Stumped by a Mazda 3.

So, my freebie Mazda 3 is not cooperating at all. 

I installed the new alternator, tensioner, and belts. Charged the battery and fired it up. It runs like utter crap.  It was throwing a code for the intake air temp sensor which according to my research is in the mass air sensor. So, I replaced it and cleared the codes. No change other than it's no longer throwing a CEL. About that time it dropped a cylinder. In checking the coils, I found the coil wiring to be a crumbly mess and all of the coil connectors to be broken, so I replaced all the coil connectors, coils, and spark plugs. That helped some but still wouldn't idle. It was surging and stalling badly. I figure WTH, I'll go ahead and throw a O2 sensor at it as well. That helped a little. Surging is pretty much gone but still not idling very well. Anything close to idle speed and it shakes like a wet terrier and stalls. 

Pulling live data the only thing I really see is fuel system one is showing an OL-Fault and the ECU is retarding timing a lot, like -11 to -14. The timing comes back up as the engine RPMs increase. The load percent is also high, like 60%. That's me holding a little throttle to it to keep it running at just under 1000 RPMs.

I'm open to any suggestions. I'm not familiar with everything the scanner is telling me so I'm going to post the data and hope one of you is more knowledgeable than I am. Any help getting pointed in the right direction would be greatly appreciated. If I need a better scanner, let me know. I'm not above buying new tools. 

Thanks.

In all this poor running has the timing belt jumped?

In reply to John Welsh :

IIRC this engine uses a timing chain. Not to say it hasn't jumped but the car was parked due to a failed alternator. I don't think it would have jumped but haven't checked. 

Correct it is a timing chain, and one that is not typically known for problems.   
 

Have you tried a compression test just as a base line?   

Toyman! said:

In reply to John Welsh :

IIRC this engine uses a timing chain. Not to say it hasn't jumped but the car was parked due to a failed alternator. I don't think it would have jumped but haven't checked. 

 

 

I would start with cleaning the throttle body.  That is the point of highest degradation (meaning the computer must adapt) and after disconnecting the battery, sometimes it is too far off for the PCM to adjust.

Carb clean on a rag, a little scrub scrub of the bore and throttle blade, put it back together.  Resist the urge to actually spray the throttle body as the stuff can damage the gears in the housing.

Next if that does nothing I would want to check exhaust backpressure before the converter.  just to rule that out.  If you do not have a pressure tap, a Miata diff drain plug (18x1.5) drilled to 1/8 pipe, a pipe to 3/16 brake line adaptor, and a foot or so of brake line will allow you to make an adaptor for the front O2 sensor bung to connect a rubber hose to a vacuum/pressure gauge.  Should be zero, should never go over 2-4psi on snap throttle.

If that is okay, next is checking timing, which is tricky.  Easiest way is to make a stop bolt of the correct dimensions (there are several), stop the crank in its check position, and pop the valve cover off to see if the slots are lined up with the valve cover rail.  The crank stopper just bonks the side of a crank counterweight at #1 TDC.  There are two different things at play: the crank to camshaft correlation and the crank to harmonic damper (contains crank sensor ring) position.  

In reply to Sonic :

I have not. Good thought. I'll throw the compression tester on it and see what comes up. 

What year is it?  If 1st Gen, they have an issue with the MAF loom - the last 1-2" near the MAF can suffer breaks inside the insulation of the individual cables damn near impossible to find, can cause weird running issues whilst often never throwing any codes.  Remove the loom insulating tape, give it all a wiggle, and see if anything changes.. (hi-tech version)

In reply to paul_s0 :

It's a 2004 so very 1st gen. Thanks!

Paul is right, I've had that, made for interesting running on track as the wires connections changed.  
 

Edit: my throttle body won't work, it is from an 07.  Lots of differences between the 04/05 and 06+.    

We had a similar issue with a Mazda 6.  It was the o2 sensor downstream of the cat.  Unplugging it solved everything.  Eventually put a new one in and it worked fine.

Easy to test!  Just unplug the sensor.

My 2007 Mazda3 also had issues with the MAF wire loom.  There were no visible breaks in the wires but, squeezing the loom caused the engine to stall.  I ultimately found two of the wires broken inside the insulation which caused intermittent contact.  The wiring is really thin and the harness could stand to be longer here.

The MAF sensor is reading 1.315 lb/min which converts to roughly 9.9 g/s. I will assume you have a 2.0 for the sake of argument. At idle you should see no more than 4 g/s. The MAF is either reading incorrectly or the harness, pcm, etc is damaged somewhere. If you unplug the MAF the car will go into limp mode but should idle a little better as a quick test.

Load is likely high due to the MAF over reporting airflow. 

Throttle body clean and relearn is also a good idea. 

If the MAF was reading high, THAT high, combined fuel trims should be in the -40% or more range.  (Less, if you are a mathological pedant)  At least, I would expect that were the engine running in closed loop. Which it isn't.  Twice the airflow means twice the fuel, it should be fouling the plugs with richness if it wasn't moving that much air.

I would expect 4 grams/second were it running smoothly at a low idle in gear.

If ignition timing is that far retarded, the engine will be struggling to make enough power to stay running and airflow will be high as a result.

If the MAP really is at 28" and that is not a false value, the engine is essentially running at WOT just to idle.

This behavior happens when the throttle is stuck.  Being drive by wire, if the throttle cannot be closed enough to idle, it will retard timing and possibly cut injectors to get the speed down.  Being drive by wire, it will also treat the accelerator as a torque request and bring the ignition timing back up as you apply pedal...

Does the throttle plate move when you ask it to?

Yes, fuel trims should be pegged if it was in closed loop. With the TP showing 23% it's possible to get around that 10 g/s, but it does still seem a little high for 900 rpm and a steady throttle position. I would imagine the throttle body is dirty, so who knows what 23% TP actually means. That could just be 13% over the learned 10% it needs to idle. 

Another thing to note is the O2 readings, they're both reading dead lean at .02v and .015v. 

Have you checked fuel pressure? Maybe you have an injector or two that isn't flowing? 

You could pull the plugs and see if anyone of them look like they haven't actually seen a combustion event. 

I did spray out the throttle body but did not pull it to clean it. The throttle plate moves by hand but I didn't check to see if the throttle pedal is moving the plate and closing all the way. I will check that this afternoon. 

I did have an issue with the throttle pedal sticking. The interior was pretty funky from being damp for a couple of years. I sprayed it out and it is working smoothly now. I may need to check and see if the pot is cankered up. I assume TP_R and TP_B are throttle pedal readings. If the live data will read with the engine off I should be able to check the pedal movement and look for dead spots. 

Disconnecting the MAF causes the engine to die almost instantly. I will check that wiring this afternoon as well. 

For the record, the engine is the 2.3. I believe it uses pretty much the same electronics as the 2.0. 

I'd second making sure the throttle body is capable of moving the plate (specifically closed) where it needs to be.  Volvos will do all sorts of odd stuff to control the idle speed when the plate is frozen in it's neutral position.  Starts by cutting ignition to various cylinders until idle speed is roughly in range.  I presume everyone does something along those lines, and retarding timing is certainly on the list of possibilities.

Naturally it was drizzling when I got home so nothing got looked at. It will be next week before I can get back to it. 

I will stress to never spray cleaner chemicals into a drive by wire throttle body. 99% of them have plastic gears in the motor section and the cleaner can get past the shaft and attack the plastic, and if the cleaner does not dissolve carbon based things then it will do you no good to clean with in the first place...

This has been drilled into me by multiple trainer technicians so I figure their words have merit 

Always on a rag, and with the ignition key in your pocket or the unit disabled electrically.  The motor is powerful and on a very high gear reduction, I saw a demonstration with carrots of how strongly the throttle will snap shut if commanded to.  

And from personal experience, never open the throttle to its stop.  Have damaged a throttle body that way, changed the clocking of its TPS just enough to cause all sorts of weird issues.  And this is how we learn...

In reply to Pete. (l33t FS) :

Well, today is a good day to learn something new. Thanks.