2.3l engine hesitation in "carburetor icing" conditions

[gnwiii]

2008 2.3l manual. Last 2 winters in damp Nova Scotia weather around freezing (classic carb.ice conditions for those of you who remember carbs or who pilot light planes) I get hesitation (sometimes just a jerk, often as if engine was turned off for a few sec.) under load. Sometimes the check engine comes on. Dealer has replaced purge valve, both ignition coils, crank sensor, a valve in the vapor recovery system, and sections of wiring harness. This week they replaced the fuel pump relay based on some other reports of problems in damp weather.

The problem usually occurs after a sharp transitions from cold (<-5C to just above zero with lots of
visible condensation). I had the problem last weekend and and again today (Sunday) which started at -12 and was up to zero at 14:00 when I went for a drive.

If I put in the clutch when this happens RPM falls to zero but engine refires when I let out the clutch.

<http://en.wikipedia.org/wiki/Carburetor_heat> explains the issue. Early Mazda RX7 used
cabs that were prone to icing.

Icing in the air induction system has been found in fuel-injected aircraft engines, but is not thought
to have significant impact on engine performance.

We also have a 2007 M5 2.3l manual that has never skipped a beat so no doubt something is
amiss with the M3.

Ideas?

[ricohman]

This isn't icing and its not cold enough fuel water/fuel problems.
What code is the dealer getting when the CEL lights up?
I would think its in the induction though. Have they measured and cleaned the MAF? And the TPS? I would start with those.
A while back I was chasing a similar problem on a buddies Civic and it turned out the be an intake leak at the head.

[slam525i]

You're a pilot too? What do you fly?

I really doubt you get significant icing in the induction. There isn't the greatly accelerated & constricted airflow in fuel injection as in a carb. Sounds to me like you have an electronics issue where something is getting wet with condensation. Is there apparent power loss if you open the throttle?

[TheMAN]

has the spark plugs been changed? what about the throttle body? has that been cleaned at least? there have been sticky throttle body issues that have caused stalling issues... an ECU flash update has also been done as a countermeasure also, so make sure that is done too... the throttle body should also be warmed by coolant hoses running through it... if you have modified your car and removed that, then it would be a major contributing factor to your problem

[stevenma188]

I would say check the MAF, and try a ECU reset (remove battery and wait etc). However those are pretty basic things to check and I would hope the dealer would have done those before replacing all those parts.

I'm also a pilot too. I think I'm going to start a Pilot's thread.

EDIT: Here is the thread - http://www.torontomazda3.ca/forum/sh...Pilot-s-Thread

[Aitch]

I've found the same thing with my 2.0L in the past few months. Got my first check engine light though just this weekend. Good idea on checking the spark plugs, as mine are probably due for a change as well. I figured it was most likely a MAF or injector-related problem.

[gnwiii]

I'm not a pilot, but in my youth (long before emergency locators) was active in Civil Air Patrol in the US. We used Stinson L5 Korean war surplus planes -- big engine in a slow plane and very tricky to land, but could climb out of downdraft situations that most light aircraft would not survive. I remember one windy day when it took 10 mins to fly to an airport with fuel on the other side of the valley, 90 mins to get back against the headwind.

Engine is dead for a short period -- opening throttle just gives a bigger jolt when it comes back. Certainly a bit of condensation in the wrong place could cause confusion in the ECU. Another post says there is heat going to the throttle body -- if that wasn't working for some reason it is conceivable that "minor" ice buildup is mishandled by the ECU. Could the loss of power come from ECU doing some sort of reset?

[gnwiii]

This isn't icing and its not cold enough fuel water/fuel problems.
What code is the dealer getting when the CEL lights up?
I would think its in the induction though. Have they measured and cleaned the MAF? And the TPS? I would start with those.
A while back I was chasing a similar problem on a buddies Civic and it turned out the be an intake leak at the head.

The codes:

PO301, PO303, PO304 -- replaced #1 and #3 coils
PO303 -- replaced #3 and #4 coil and 2 plugs
PO303 -- replace purge valve
PO302 -- replaced #2 coil and a plug
PO455 -- replaced purge valve
PO455 -- replaced purge valve

I'll mention the other items to the dealer.

[ricohman]

They are throwing parts at the problem.
For example, when my V10 Triton spit out a code for COP's #7 , 8 and 9 the fix was a weakened vac line at the EGR switch plate.
And my brother-in-law just fixed his 5.4 that was backfiring through the intake and throwing codes to all the plugs on the left bank.
The problem was a cracked rocker.
The tech needs to start thinking about why you have a lean mix and less about replacing parts. You have a problem with a lean condition or a complete ignition failure. And the plugs and COPS are the last things downstream...

[gnwiii]

Just to update -- the dealer has continued to just replace coil/plug assemblies. In reposne to the suggestions made here they say that most of codes are for a consitent missfire on one cylinder. Anything to do with air fuel would give a random misfire. However, once the check engine light comes on the engine runs fine, so I doubt there was anything wring with the plugs. They told me the problem was due to either a mouse nesting under the engine cover or Krown Rust treatment. Today was the first carb ice weather this season and like clockwork the check engine light is on again -- I was driving on an ice-covered 100-series highway in heavy traffic when the engine was cutting out -- had to use the clutch to keep the thing facing the right direction. Good studded snow tires probably prevented a serious accident. I'm waiting and waiting and waiting some more for the service manager to return my call.

[Booter22]

only 3 years later.. good thing i wasnt waiting..

[gnwiii]

The rest of the latest episode:

Drove the car Monday and Tues. with the "check engine" light on solid but engine running well. One the CEL is on my experience is that the engine runs well (might be down on power, but this usually happens on wet icey roads so power is the last thing I need). Didn't touch the car on Wed. except to clear off the snow, but make an appt with dealer for Thurs. When I started the car teh CEL was off and stayed off with engine working well. The dealer found "lots of codes". The EGR checked OK. They cleared the codes, drove it, and got more codes, tried swapping plugs but codes weren't following plugs. Then they looked at the MAF and found a defective gasket. They think it was still sealing most of the time, but under some conditions might let moisture in.

I've learned to recognize the syndrom: carburetor icing weather, usually coming off cold (-10C) with lots of condensation on every metal surface and wet ice on the road. The car lives outside. About the time the temperture reaches "normal" the engine goes dead for a couple seconds (if I push the clutch RPM's drop to zero and I can get it pointed in the right direction) then refires. Sometimes this happens a few times without CEL activity. It can be pretty hairy if I'm trying to merge with traffic on a 100-series highway when the engine goes dead (ramps tend to have slush on the ice, so a bit of power helps keep things pointed forward). A couple times it needed the starter to refire (not enough traction to do safely restart with the clutch). The dealer says the computer detects a problem and shuts down the engine. They seem to have a lot of confidence in the computer codes even when they don't make sense. They do say that when they get a consistent misfire it doens't normally go away and there is often visible damage. The cars has spent several weeks total at various times, but I don't think it has ever misbehaved on a test drive. Of course, working on the car inside would warm things up, so it the problem is due to condensation or ice forming in the air intake path you wouldn't expect it after work has been done.

I also have a Mazda 5 one year older but also with the 2.3 engine. At its best the 3 has never been as drivable as the 5. Idling at a light the 5 shows a steady RPM while the 3 tends to hunt or surge. When I accelerate the 3 tends to be flat and then surge slightly while the 5 has smoother response. The 5 has hesitated a couple times in the same weather conditions, but not to the point of shutting off the engine.
The was a Mazda service bulletin for hestation in cold wet weather with a computer update that was applied to both cars. Both cars have had Krown anti-rust treatment annually at the same location. The dealer keeps mentioning the Krown might be connected to the problem. That might make sense if some defective seal allowed water into an electrical component and then Krown followed the water. What happens when water+Krown freezes?

The dealer seems to have a lot of faith in the "codes", but in tha majority of these episode the codes have indicated a consistent misfire in one plug while the engine runs well with the CEL on. The service manager suggested that the computer may make some change to ignition timing and mixture to compensate, but I think it would be pretty easy to recognize a consistent misfire on one out of 4 cylinders and the loss of power would also be very noticeable.

Can anyone explain why the CEL would be out after the car was sitting for a day? In the past I've driven several days with the CEL on before I could get to the dealer. What causes an engine to shut down but then refire? What changes to the engine are made when the CEL is on?

[Booter22]

has there been any attention to the coils? any hair line cracks could cause the random misfire and wore when there is snow or water / condensation and be better when it is warmer. if the coils are the same ( check with the dealer ) and maybe switch them from the 5 to the 3. see what happens. also we have seen some issues with throttle bodys that need the carbon to be cleaned and then they need a re learn procedure and software update after.

[gnwiii]

I'm not touching the 5. For the 3 the dealer has been replacing the plugs and coils as a unit. All 4 were replaced last spring (6 months ago) at 5 years. Last Thurs the dealer tried moving them around to see if the codes followed a plug. I have one here that was replaced last spring -- it looks new except for a bit of carbon on the plug. On the side of the coil there is:

FoMoCo
6M8G-12A366
099700-1064
N7431
1520

It lloks like <New-Ignition-Coil-Pack-Mazda-6-CX-7-3-2010-2009-2008-2007-2006-MX-5-Miata->

None of them have shown visible cracks or damage, and by now they have all been replaced at least twice. Why would the engine run OK when the CEL comes on if a plug or coil was bad? Why would carbon in the throttle body cause weather-dependent problems, and wouldn't that be more likely to cause a random misfire indicating a fuel/air problem? Most of these episodes have come up with a misfire for one cylinder, only a couple were "random", which is why they have been going after the coil assemblies. I'm inclined to agree that moisture is getting somewhere it shouldn't, but in a connector or sensor, not the coil assembly.

[TheMAN]

one thing that seem to have been overlooked is the ECU... any wiring, connector, or seal pack damage on the ECU could cause issue under humid conditions... the ECU sits in separate compartment next to the battery but it also gets fed fresh air from the battery cooling duct system... I have seen ECU connector issues on RX-8s before, which also sits next to the battery and is also air cooled... might be worth a look

at this point, I seriously doubt it's any sort of sensor issue but either leaky gaskets/o'rings or wiring
codes don't mean a whole lot when they start showing up light christmas lights... I've seen cars behave weirdly and have a code that had nothing to do with what the real cause was

[gnwiii]

I agree ECU should be investigated, but this problem has been remarkably consistent for 5 years. Most connector or water leak issues would have gotten much worse by now, so I'm thinking more along the lines of condensation or ice forming inside sensor, connector, or ECU (e.g,, distilled water rather than the salt-enhanced stuff that you get by simply breathing in NS).

In any case, I've decided enough is enough. This car has tried to kill me one time too many. My wife refuses to use it. Even if I found something that might be the cause, I wouldn't want to put it to the test, and I want to know that CEL coming on means a real problem.

I'm tempted to go back to VW diesels. I drove VW's for years back when they didn't use computers -- once one fired it would only stop when it couldn't get fuel -- but I fear those days are gone.

I do appreciate the Mazda community -- clearly a goto brand for people who appreciate cars.

Anything to watch for on 2012-2013 SKYACTIV 2.0 litre 6-speed? Reviews have been positive, but that doesn't mean they won't all blow up after 6 or 7 years. I did find one TSB for headlamps flicker.

[Booter22]

Anything to watch for on 2012-2013 SKYACTIV 2.0 litre 6-speed? Reviews have been positive, but that doesn't mean they won't all blow up after 6 or 7 years. I did find one TSB for headlamps flicker.

well.. i guess we need to wait until 2018 - 2019 to find out if they blow up or not havent seen any issues with them. just follow the maintenance plan

[TheMAN]

did you think about this too?
http://www.protegefaq.net/tsb/01-007-06.html
yeah, it's "no start", but it's possible that it could cause intermittent problems too