2010 speed 3 with P0012 code

[rickkari]

I'm trying to repair a P0012 code on a 2010 Speed 3 for a fellow church members grandson.

History: The kid came home with the car about two months ago after buying it from a buy here pay here lot. The car had a check engine light on. The boys parents were not thrilled as the car is a modified ricer/tuner. It has had some modifications done and some that have been undone as evidenced by the hole in the right fender, front mount intercooler and piping in the back hatch area as well as 2 non stock Mazda air intakes, wiring harness re-routing under the hood etc...

The kid took it to a repair shop that had it for two weeks and then informed the kid that there was nothing they could do about it because of the "modifications". That's when the grandfather asked me if I could look at it.

So.... the kid drops the car of at my house while I'm at work. His grandad told me he thought it was a Mazda 6 so up to this point I wasn't sure what kind of Mazda it was as the grandad was only guessing. He said he just needed me to fix his check engine light issue to make it through emissions inspection. I pull up to find a lowered, tinted window, cat exhaust only with a 4 inch tailpipe, hole in the fender, tires worn to the steel belts and leaking due to lowered negative camber suspension, fuel gauge/odometer inop., short shifter with torn out boot, wrong air intake with no filter, rode hard piece of junk.

Only modification I see is a aftermarket blow-off valve. Factory top mount intercooler installed.

Now for my repair history. Installed new tires as it wasn't safe to drive. Installed a Corksport stage 1 SRI just to have a filter as it was way cheaper than buying the OEM assembly. Pulled 3 rear O2 codes, P0012- VVT cam over retarded, 2 misfire codes.

Found the rear O2 had been replaced with a universal connect to the original harness type. The wiring was corroded green and a couple had broken. Replaced the sensor with Denso unit which fixed the 3 codes and allowed the Cat monitor to finally run and thankfully passes every time.

I first addressed the P0012 code by doing a synthetic oil change as the oil was over a quart low and looked like black water. Didn't fix it.

After driving it with the new SRI, rear O2 and oil change, everything is good except the recurring P0012.

I tried to access the ECU with a Snap On Solus Ultra scan tool but could not communicate. Was trying to get into it to do bi-directional tests on the VVT system.

Went the manual route. Connected a cheap glorified code reader to access the VVT duty cycle and VVT variation from desired position PID's. While revving the engine the ECU called for duty cycle and the variation would go to a negative reading. Connected a VVT solenoid to the connector that I had from a old NA 2.3. It is the same P/N. While revving the engine I verified that the ECU control to the VVT sol. was indeed activating it and moving the plunger from the static retard position to the advanced position. I then connected power and ground to the VVT sol. on the engine with it idling to verify that the sol. and VVT gear were responding. Connected, the VVT would go to full advance and the engine would idle rough indicating that it was working. Did this several times. The scan tool read minus 20ish degrees with VVT energized. As I did it repeatedly it eventually read minus 6ish degrees although it still did the same rough idle.

Decided to check the crank/cam timing. Pulled valve cover and HPFP and installed the crank pin and cam tool. Timing is perfect. Noticed that the crank sensor was a bit retarded on the 20th tooth. Adjusted crank sensor to dead center postion, wiped cam sensor clean of any debris, installed the VVT sol. from my 2.3 NA just to verify that the VVT sol. was ok. Checked VVT filter in cam cap- OK and free of debris and reassembled everything to try it again and see if crank sensor adjustment, different VVT sol. would make a difference. It did not. Set P0012 again.

Verified so far that is good:
VVT solenoid
VVT signal from ECU
Engine timing chain and timing all OK
Oil pressure light goes out and stays out when running
VVT activates when jumped and also by ECU per scan reading

Two things unknown and more if you see something I'm overlooking:
-ECU issue- why does it read minus numbers when it is advanced either by ECU or jumping?
Is this normal and is it possible this ECU has a tune on it which could be causing the issue and not allowing a regular scan tool communication. Only my 2 different code readers work.
-VVT gear issue- could it be bad even though it is moving when jumped or by ECU activation per rough idle or scan reading respectively?

Look forward to everyone's insight. Thanks for reading.

[kTaLgsTo]

A bit of reading... I didn't notice anywhere you saying you checked the cam sensor, only the crank.

Good luck, I'd run away from that car. You seem to be knowledgeable, but if that car was rode as hard as it it sounds then it's someone's stock out and headache... These cars are not good cars if they don't get treated well, I guess I am saying they aren't so forgiving...

p0012 plz help


Error Code P0012 - Unsolvable Mystery

[rickkari]

Originally Posted by kTaLgsTo A bit of reading... I didn't notice anywhere you saying you checked the cam sensor, only the crank. p0012 plz help Error Code P0012 - Unsolvable Mystery

I didn't check the cam sensor as far as a scope is concerned. There isn't a code for the cam sensor. I have access to the factory service manual for OEM testing of this code and it doesn't list the cam or crank sensor as a issue nor does it have one testing them in the diagnostic tree. I just noticed the crank sensor not lined up correctly and was hoping there was an outside chance it was the issue.

The factory test lists the following causes:
-VVT solenoid
-VVT gear
-ECU
-Wiring issue to VVT sol.
-Cam timing
-Oil passage blockage

Based on what I've done all is eliminated with a possibility of it being the ECU or VVT gear issue. At this point I am supposed to remove the VVT gear and physically check the pin and rotor orientation. As far as the ECU the Service Manual says to check everything else and if it is ok, swap the ECU. Short of parts swapping at this point I was hoping someone has had the exact issue and results and could tell me what fixed theirs. If nothing comes up, I suppose I have to disassemble the engine to pull the VVT gear off the intake cam. It just seems to be operating properly by the manual jump test.

[sawo]

Did you ever fix this? I have the exact same issue! There are multiple threads on this and every time the people starting them disappear.

[MSMS3]

Originally Posted by sawo Did you ever fix this? I have the exact same issue! There are multiple threads on this and every time the people starting them disappear.

Maybe PM him and/or the others? He/they might not be following a year old or older thread anymore.

[sawo]

Good idea! Too bad I can't send PMs yet, might search around, see if I can help someone first
I'll be troubleshooting this tomorrow, it is interesting that I've found another guy who had the opposite problem (over-advance), which triggers P0011 and does not come up on searches for the over-advance code. In his case, turns out to be some weird issue where the cam ring gear that triggers the camshaft position sensor gets moved to incorrect position, causing false readings on the timing by the cam sensor even though the physical timing is good.

Not sure how can one check the reading on the camshaft position sensor, though - can't seem to find any PIDs on the OBD2 for that. I don't have AP yet so the best I've found was using Bluetooth scanner and the Forscan app, which can show actual valve timing and difference between the target timing. Seems like most of the time the actual timing sits at 0 and the difference with target timing goes as high as 30 degrees when the car is at load in lower RPMs. Not sure if this is expected, pretty sure it isn't. Too bad he only tested at idle, there my timing is off by -0.xx, but once the car is moving, things are very differnet.

To me, it seems as if the VVT actuator is not engaging fully, even though it changes the idle when 12V are applied and the ECU sends signal to it (voltage and resistance measured on the harness side). I will be checking that cam ring gear tomorrow and the filter on the VVT actuator cap, seems like it can easily be clogged, which may cause issues.

[rickkari]

Originally Posted by sawo Did you ever fix this? I have the exact same issue! There are multiple threads on this and every time the people starting them disappear.

Did not fix. Young man wanted his car back before he shipped out to Army basic training last August. My next step would have been to replace the VVT cam gear. All other things checked out. I believe I checked the cam timing ring position with the cam sensor pulled out of the valve cover at TDC and it was fine, not sure though. This is where others were directing (VVT gear) also as everything else checked out. Others said that it is the nature of the beast. IDK. It did run good other than setting the CEL and P0012 code after hard acceleration.

Back to the cam timing ring, it would have been easy to mark it and use a punch to move it a little toward the advanced side to see what would happen. It could easily be tapped back to original mark. I wish I would have tried that just to experiment. I'm assuming that it is pressed on the cam like the 2.3 N/A engine one is.

Hope you find a remedy for yours.

[sawo]

Originally Posted by rickkari Did not fix. Young man wanted his car back before he shipped out to Army basic training last August. My next step would have been to replace the VVT cam gear. All other things checked out. I believe I checked the cam timing ring position with the cam sensor pulled out of the valve cover at TDC and it was fine, not sure though. This is where others were directing (VVT gear) also as everything else checked out. Others said that it is the nature of the beast. IDK. It did run good other than setting the CEL and P0012 code after hard acceleration. Back to the cam timing ring, it would have been easy to mark it and use a punch to move it a little toward the advanced side to see what would happen. It could easily be tapped back to original mark. I wish I would have tried that just to experiment. I'm assuming that it is pressed on the cam like the 2.3 N/A engine one is. Hope you find a remedy for yours.

I believe I found the issue, but it is not pretty. Here is a video of the actual cause:
https://youtu.be/Lvl61UKk05s
Basically, due to camshaft wear (irregular oil changes, possibly even loose timing chain, etc.), the camshaft moves around and the lobe gets severe scoring. So you get a nasty leak from the cam lobe and the OCV just can't hold oil pressure. I will see if I can source camshaft first and try that, hopefully, I won't have to replace the entire head.

This can be verified very easily. You need a proper scanner tool that can get live feed from the PCM, rather than just OBD data. For ELM diagnostic devices, I've used Forscan which is specially designed for Mazda/Ford, but at least with my ELM dongle, it is very hard to get it to read the PCM streams. Out of 20 attempts, it connected only once. Anyway, once you connect to the PCM either with Forscan with ELM or some proper diagnostic dongle, you need to check "Actual Valve Timing" from the live data stream. If you get reading of 0.5-1 degrees, then your OCV is not doing anything, most likely you aren't getting oil pressure to the VVT. If you rev the car while stationary, you should get readings of up to 25 degrees. You can also monitor "difference between target and actual valve timing", that would show if the computer is seeing problems, but the car has to be moving for this. In my case it is easy to verify it in 4/5th - at 1500-2000rpm go WOT, you would see up to 30deg difference (the computer raises P0012 when there is 5 deg difference for over 5 seconds when the coolant temp is over 70C and the RPM is below 4000).

I guess if the leak from the cam lobe isn't that bad, the OCV would probably work when its colder and you don't stay at low RPM that often, but during the summer and with city driving, it would be the worst. Some people suggest filing down the cam cap a bit using 600 grit sandpaper, but in my case, it is so bad there is no difference, probably because there is a gap sideways as the chain now pulls the cam to the side. In my case, I bought the car with a bad chain, the previous owner didn't think much of it, but we managed to just take out the chain without touching tensioners or anything - it was that bad.


Hope that info helps other people, took me 10 days to triple-check everything until realizing what was going on.

Here is a picture of the cam lobe underneath the cam cap that holds the OCV too:

https://i.imgur.com/Zh2vZkf.jpg