Frankenstein Mazda 3

Justin Mobile Technician Utah Posted   Latest  
Case Study
2005 Mazda 3 S 2.3L 4-spd
Crank / No Start

2005 Mazda 3 cranks, no start. The shop that called me on this car had installed a used engine. It would attempt to start for a second then just crank over without firing. This car had been at this shop for six months since the engine was installed. When I got there it was in the typical forgotten about state. Snow on top of it, no battery and a flat tire. We pushed it inside and found a battery for it. So no codes to go off of. I confirmed the complaint. It did flare up to start and then cranks steadily. Nice rhythmic sound to the engine. 

I started my diagnosis with a scan tool. No codes would store after several attempts to start the engine. I then went into data and watched the RPM pid. It had a steady 180 rpm while cranking. No other pids appeared out of a normal range. Next step was to check for spark. My gut feeling walking into this job was a cam/crank timing issue. It’s fairly common to see someone unbolt the crank pulley and loose the timing of these engines due to the no keyway. So I jumped right to pulling out the scope. I connected to cyl 1 coil. 1 lead on each of the two wires. Ground is there as well as power. The ECM would ground the coil 3 times and stop. During the firing of the coil the engine would attempt to start. Ok so I have found an ignition problem and for now I feel we have fuel. I didn’t go into the fuel testing at this point. Fix what you know is broken first is the way I approach a repair. 

I ask the shop owner what he did with this engine. He informs me the camshaft cover was damaged so he used the old one on the replacement engine. I asked him if he took the dampener off. He said he did because the timing cover had been broken also. I asked him if he used the alignment tools to time the engine. He was very familiar with the process and I felt he understood it well enough to move forward without checking them at this time. 

I decide to get the cam/crank sync capture to compare to the known good I have for this engine. The signals are there and plenty strong enough to be seen by the ECM. Comparing the pattern to my known good I see an issue. It doesn’t look like my good waveform. So much not like my good that I can’t begin to check the relationship. I don’t have an ignition sync I can feel good about using as it doesn’t repeat. I decide to use in cylinder pressure to determine where top dead center is. 

I noticed enough of a difference between top dead center in relation to the crank pattern compared to my known good to make the call and recheck the timing. We take the valve cover off and install the pin in the block for the crank to stop against. The flat bar slides almost perfectly into the back of the cams and the hole through the crank pulley is perfectly in line with the threaded hole for the bolt. I decide to have another crank pulley ordered to make sure the clicking of the reluctor is the correct one for this year. New part is exactly like the old one. Moving on this engine was and is timing perfectly. I should have gone after the cam first. My gut and experience took me to the mechanical timing. 

I have another diagnostic mobile tech look at my waveform to verify they see the same thing I do. We both agree the ignition timing is way off and the cam pattern is not the same. I tell the shop owner I feel it has the wrong cam reluctor in the engine and tell him I am going to go home and do more research on the problem. I showed him how the pattern his engine had was 2,2,1,1 and what he needed was 3,1,1. He seemed to understand what I was showing him. On my way home I was on the phone with another mobile tech discussing what I was dealing when it hit me. I had looked at a 2006 pattern. It was way different. In 2006 these engines went to a digital sensor. The engine I have uses analog. I went back to the 2006 and looked at the cam pattern. Sure enough it has the 2,2,1,1 pattern. But the crank pattern is not even close to the same. I determine that the shop owner has a 2006 engine that he transferred the crank pulley from a 2005 as well as the analog cam sensor on. This thing had no chance of starting. 

As I was starting to look for pictures of the different cams to show the shop owner in the morning I got a call from the shop owner after hours. He had found a head from one of the 2005 and older engines in his scrap pile. He checked the pattern and it was indeed the pattern I told him we we looking for. He had installed the cam and now the car was running. 

I would not have felt as confident as I did had I not had the library of known good captures showing the differences between years on these engines. I felt confident they were actual known goods because I gathered them myself. Not someone else telling me they were good. I have been burned by that in the past. Capture and save your own. Don’t count on someone else to have one at the ready for you when you need one. 

Michael Diagnostician

Nice write up and diagnosis Justin!

Ð Awarded
Chris Technical Support Specialist
New York

Great find, as I was reading this I remembered a used 2.3 Mazda I had chased with a different year crank pulley. These can be tough without known good waveforms.

Ð Awarded
Ray Diagnostician

Justin, good work and thank you for the known good wave forms.

Ð Awarded
Geoff Diagnostician

Keith Perkins, had a similar situation recently too. He found a perfectly "good" cam-crank relationship for the wrong year engine on that one. It was a video. youtube​.​com/watch?v=FSQocM… Man-made problems, man, man-made problems.....they just keep getting worse! Great job Justin.

Ð Awarded
Keith Diagnostician

Great catch Justin! And good word on saving your own waveforms!

Ð Awarded
Randall Technician

Justin, Your case was eerily similar to the one that we are working on. Our crank signal appears to be wrong. Thank you Randy

Ð Awarded