Dangerous cylinder deactivation trouble on 2018 CX-5 and Mazda6 models

[davmac]

Here I thought I was a pessimist - a curmudgeon. But 32 pages into this thread and with nearly 10K posts between you: you guys put me to shame.

I'm also afraid Mazda is dealing with this problem from a cost benefit standpoint, to everyone's detriment. *
Not good by all appearances.

And whoever in Mazda Corporate he/she ignored the bad reputation from history and came up this idea of the FIRST 4-cylinder engine with cylinder deactivation for North American market should get fired! *

God bless you guys. Thank you. I feel like an optimist. I think Mazda is taking appropriate action with the recall. If / when I get a recall for my 2018 I'll take it to them and accept the fix. If I have any concern I'll pop the valve cover off myself and take a look.

 

[yrwei52]

One could see the rocker not in contact with the camshaft with the engine NOT running. Ed

But not through the spark plug holes with a scope ...

 

[jjm86m]

Hi JJm86m (and Go Hawks too).....I thought I understood what need to be done to check for dislodged rocker arm, but after reading your post here I got really confused.

Please help me(and other viewers) clarify or confirm a few points:

1. Are you saying that after the reflash of your PCM, you tried to rev your car to high RPM in effort to get the car to go into limp mode and you could NOT?

2. Your post sounds like the reflash of the PCM prevents the car from going into limp mode at high RPM and AT THE SAME TIME you still have the dislodged rocker arm. True?

3. If #2 is true, then why do we care if there is a dislodged rocker arm..after all, the car no longer goes into limp mode anyway (unless there is some power loss at high rpm due to less # of cylinders in operation).

4. What I understood was.if you want to check to see if you have a dislodged rocker arm, you can rev the car to high enough RPM (like 5500-6200 and hold it there a bit) to see if the car would go into limp mode and if it does, you have a dislodged rocker arm I this still true?

Tran,

Below are my answers based on my experience.

1. That is correct. I could not get the engine to misfire after the dealership reflashed my PCM. The tech was able to replicate the problem prior to reflashing and neither him or I were able to after the reflash.

2. That is correct. Only caveat is that I wouldn't go as far as saying the PCM is disabling the limp mode, however it didn't happen after the update. There could possibly be other factors besides simply disabling limp mode.

3. You don't want loose parts floating around in your engine. Per the recall "In the worst case, engine damage may occur, resulting in engine stall while driving without ability to be restarted, which may increase the risk of a crash." I'm assuming engine damage isn't occurring because of the misfire but because the rocker arm is jamming a cam or worse.

4. Kinda. I never held my RPMs, I redlined it and expected the transmission to auto shift to the next gear @6200. In my case the engine immediately went to limp mode instead. 1st to 2nd gear and 2nd to 3rd gear was enough to get the limp mode to occur. The safer but more costly but more sure way would be to have your dealer remove the valve cover to check. The RPM check has only worked for me Pre Flash.

 

[jjm86m]

Autoblog is covering the story and talking about the rocker arm

https://www.autoblog.com/2019/07/12/mazda-3-6-cx-5-recall-engine-stall/

 

[Mulligan]

Your way won't work because it's the "hydraulic" lash adjusters which is pushing the valves down. If the engine is not running, the HLA can't push the valve down as there's no oil pressure to supply the force within the HLA.

I'm not sure any of us completely understands how the HLAs work when switching between normal and cylinder deactivation modes. So I'm open to the possibility that the following reasoning is flawed, but here's how I see it.

What little CD documentation exists online (see this page) describes a lockpin in the HLA that is released to allow the HLA to collapse, which results in no valve lift.

I suspect the default condition for the HLA is to have the lockpin engaged, and that oil pressure is utilized to release it. My thinking here is guided by the expectation that when starting the engine, when there's little if any oil pressure, you'd want all cylinders to be operating. If that's right, the valves in cylinders 1 and 4 would operate when manually turning the crankshaft.

On the other hand, if the default condition is for the lockpin to be released, it's true that the valves in cylinders 1 and 4 won't be operating when the engine isn't running (i.e., when the crankshaft is turned manually).

Whether the lockpin is engaged or released by default, it does seem clear that oil pressure is used to control the lockpin. Oil pressure doesn't directly "supply the force within the HLA" that the rocker presses against when operating a valve.

 

[shadonoz]

Thanks for that reference, Mulligan, it really helped my understanding of how it works. Here's my reading of it.

The rocker arm is what opens the valve, not the HLA. When locked under normal operation, the HLA keeps the rocker arm in contact with the cam such that the valve must open when a lobe comes around.

Under CD, a locking pin in the HLA is released, allowing the HLA to basically collapse [or an internal spring to release?]. That way, the valve springs resist opening, and the HLA moves instead [kind of mimicking the movement of the valve].

If the HLA collapses too much, the rocker arm can fall over since it only has 3 points of contact: the HLA, the cam, and the valve. If it loses one point of contact, it's free to move out of position. So an unexpected failure of the HLA regardless of cause could lead to rocker arm displacement. Seems like a really vulnerable design to me.

Not clear on how the oil pressure relates, because yes, it seems the default must be locked with no oil pressure. That would suggest that pressure must be applied to unlock the HLA. Too much or too little pressure might make it "collapse" too much and allow the rocker to fall. Or maybe low pressure keeps the HLA from relocking after it has collapsed, revs rise, and the rocker arm goes flying.

Just thinking out loud.

 

[Tran7270]

Tran, I can't say whether or not the reflash disables limp mode. My car has a new engine and reflash and so far I have had no further issues. My biggest concern about the recall only doing the reflash is the fact that the owner really has no indication of the rocker arm being dislodged. If it is already dislodged and they reflash the PCM, they are preventing nothing, the damage has already been done and still exists unbeknownst to the owner and the dealership. Yet both parties feel as if the problem is remedied. If your rocker arm is off and the reflash does disable limp, obviously it will not go into limp mode as it would have before, but that doesn't mean there is no problem. This is why people are saying the reflash could possibly mask or hide the issue. So now you may not have to experience the car going into limp mode like myself and others have, but that rocker arm could possibly work its way into a place that could do catastrophic damage to anoher valve or worse. And there is no guarantee when or if this will happen, but Mazda definitely knows this is a possibility. This is just one of the reasons why I still believe they should be doing a physical observation along with the PCM reflash.

I ask again. If you do the reflash and your rocker arm is already off and thousands of miles later your engine craters because of this preexisting condition that wasn't discovered due to the lack of a visual check, who is financially responsible for this repair if you are no longer under warranty? And how will you be able to prove that this was preexisting? Nobody knows how many miles one can drive with this rocker arm dislodged. As near as I can tell mine may have been off from the time we drove it off of the lot. And we had 9000 miles on ours before we had our experience. We never had any indication of a problem. Mazda KNOWS there is possibly vehicles driving around right now with a missing rocker arm. It appears to me that they are trying to take the easy and cheap way out. And quite frankly it disgusts me.

Thank you Go Hawks for sharing your experience & concern.
I already made an appointment with the dealership to have my PCM reprogrammed per the recall and will also ask them to open up the valve cover to do a visual inspection of the rocker arm, but chances are they won't because it is not part of the recall. Fine, I will pay for the visual inspection work just to have a piece of mind(since I plan to keep the car for a long time) . I only have 3100 miles on my 2019 CX-5, so the chance that they find a dislodged rocker arm is small, but who knows. Hopefully it will not be more than $220 (as quoted to JJm86m from his dealership).

 

[Tran7270]

Tran,

Below are my answers based on my experience.

1. That is correct. I could not get the engine to misfire after the dealership reflashed my PCM. The tech was able to replicate the problem prior to reflashing and neither him or I were able to after the reflash.

2. That is correct. Only caveat is that I wouldn't go as far as saying the PCM is disabling the limp mode, however it didn't happen after the update. There could possibly be other factors besides simply disabling limp mode.

3. You don't want loose parts floating around in your engine. Per the recall "In the worst case, engine damage may occur, resulting in engine stall while driving without ability to be restarted, which may increase the risk of a crash." I'm assuming engine damage isn't occurring because of the misfire but because the rocker arm is jamming a cam or worse.

4. Kinda. I never held my RPMs, I redlined it and expected the transmission to auto shift to the next gear @6200. In my case the engine immediately went to limp mode instead. 1st to 2nd gear and 2nd to 3rd gear was enough to get the limp mode to occur. The safer but more costly but more sure way would be to have your dealer remove the valve cover to check. The RPM check has only worked for me Pre Flash.

Thank you JJm86m for the clarifications and additional inputs. Really appreciated. I plan to have the dealership do a visual check of my rocker arm for a piece of mind (cost on me, of course).

 

[cz5gt]

Bear in mind its labor time + new valve cover gasket if you go that route. plus removing few parts, reconnecting them after and there is the exhaust oil control valve seal as well. Thats if you want to go by the book

 

[yrwei52]

I'm not sure any of us completely understands how the HLAs work when switching between normal and cylinder deactivation modes. So I'm open to the possibility that the following reasoning is flawed, but here's how I see it.

What little CD documentation exists online (see this page) describes a lockpin in the HLA that is released to allow the HLA to collapse, which results in no valve lift.

I suspect the default condition for the HLA is to have the lockpin engaged, and that oil pressure is utilized to release it. My thinking here is guided by the expectation that when starting the engine, when there's little if any oil pressure, you'd want all cylinders to be operating. If that's right, the valves in cylinders 1 and 4 would operate when manually turning the crankshaft.

On the other hand, if the default condition is for the lockpin to be released, it's true that the valves in cylinders 1 and 4 won't be operating when the engine isn't running (i.e., when the crankshaft is turned manually).

Whether the lockpin is engaged or released by default, it does seem clear that oil pressure is used to control the lockpin. Oil pressure doesn't directly "supply the force within the HLA" that the rocker presses against when operating a valve.

One of the function from hydraulic lash adjusters or usually people call them hydraulic lifters is to eliminate the need of periodical valve clearance adjustment. So if the engine is not running, the HLA will be collapsable without oil pressure supplied. I fully aware there's a lock pin in HLA for those cylinders with cylinder deactivation. But unless Mazda eliminated the function of automatic valve clearance adjustment from those HLA's, they will get collapsed when there's no oil pressure, hence the valve won't open, or open only partially.

 

[yrwei52]

⋯ The rocker arm is what opens the valve, not the HLA. When locked under normal operation, the HLA keeps the rocker arm in contact with the cam such that the valve must open when a lobe comes around.

Yeah, one end of the rocker arm push down the valve, but the HLA at the other end of the rocker arm is controlling the valve opening by how much it would collapse with proper oil pressure.

 

[shadonoz]

More info from Mazda in this pdf for our edification.

Note the verbiage: "As a result, an intake valve rocker arm may come out of position and make contact with internal engine parts, which may cause an engine misfire and loss of power and/or Malfunction Indicator Light (MIL) illumination. In the worst case, engine damage may occur, resulting in engine stall while driving without ability to be restarted, which may increase the risk of a crash."

No mention of lean condition; maybe that was speculation.

Also: "This occurs when transitioning from cylinder deactivation to full cylinder activation modes while driving, due to inappropriate control logic of the PCM."

 

[GAXIBM]

I think it would be possible to see if all the valves are opening by pulling the spark plug and putting a remote endoscope in the plug hole and manually turning over the engine by hand. The intake valves should open together every other time the piston is (slowly)moved down. (like this one for $20 https://www.amazon.com/dp/ (commissions earned)) You could also run a torque and hp chart on a dyno and see if it is putting out to speck. If one of the valves is not operating it should reduce power by 12% or so at top end.

 

[cz5gt]

Its not lean but rich in the scenario when one valve is stuck closed.
mpg will suffer a bit, also map sensor live data can be used. but its too complicated for the average user to diagnose.

 

[Tran7270]

Let's start researching on how to disable CD permanently on these Mazdas. But keep in mind one problem - Assume there is a way to disable CD, how are we going to know we really disable it when the transition from Cylinder deactivation to Cylinder activation or vice versa is so seamless or unnoticeable as others have reported?

 

[p38cyq]

Let's start researching on how to disable CD permanently on these Mazdas. But keep in mind one problem - Assume there is a way to disable CD, how are we going to know we really disable it when the transition from Cylinder deactivation to Cylinder activation or vice versa is so seamless or unnoticeable as others have reported?

Sigh.

 

[shadonoz]

More from the auto press: A Quarter Million Mazdas Have A Very Embarrassing Problem

"The result of that glitch is an engine stall and possible contact between the intake valve rocker arm and internal engine components. Thankfully, the fix involves nothing more than a software update to the PCM."

Love the spin at the end. They forgot to say "unless there has already been inappropriate contact", shall we say in the vernacular of the day.

The true scope of the problem is slowly coming out. This is a whole lot more than embarrassing.

 

[shadonoz]

Let's start researching on how to disable CD permanently on these Mazdas. But keep in mind one problem - Assume there is a way to disable CD, how are we going to know we really disable it when the transition from Cylinder deactivation to Cylinder activation or vice versa is so seamless or unnoticeable as others have reported?

Once you get the "fix", there should [SHOULD] be no more danger of the RA falling off. The problem is not knowing if it has already fallen off. And the slight lack of power isn't a big problem, but damage to the engine could be.

Someone has to determine a way of doing that other than removing the cover, since Mazda seems to have eliminated that possibility. Perhaps the rich condition could be measured somehow with OBD?

 

[Go Hawks]

Once you get the "fix", there should [SHOULD] be no more danger of the RA falling off. The problem is not knowing if it has already fallen off. And the slight lack of power isn't a big problem, but damage to the engine could be.

Someone has to determine a way of doing that other than removing the cover, since Mazda seems to have eliminated that possibility. Perhaps the rich condition could be measured somehow with OBD?

Also, if the reflash does not disable limp mode, and the rocker arm is already off, the danger to the engine may be the least of your worries if you are presented with the situation that my wife and I were in. The danger to your life becomes the real issue.

 

[7eregrine]

Let's start researching on how to disable CD permanently on these Mazdas. But keep in mind one problem - Assume there is a way to disable CD, how are we going to know we really disable it when the transition from Cylinder deactivation to Cylinder activation or vice versa is so seamless or unnoticeable as others have reported?

Not happening.