Nikzilla

Ok, I have DataScan running, and here is a brief explanation of what happens.

Conditions:

• Very light throttle acceleration, or constant speed cruising.
• TP Voltage between 1 and 1.5V.
• No boost.
• RPM is at least 2500.

Issue:

• Narrowband spikes to the rich end (showing 100-120) and stays maxed out.
• Wideband spikes to the lean end (20+) and stays maxed out.
• Car stumbles.
• MAF voltage and Duty Cycle stay consistent when this happens, and the fuel trim doesn't budge at all (stays at 100 during the whole process).
• Quickly increasing throttle resolves issue for a second (fuel enrichment during fast TP acceleration).

Here is a screenshot of the data log with explanations on the graph.


Can anyone help me figure out what's causing this?

Also, I did replace the O2 sensor. I thought it would fix the issue, but it didn't.

If I unplug the O2 sensor, I don't get a lean spike on my wideband. However, it still runs too lean before boost (like at 16-17), but runs fine under WOT with boost (11.0 AFR). The difference is, it doesn't lean out continuously until it maxes out the gauge, like it does if the sensor is plugged in.

Kingtal0n

Normally, when the O2 pegs rich, and the engine runs lean, the narrowband is bad. Think about it, what do you trust more? your narrowband, or your wideband? If the narrowband says rich, and the wideband says lean, the engine must be lean right? so the narrowband is bad.


You could have it wired wrong, or there can be an exhaust leak before the wideband, but after the narrowband, and other possiblities. Furthermore, Is the engine in closed loop? When you say the fuel trim does not move, that tells me that the ECU is not adjusting the A/F ratio for closed loop, which puts you in open loop for some reason. Lots of little inconsistencies here.


Just make sure you do a boost leak test before any more diagnostics. Fill the plumbing, including the inlet before the turbo, with 10-15psi and check for leaks. Could be an air leak after the maf, before the turbo.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

The O2 sensor is literally 24 hours old. It's a 95 z32 TT MAF (I've got a blacktop).

I'm at attempt #4 at building a boost leak tester. My PVC didn't fit properly last time and flew off, but I could pressurize it enough so that once I took of a vacuum line above the throttle body, it took it like 3 seconds to depressurize.

About the exhaust leak, why would the narrowband be showing rich in that case?

I'm starting to think it could be wired wrong. The guy who had the car 2 owners before me did the swap. Maybe no one just realized cuz I just put the wideband and the consult in. Is there anything I should check for here?

Kingtal0n

check your signal wire at the ECU by following a pinout, and check(compare) for continuity at the signal wire of the sensor on the harness end. You can also check for voltage and ground, it should have neither. The narrowband sensor creates its own voltage signal (which is why you see 1-wire O2 sensors)

PVC = poly vinyl chloride
PCV = positive crankcase ventillation

An exhaust leak would show a lean condition. I only mentioned it because if the leak is near the wideband, the wideband will show lean. (meaning that the leak is far away from the narrowband)

furthermore, if you want to cheaply change the o2 sensor, buy the cheapest 1-wire you can find from a local auto parts store. You can literally ask them if you can look in each box and price check the 1-wire sensors until you find one that is about $8 or so. then wire that 1-wire to the signal wire going to your ECU. Once the car warms up, it will start working.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

Ok, but since I actually have DataScan, the fact that I can see the O2 oscillate in datascan, doesn't that mean that it's actually hooked up properly?

Maybe the wideband is leaking. I only torqued it by hand. Didn't have a crow foot on me at the time. I'll go re torque it properly.

Also, I got another coupling so I'll try to pressurize the intake again today.

Kingtal0n

You said that "Narrowband spikes to the rich end (showing 100-120) and stays maxed out. "

And now you are saying that its working fine.

Which is it?

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

It only does that under certain conditions described in the original post. Most of the time it oscillates like a happy sensor.

Kingtal0n

So you are saying the O2 sensor works fine. but sometimes, while cruising, it stops working, and goes full "rich" while the wideband shows a lean condition? And the engine struggles? Does the car decelerate? Does the throttle position increase to keep the rpm constant?

It sounds like the engine flips into open loop during your "condition", which would explain the sudden locking out of the fuel trim.
That could be the maf hot wire (clean it, what does the maf voltage do compared to the rest of the time? ) or an air leak between the maf and turbo (you are not getting enough air to pass the maf hotwire). It could also be a wiring issue (a sudden intermitten ground causing a sensor to fail for an instant) , a faulty tps (you would see the tps signal fluctuate, or spike).

But none of those things forces open loop, unless you meet the condition for open loop, or the O2 sensor fails, I am not sure why it would go there. If the maf voltage suddenly spikes up, or the TPS suddenly spikes up, yeah that would force open loop, your O2 would peg rich (narrowband)- but you would probably see the voltage spike on the data tool.

I still say, check for air leaks, clean the maf, change the maf hotwire (SOHC KA maf hotwire for redtop #62 ECU easy to find) first.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

Yes exactly! While cruising, under constant light throttle, the narrowband suddenly goes what the fuck rich and wideband goes what the fuck lean.

I cleaned the MAF last week. But there is NO fluctuations. Look at the screencap I posted. The MAF voltage, TP, RPM and Duty Cycle are all consistent. This is why I can't figure out wtf is going on.

Kingtal0n

this makes me think you have an intermitten wiring issue. The O2 sensor where the previous owner wired it (IT NEEDS to be lengthened) is probably rubbing a 12V signal due to vibrations. Find the spot in your harness where the O2 sensor was wired and check how it was done. Probably with tape and twisty 1 2 3 method.

The only thing with this... the duty cycle of the injector should be dropping when the conditions present. I would only be a little bit- maybe .3% or something. But it would be changing. See if you can increase the resolution of the duty cycle, or injector pulse width during the situation, look for a small adjustment down...
I cant read your screen shot, its very blurry. Have you tried to look at it? I cant tell one line from the other.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

That's solid advice.

Btw just wanted to show my appreciation to your posting. You're always answering everybody's technical issues with good answers.

Kingtal0n

thanks. Its really hard to diagnose anything over the internet- but it is alot of fun to try What else am I doing at 2AM with 6 hours before practical neuroscience class? sleeping? hahahaha no.

I'd be really surprised if it was wiring. Thats a heck of a guess but I suppose you had better check for peace of mind anyways. Backing up the diagnosis with confirmation from your data tool would increase the odds. But I cannot tell if the engine is flipping into open loop because it senses voltage on the O2 signal wire or what, having 12V at the ECU on that wire might actually damage something. good luck...

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

Got my unit to hold pressure! Woot! Took me only 3 high velocity PVC launches to succeed.



It goes Tomato Sauce > 2/12 in to 3 in intake coupler > 3 in to 3 in coupler > 3in PVC end cap > Tire Valves > Pressure Gauge / Bike Pump

I shall try it out in the morning.

Wish me luck!

Nikzilla

UPDATE:

So I tried to pressurize it, and I couldn't get more than 5 psi with a bike pump. Once I pull the pump off, the pressure veeery slowly goes back down to 0 on the gauge, but if I pull a vacuum hose there is still a lot of gas velocity.

So I hooked up my DIY smoke machine (electric air pump > paint can with rag in it > PVC). Lit the rag on fire, let it burn for ~30 sec, turned on the air pump, and put the lid on. The air pump couldn't get more than 3 psi, but I couldn't find where the smoke was coming out. All I could see was very very very light misting around the turbo, so light that I think I was imagining it. I'm thinking of taking it to a shop to have em run a real smoke machine through it, not sure if they will be SR friendly thou.

In terms of wiring, there is no extension! The O2 plug runs straight into the harness after 2 inches. There is however a clamp on the middle wire that goes to the narrowband under the dash that the previous owner installed. It doesn't work thou cuz I accidentally snapped the wire while doing my gauges. I unplugged the sensor clamp just now though just in case it's accidentally grounding it on the exposed end.

Kingtal0n

See if you can get a better picture uploaded that I can look at. For now, Ill just throw some ideas your way.

You say the injector duty/maf/tps do not move when the problem occurs. Only the oxygen sensors report a problem. You never answered whether you can feel the engine hesitate or if you can notice that the engine runs differently when it happens?

IF the ECU is not changing any values, but the fueling suddenly changes, something in the fuel system may be causing it. Possibilities are, fuel pump, fuel filter, fuel regulator to name a few. Since you are cruising, and the problem corrects with throttle, the fuel pump is unlikely. If vacuum at the regulator was the issue your wideband would show us a rich condition, not lean. I am still not sure if the engine is flipping out of closed loop, you said the fuel trim sits still. Does the fuel trim normally move while driving? I am unfamiliar with that data tool you are using.

If the fuel isnt changing, that is, if the injectors dont change, and the fuel pressure doesnt change, then the engine cannot suddenly go lean with no reason. Unless the ignition timing is suddenly going really wrong, or getting really weak, such as if you retard the timing 20-25*, THEN you would get a nice fat lean reading suddenly on your wideband, and the engine would lose significant power, and require more throttle position (ive asked already now twice if this happens you never answered). If the car continues driving normal-like, with the same throttle position etc... then we look elsewhere. Also check the ground for the coil packs at the back of the head. And look at your plugs, change them if they are very old or dirty.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

I stated in the original post that the car stumbles during this.

And yes, usually the fuel trim does move, but when this particular issue happens, it stands still at 100.

I also have a fuel pressure sensor, and I can say that the fuel pressure remains constant and does not drop not spike when this happens. Brand new Walbro 255 installed too, old one crapped out after the sock fell off. Z32 filter. Perhaps the only thing in the fuel system which I'm skeptical about are the injector O-rings. The injectors never have been touched as far as I'm aware of, but could bad O-rings cause this somehow?

I replaced the coil pack wires recently, and I remember screwing in the ground, but I'll check it to make sure it didn't slip off or something.

Now, timing, could it really affect the AFR? I didn't actually check that, I'll go pull out the log, brb.

Kingtal0n

right, I doubt its your fuel system. Just something to throw in the mix. If we rule that out, then we must consider electronics, wiring, timing, misfires, air/compression leaks.

The engine stumbles as in, misfires? If the engine mis-fires your wideband goes full lean. That sounds more like an ignition problem, coil packs, grounds, plug(s). Could even be just one single plug mis-firing.

Other than that- a mechanical problem, such as a sticking valve, can cause a stumble/misfire. Something else to keep in your mind. Any funny "metallic sounds" when turning the engine over?

If you havn't changed the plugs... do that.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

I'm no expert, but I'm pretty sure that classifies as a misfire. It jerks a bit.

I pulled the plugs a couple of weeks ago and regapped them to 0.32. They were gapped to much. Other than that, didn't look too bad. They were just showing a tad rich cuz I was running rich back then cuz my turbo elbow wasn't bolted in properly (goddamn previous owner, I HATE UUU), so it ran rich in open loop.

No mechanical weird noises, compression was at 150psi couple of weeks ago. There is a funny rattle though, which I'm 85% certain is the exhaust cuz I'm missing the insulator piece that connects the downpipe to the tranny crossmember (goddamn previous owner), and you can't hear it from the engine bay, only from the side of the car.

Kingtal0n

Just put some cheapy plugs in it and see if the problem goes away. If it does not, the next step is start replacing ignition parts, like coil packs. camshaft angle sensor if you have a spare. Igniter if you have a spare. Coil pack harness. Could always be something in the harness rubbing something else of course, but you wont find that unless you start taking off the tape and looking at whatever's down there.

Is the misfire constant? Is it like turning on a switch that stays on for a little while? Or does it happen real QUICK and then stop suddenly just as fast as it started?
And does it feel like a single cylinder? would be great if it was just one cylinder coil pack going bad.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

What's an "Igniter"?

Coil Pack Harness is brand new, just replaced it.

Misfire is like a switch. It's pretty much consistent in closed loop when there is no fuel enrichment. If I move the throttle quickly thou, the fuel enrichment fixes it.

I've actually tested the coil packs by taking them out and cranking the engine with the spark-plugs touching a ground, they all seemed to work fine. Of course, under load is a different thing.

NOW.

I was looking at the data log, and the timing stayed the same, but I noticed something else.

So I said previously that the duty cycle stayed the same. But, duty cycle is inversely proportional to RPM. So if RPM is rising, and injector pulse is consistent, the duty cycle should rise too.

So I mapped out the injector pulse, and what do ya kno, it goes down! However, it's not directly proportional with the narrowband readings. The Narrowband actually first reads slightly LEAN, all while the injector pulse is going down, but then the narrowband has a bipolar moment and switches to full RICH.

Here is the graph so you can see for yourself:

Kingtal0n

first thing i notice is your compressor surge. You need to fix your bypass valve.

Secondly, there is a whole section of tps missing. But Ill assume its flat.
Holding the TPS flat, but maf is climbing, rpm is climbing, and duty cycle is dropping. perplexing? maybe. The only way this is possible is if the narrowband was pulling fuel, or the ECU was mapped wrong. I would say try a different stock ECU. You HAVE a stock ECU right?

You said you tried un-plugging the narrowband. that would prevent the fuel trim, I cant tell from the graph what the fuel trim is doing. can you coorlate the fuel trim with the dropping injector duty cycle? If not, there is foul play afoot in the computer, because a rising maf signal, a rising rpm signal, should both yield - a higher duty cycle.

Ill keep thinking about it though, I might be missing something simple. its always the simple things.

Also, when the narrowband goes to the top, thats lean? or rich? because to me up means more, and more means more voltage, which means rich. but I think you said thats actually lean. clarify?

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Kingtal0n

Here are some of my logs with a power FC. This is accelerating through the gears. Notice when TPS held constant, RPM rising, duty cycle either moves up or stays about the same- never drops.

It isnt inversly related to anything though. It just goes by whatever the ECU tells it to do. Usually, more airflow = more horsepower = more duty cycle. if anything you could say duty cycle is directly comparable to horsepower output, but even that falls apart once you consider brake specific fuel consumption getting worse and worse as the engine richens up for power delivery.

Attached Images

	1done.jpg (38.0 KB, 6 views)
	2done.jpg (37.6 KB, 5 views)

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Kingtal0n

And also, since this is tuning into a tuning night,
What you are probably seeing is called transient delay. The injector has to inject the fuel, the engine has to burn it, and it has to get into the exhaust system, before the narrowband can "see" it. That all takes .250-.700 milliseconds depends on the RPM where you cruise and where the sensor is. So when you are looking at those graphs, the narrowband is always slightly behind whatever the injectors are doing.

I keep looking at your graph and I dont really see anything that weird, besides the slight trend down for the duty cycle. the dip lean in the narrowband corresponds to the dip in the TPS if you notice you lift for a moment then get back on it. Seems to react pretty fast- Ill assume the engine RPM was pretty high (2500~). when it goes "full rich" thats expected considering your throttle position.


please be more specific.
lets scenario. Your driving at constant throttle. The O2 sensor is flipping back and forth. You keep driving this way forever. Suddenly out of no where- the car acts up? For how long? If you just keep your foot there, does it start working again like normal?

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

Don't hate! I love the chu chu chu sound, no BOV. Don't really care, at only 7 psi it's not necessary.

ECU is 100% stock, never touched.

Here is the requested fuel trim. I'm no expert, but shouldn't that bitch be oscillating with the narrowband?


P.S. How could you tell I have compressor surge?

Someone is definitely fucking with me then

Nikzilla

That's pretty interesting, but what I really was talking about was the fact that the 2 graphs go in opposite directions.

Kingtal0n

the fuel trim looks right, ill assume 100 = untouched (100% fuel from the map). When you step down on the TPS it locks at 100 because the ECU stops looking at the narrowband during open loop. Assume when you see 100 fuel trim you are in open loop.

While the fuel trim is 97 Ill assume that 3% fuel has been removed. Basically not important. the ECU is not pulling that much and it has a max adjustment anyways to prevent a bad O2 sensor from causing serious problems.

focus on the scenario:
lets scenario. Your driving at constant throttle. The O2 sensor is flipping back and forth. You keep driving this way forever. Suddenly out of no where- the car acts up? For how long? If you just keep your foot there, does it start working again like normal?

compressor surge shows up on the maf signal- see the jagged edges?

You dont really have enough of a graph to say for sure what its "supposed to be doing" and even I have never seen a fuel trim graph from an SR20 so I am not expert either. It would make sense that it might adjust over time- but most ECU have a memory that stores up hours worth of data so a fuel trim might not move very much at ALL while you drive. In other words, it remember from the last few weeks, that in each cell, the fuel trim is a certain place to give the closest possible A/F to make the narrowband move through its center of 0.5volts, and it might only take 1-2% of fuel "trim" to cause the swing, which would be moving cell to cell since each cell has it's own fuel trim memory stored. But this is all just guess based on what I know about other factory computers.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

Ahh see that's where you're wrong! This is during (supposedly) CLOSED LOOP. Not open loop. (Yes 100 is 100% of the base map).

The TPS voltage you see here is only 1.1. This is basically 15% throttle, just enough to barely accelerate in 2nd gear.

If I open up map trace, it actually shows me in closed loop during the stumble.

focus on the scenario:
I apply 15% throttle at 2500-3000 RPM, attempting to maintain constant speed. Everything is normal for about 1 second, narrowband oscillated at an average of 0.5. Then narrowband all of a sudden goes 100+, wideband jumps to 17.2, 1 more second, wideband jumps to 20. Car starts stumbling. Car will KEEP stumbling (misfiring) if I do not move throttle. I don't let it do this for too long. I usually just floor it for 1 second from here on to throw it into open loop and give it more fuel. The only way to maintain highway speed safely is to drive 0% throttle / 100% throttle.

Kingtal0n

I still think you should put some cheap spark plugs in it and try that before doing any more diagnosis. dont even gap them, dont boost the car. Just drive it normal. Even with a .044 gap it will be fine.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Kingtal0n

It might be sucking air from somewhere. An enormous vacuum leak would cause that. Did you pressure the plastic inlet tube in front of the compressor? between the maf and compressor that plastic tube always leaks causing problems. An air leak accounts for all your symptoms, dropping duty cycle, lean misfire, and everything else appears to be working perfectly fine.

__________________
5.3/4l80e/turbo 240sx playthrough:
https://tinyurl.com/turbo240

Nikzilla

That's long term trim. Short term trim I believe is supposed to directly respond to the O2 input during closed loop. Speaking of which, I cleared my ECU, and my long term fuel trim refuses to ever budge from 128. Also, do you want more data? I can gather more logs. That one is pretty long though, that's like a good 5 seconds of engine exposure to leaniness.

Can you give me a recommendation? Not like I can just walk into autozone and say "Nissan 240sx, SR20DET".

I pressurized it from the MAF. In other words, I pulled the MAF/Pod filter combo off, and stuck my tomato sauce can in there. There is a boost leak somewhere I suspect, but it's so fucking tiny I can't see the smoke coming out.