Protege rough idle issues

I have a 2002 Mazda Protege 2.0 with a rough loping idle. The car has a JDM engine with ~50,000 miles (85k on the odo).

The car starts up and runs perfectly when cold, but after 2-3 minutes, when it goes into closed loop it starts to idle poorly and shake. The condition gets worse as it fully warms up, sometimes to the point of stalling, and has hesitation up to about ~1500 RPM. After that it runs perfect.

Hooking it up to my scanner, when warm;
it has a LTFT of around +8-10, sometimes higher (STFT stays close to zero). At 2500 RPM LTFT drops to around 0-+3
O2 sensor voltage cycling looks normal
MAF PID looks fine around 2.6-3.0 g/s idle, increases linearly
TPS output looks normal, no dead spots. Although if I unplug the sensor the car runs fine as it's in open loop?

The first thing I checked was for vacuum leaks, using my smoke machine I did find 2 leaks (intake gaskets and injector o-rings). Fixing the air leaks didn't make any change.

Other things I've done are clean the EGR valve, check fuel pressure, which holds good while running but drops off quickly.
I've checked for any loose or broken wires, exhaust leaks, and checked for power/ground at literally every connector. I even did a compression test and checked the engine timing. I can't seem to find anything wrong or broken with this car.

Other thoughts; sometimes it takes a few tries to get into live data on my scan tool, says it's not supported. Also occasionally disconnect saying it failed to communicate with the ecu.

Thanks for any possible help.

Just a thought but have you looked at your 5v reference while vehicle off and also under load?

TPS reference with the throttle depressed? If so, I just checked it like that and still a good 5v.

Ok I'm going to bounce ideas your way. Some to keep in the back of your mind and some to think about or check. Remember wires that are problematic may only cause issues under load. Is the STFT & LTFT data from before or after leak repair? 5v reference is on all of your sensors. (i.e. O2, air/fuel, cam, crank, etc.) Alot of times they are tied together running back to pcm. Possibly why unplugged TPS could act like unplugging an O2 taking you back into open loop. If one sensor has 5v then chances are, being tied together, all sensors have good 5v and thats not the problem. ScannerDanner channel on YouTube has some great video on 5v reference, it's importance and usefulness in diagnosing problems. Without knowing specs on your car and doing google searches there's some more info that could be helpful. Size of engine, # of cylinders, on and on. If you can find diagrams/schematics and post them here it would help. I don't know where you are with this vehicle, but Mitchell vehicle diagrams offers a subscription, $20 for one vehicle for one month. You can print out what you need before the month is over and keep it. Something to keep in the back of your mind as you go further down the road and some of the information in there can be very helpful and educating. If that fuel trim data was after you fix the leaks then I would plug in your scan tool and take a drive. It may or may not be the cause of your problem but it still needs to be fixed. I'm not sure what you know about STFT & LTFT but I can offer a great YouTube video that would bring you up to speed. Good luck and I'll be looking for a reply to help out.

Are there any chk engine codes? Forget to ask.

First off, thanks for the reply.

The car has a 2.0L inline-4 DOHC, it has no codes and has not since this issue arose.

I guess I should've mentioned that the scanner data is post leak repair. I did not have my scanner before I fixed the leaks, so I can't say what the data was like before. The car has been sitting awhile since then.

Fuel trim as I understand is the amount of fuel the computer adds or removes from normal, STFT being instantaneous and LTFT being the new base-line.
So this car would be adding 10% more fuel by default to compensate for a lean condition. Is my understanding correct?

I would take the car on a drive, but as it's been sitting the tires are junk. May be able to scoot it around the block anyway

I've attached some relevant diagrams, maybe they'll be helpful?

Yes you have it correct. I'm running out the door to a performance and will be back tonight and see any replies but till then check out this video. There is a part two as well.

I don’t want to send you down the wrong direction, but have you performed a leak down test? If you give it just a little throttle does the problem go away? If so I would consider running a leak down test as it would indicate a valve issue. If live data seems okay I would look at TSB’s for programming updates. Something else to consider is a sensor failing when it gets hot.

I don’t want to send you down the wrong direction, but have you performed a leak down test? If you give it just a little throttle does the problem go away? If so I would consider running a leak down test as it would indicate a valve issue. If live data seems okay I would look at TSB’s for programming updates. Something else to consider is a sensor failing when it gets hot.

Just realized you said the car has been sitting... fuel quality and injectors should be considered. Could be gummed up and not allowing enough fuel to come out and the computer compensation may be confused on its fuel control strategy.

John Curtis makes a great point, fuel quality is something to consider as well as it's effect on injectors. Was it parked for a bit prior to the onset of your problems or was it parked because of these issues? If so then the gas might be causing issues with combustion and or clogging up injectors. Either way its something to consider and at this point you might want to make sure there's only fresh gas in it from here on out. It could hide the fact that you've fixed it by masking it with a new problem. I'll admit that I know just enough to get myself into trouble sometimes and as a wise man said once, "no stupid question only stupid answers". I hope to only help with the info I offer up and hope anyone will correct me if I'm incorrect. I'm here to learn just as much as to help out. That being said... One of the pieces of information that caught my eye was about the fuel pressure. From what I can find on the internet it should be around 39-45psi. What was your psi? What ever the problem, it seems to be a "hot" issue or worsening at temperature. When and if it would stall, how was restart? Any difficulties or did it start right back up? For me, and I'm thinking out loud, I'm wondering if it's a fuel issue. Before we head down that road, something that would help to put your scanner data into context. Have you looked at/cleaned your MAF sensor? With a scanner hooked up looking at MAF Hz, RPM, and HO2S(mv) at the same time/real time. If you do a wide open throttle rapid snap and release once or twice it would tell you some good info. Maybe include TPS%, that would verify wide open. Looking at MAF and HO2S at the exact point of snap you could possibly verify or rule out a fuel issue with how the HO2S reacts. MAF info could tell you the state of that sensor, if it's working properly or needs to be cleaned if it hasn't been already. I know you mentioned that it ran well after 1500 RPM and even though that fact isn't lost on me, I think I would want to start ruling things out. This would be a simple and easy way to do it. Just a thought. Keep us posted, good luck.

Thanks for the suggestions. I should mention that I drained the tank of all old gas before starting working on it.

@John Curtis, I actually performed a leak down test awhile back, must've slipped my mind, and it was all good. Question with the injectors. If they were clogged enough to cause this issue, wouldn't they also cause it to run poorly at higher RPM as well, as the engine is using more fuel? Also no related TSBs that I can find.

@Otbrecords, I rechecked fuel pressure at operating temp, it sits at around 30-32 PSI (spec at idle is 30-36). I also checked with the pressure regulator vac line unplugged, it stays around 34-36, which is below the spec (39-45 PSI). It also drops to around 28 PSI when I drop the throttle. Could this be part of the issue?
I also did what you suggested to check the MAF, TPS, and O2, and evrything seems to be normal there. I also tested the ECT sensor at operating temp, no issues. I think at this point I can safely rule out these sensors.

Strange thing though, with my scanner, when I try to select individual items from data stream I get a ecu communication error every time, but fine when viewing all data. I tested the scanner on a different vehicle, so I don't think it's an issue.

Thanks again for all the help so far.

I searched around the Web yesterday to see if the OBD2 issue was in any way a Protege or Mazda specific issue. Didn't say anything out of the ordinary. With the rpm snap the O2 didn't go lean (low mv) at the exact moment of snap? Also the MAF was somewhere above 7000Hz? Did you do any type of work to the vehicle prior to the issues? As I'm sure you know disconnecting the battery in sometimes cause issues. Is it Fly by wire throttle?

At cold start up the engine calls for extra air until it warms up, then RPM drops and the computer reduces idle air bypass. A couple things can cause issues at that point. First one was what I mentioned disconnecting the battery sometimes causing the computer to forget idleair adaptation. Not a problem with colder dense air but a warm idle will show issues until it's up to RPM sucking more air. How does your throttle body look, clean?

Okay, so I just took it for a drive and discovered it's quite a bit different than sitting still.

First off it hesitates/bucking/jerking up until around 3,000 RPM. Scanner data shows a LTFT jumping rapidly between 0 and +19% and STFT anywhere between -23 and +23. MAF reading at 2500 RPM while moving between 9 and 23 g/s depending on throttle.
Also close to the end of the drive the check engine light started flashing, then steady on. Pulled codes and got Cylinder 3 misfire and a pending p0660 for the variable intake control circuit open. Fuel pressure and O2 readings are sthe same under load as not.

@Otbrecords, The car has a cable actuated throttle. The TB and intake were cleaned when I replaced the gaskets. Before driving as per your suggestion I took a closer look at the O2 output. At the moment I hit the throttle the voltage went up to ~850mV. Also at idle the voltage tends to stay lower longer(around ~120-340mV) before cycling up again.

About the MAF Hz output, I don't have a frequency PID on my scan tool.
Lastly, no other work was done prior to these issues.

Interesting... Glad you were able to get it out and drive it. Changes the whole out look. I'm going to go work on my own elusive mystery vehicle issue. See 03 Tacoma #1 & #4 no spark in the forum for a good read. Might help you to forget about your problem for just a moment, probably not but... While I'm doing my best to play pro mechanic on my vehicle, I'll chew and digest the information you posted about yours and get back to you. It's raining here where I am, hopefully you have better weather. Praying to the car gods.:side:

I just took another look at your long-term and short-term. Whoa! There's some craziness going on there.

You know just for just shitz and giggles, check to see if your O2 sensors are plugged in correctly. I don't know why they would be unless you recently replaced them. I had a buddy who had his plugged in backwards causing that exact same scenario, short-term long-term jumping like that. Just a thought while I was walking out the door. Good luck

appears you're having a fuel delivery problem... I'm seeing a possible weak fuel pump. psi at idle and WOT are low... should have approx. 10psi difference between idle and WOT. Snap test reading should be 35psi at idle, 45psi at WOT (same as when you remove vacuum from pressure regulator). Pressure will drop about a psi or two when the throttle quickly drops to idle then back to the 35psi.

Before you condemn the fuel pump check for an inline filter it may be restricted or a restriction in the line as your fuel system was sitting for some time.

You did not mention if your using your original injectors or one's that came with the replacement engine.

ScannerJohn is having the same thought as I did and kinda still do. Also check to make sure powers and grounds are good on the pump.

Okay, so I am using the original injectors, but I also tested with a spare set I had laying around.

I'll inspect the pump and try to flush the lines out tomorrow. Thankfully all of the fuel system is very easy to access on this car. Funny enough, I had to replace the pump twice before, filter was changed both times also. First time I replaced it was when I swapped the engine. then the car sat for around a year an the pump was dead again. Maybe a wiring issue, or could a clogged line cause this also? Both times the pump failed it wasn't putting out anything.