etc.

Interesting thought about the O2 just seeing an excess dump of fuel and it taking time to burn off. I will look into that further... however I think the car is actually running rich just because it stumbles if I get into the throttle while it is in the "rich running" condition and AFR dips into the 9:1 range.

Im looking through the FSM for how the ECU and factory O2 interact... it says that the ECU sends a voltage ~1V and reads the output voltage, which changes based on the sensor's resistance that goes up and down due to AFR. It then uses this output voltage to adjust AFR. Since there are only 3 wires - a battery + (I assume for the heater), ground, and a white wire connected to the ECU (pin 29) - what wires does the ECU use to receive and send a voltage? If it sends 1V through the white wire, does it check the voltage across that and the sensor ground? I'm not seeing how that circuit connects back to the ECU...

LvR

The description in text is up-to maggots imo. The ECU cannot both generate and sense on the same single wire. The sensor's active part (resistor changing its characteristics based on oxygen content its in contact with) is electronically mounted and connected in such a way so as to ensure that typical sensor failure will result in the ECU being able to protect the motor (read the requirement to run rich if the sensor wire is either broken or if the sensor's heater fail or even if the sensor itself fails).

While the sensing action is indeed a simple voltage divider circuit, the "reference voltage" is generated external to the ECU as part of the O2 heater circuitry. You can easily measure and confirm that to be the 1V they talk about in the diagrams there too - leave the heater connections on the O2 plug and remove/cut the wire from the plug that goes back to the ECU - refit the plug and now if the ignition is switched on you will be able to measure 12V on the heater wires, and at the same time you should be able to measure the supposed 1V reference voltage on the O2 sensor wire (sensor side) going back to the ECU wrt the ground wire.

That wide-band O2 of yours - is it also a resistive design (change resistance with O2 concentration change) or is it a crystal design (generating variable voltage with O2 concentration change)

WRT you using a MAP sensor - is that only to effectively measure vacuum or have you got an inlet temp sensor hanging somewhere to actually check on and measure proper intake mass flow volume for the blue's control under boost conditions? Hmmmm - if its running rich to the point of stumbling my suggestion as an explanation sucks - what I suggested is undoubtedly happening but it cannot possibly account for a "rich stumbling" situation - sequential injection is triggered soon as the motor is running - while I can see the simultaneous injection excess hanging around for a while (as in seconds) it should certainly not be "that bad".
Just to get back to that initial description of yours - how sure are we that the blue isn't itself causing this as a direct result of its own internal design logic?

LvR

Another thought as I sit here drinking a beer .................... since the "rich" issue only occurs after each and every start action, where exactly is your O2 circuitry "earthed"? .................. at exactly the same location as the ECU and the blue (ie the engine block/head or right on the alternator and not the body or even worse still in different locations) ............... the idea being that after start-up the alternator is charging the battery for a period of a about a few minutes with serious currents flowing and if you have not earthed the individual O2 and ECU items properly you may be setting up earth loops affecting the O2 signals/readings

Maxpwer

iTrader: (7)

You must be an Electrical Engineer or work in the field. These types of concerns are far beyond the understanding of even most car enthusiasts. However, the design of the wideband O2 sensor assumes this and the circuitry is setup to be accurate despite differently grounded parts. Now the OEM ECU wasn't designed with this assumption, so when using the wideband's narrow band output for ECU, noise filter capacitors and groundings must be as close to the ECU as possible.

LvR

Yes I am and yes I do have experience with old technology small signal analogue control systems.

As for the grounds - sorry I have to disagree again - non OEM O2 suppliers in my experience always make a huge thing of the proper earthing/grounding practices - most the time in their installation documentation but often only in fault finding forums etcI can go into the theory but I can very easily demonstrate its simply not possible to setup any control or measurement circuit as long as you are aware of the possibility that earth loops could be influencing your results - as soon as you have different earth points and huge non-related currents flowing through thin conductors you will have earth loop issues - guaranteed. The idea with using the exact same earth point is to prevent unrelated currents effectively generating differences in "earth" for individual circuits by keeping non-related currents from flowing through ANY part of the control related wires/cables. I am willing to bet the OEM ECU was definitely designed with these concerns in mind - its basically impossible to design a reliable control system while not adhering to these principles. If they are using non-common earth points then its because they knew the rest of the design of the vehicle's wiring harness and current flows - I certainly don't so I cannot argue with their choices 25/30 years ago - what I can however tell you is ALL of the 3rd gens are of that age and as such the assumptions made then are no longer valid for all earth points/joints/connections - think corrosion/cracks/broken wires/rusted panels/huge aftermarket sound/different battery mounting locations/capacitors mounted in the boot etc etc etc etc

Unfortunately this is also not true - the general and good engineering practice is to have filtering components as close to the source of interference as possible - this ensures that funky interference currents have access to the shortest possible cable runs so that their EMI "ability" is kept to the absolute minimum by having their "antennas" as short and as few as possible



Note I am not saying this IS the OP's answer, but I am asking the question because the logic of his problem's frequency of occurrence fits right into the possibility and thus far there is no other logical explanation bar a funky blue controller with a mind of its own

Brad92SE

iTrader: (9)

So I've been trying to drive it lately (in between rainy days) and pay really close attention to what happens when the car does run rich. I've gone WOT at <3500 RPM while it does, and like I mentioned before the AFR dips into the 9s.

I haven't noticed any misfiring or bogging like I had before... which leads me to believe LvR's thought above about the simultaneous injection at startup causing excess fuel to get thrown into the exhaust which takes time to burn off. I guess the misfiring/bogging I had before may have been unrelated, or as with most diagnostics, doesnt happen when you are actually trying to get it to happen.

I'm going to keep paying attention to the situation to see if there are any noticable drivability changes during this "running rich" situation, but for now I'm chalking it up to the problem above.

Thanks for everyone's help and advice!!

Brad92SE

iTrader: (9)

To post a final solution to my problem -

The car actually was running very rich after startup... while the wideband was reading extremely rich the car was bogging and I got it to misfire again. Today I hooked up my LC-1 wideband analog output (simulates the factory narrow band) to the ECU O2 sensor input, and had no more instances of the car creeping into the 10:1 AFR range while cruising after startup. So far so good, although the AFR does bounce around 14.7 quite a bit more during highway driving... although I guess that is better than the alternative of it running rich after every time I start the car up.

LvR

Its unfortunately just one of those things that bitter and expensive experience taught me as I said in the beginning...........................

So I guess the obvious question then:

What is it that you were chasing with the mod you did and how much of that goal/satisfaction are you losing as a result of the O2s being present again?

Maxpwer

iTrader: (7)

Thats great news! Did you hook up the resister to the O2 heater wires?

Brad92SE

iTrader: (9)

No, I just hooked the LC-1 output to the white wire in the O2 sensor harness which goes to the ECU. The other two wires are still not connected to anything. It feels great actually having it running consistently now!

To LVR - I initially disconnected the O2 so that the ECU wouldnt fight me trying to lean out the AFR during highway driving. I wanted to lean it out to improve MPG a tad. It worked, but obviously this problem came up. I suppose if I cared that much about it I would buy a Hybrid Or maybe I should just invest in a straight-cut gearset with a higher 5th

nickdoof

iTrader: (2)

Mo' power now?

hristov

I have 3 wire oxygen sensors so where do I connect the resistor?

LvR

What resistor? ............. what are you trying to do?

Do you have a FSM for your engine (stickies have download links) - all the standard connection details are in there