[FIXED] 2006 Subaru Forester 2.5L hesitation/misfire

Ahhh sorry about that, I forgot to post a picture of my actual capture. :blush: Was on the group chat at the same time. May or may not have had a few. :silly: I updated the post with my capture and clearer labeling.

It looks like late cam timing, to me. 

 

But, your Cam/Crank match the known good.

 

It makes me wonder about the tone ring on the crank. Is it possible to put it on backwards? I hope I am not getting you onto a wild goose chase. I am not familiar with Subaru, at all. 
 

Don't worry at all, sir. I'm already deep into this goose chase. :lol:

Did you notice a difference in cam timing bank-to-bank, or just late timing in general? I was comparing IVO on Friday. I could see a difference if I really wanted to, but it wasn't obvious.

Neither the cam or crank wheel can be put on backwards and still produce signals. The cam wheel is on bank two, which means bank one COULD be out of time and not show in the waveform.

Did you notice a difference in cam timing bank-to-bank, or just late timing in general?

Late timing, in general.  I can see some differences from cylinder to cylinder, but each capture is at a different RPM. RPM ranges from 700 - 1500 RPM between captures. If your not tired of pulling spark plugs, yet, take a 5 second capture of each cylinder at a steady idle.

Here is a side by side of your previous captures. 
 

On a side note, Weycraze gave me a tip when it comes to graphing crank frequency.  Fill the entire buffer with data. Having large blank sections of data in the buffer seems to play havoc with the Frequency Math channel.  Shorten your buffer/time-base to just long enough to capture the data you need, without empty space at the beginning, or end, of the buffer.  It seems to make it easier to dial in a nice Frequency trace. It would also be beneficial to crank up the sample rate, too.

I see what you mean about the timing. If I get time tomorrow, I'll recheck one cylinder on each bank at least.

I'll retry the frequency trick, too. These are the nice little tidbits I pick up around here.

wrote:

I see what you mean about the timing. If I get time tomorrow, I'll recheck one cylinder on each bank at least.
 

If you do get chance to retake running in cylinder add an ignition sync. Preferably coil current. If the CKP pulley if "off" it'll probably affect the ignition timing too.

Calling this one [FIXED] for now. Silver bullet version: Swapped cam and crank sprockets from original engine to replacement engine.

But before that, I gotta say thanks to everyone that replied. This was definitely not one of my best diagnoses. BUT I did learn some cool scope tricks along the way. Matt T suggested scoping the cam and crank with a channel on the MIL control circuit, so we'd have a reference for when the 'symptom' was happening. Brilliant.  

After my last post, I did a TON more testing. Didn't find jack or squat. Finally whittled my options down to the ECM or the cam/crank sprockets. There just HAD to be something wrong in the misfire detection strategy. Something about the way the ECM detected misfires at higher RPM was not working correctly. But if the ECM had truly failed, why were there no other symptoms? If the cam/crank wheels were incorrect, why did the symptom only show up at higher RPM? And why do my signals match posted known good waveforms perfectly?

Well, I've never seen one of these ECM's fail. But I still had the core engine around. I'm already this deep, might as well try swapping them. :silly: I took several shots for comparison:

 

This is the crank sprocket that was causing the problem:

 

And this is the one that fixed it:

 

Aside from the color, I could find ZERO difference between them. None. For further comparison, I took crank waveforms before and after, along with the 1-2 and 3-4 coil firings for reference.

Before:

 

And after:

 

I know, I didn't keep the channels consistent. :blush: What a hack, right? But still, aside from the ignition timing (which I think is down to engine temperature), the CKP waveform itself is pretty darn similar.

As usual, the Pico captures will be available to anyone that wants them. wrenchturnsyou@gmail.com. If anyone can tell me what exactly the difference is between these crank sprockets (besides the color), I'll buy you coffee. It would seem that, for whatever stupid reason, the crank wheel used on this generation of Suubahwoo is somehow tied to the transmission or the ECM. NOT the engine. :silly: Neither service info or the documentation that came with the reman engine said anything about needing to swap the sprockets over.

TL;DR If you're long block swapping an EJ253, carry the timing belt sprockets over. The Subaru engineer that thought this was a good idea deserves to get field goal kicked in the nuts.

crank wheels for 36-2-2 are all the same.

Right from 2001 when they first appeared, right on up the years. Turbos use the same wheel.

There can be variances in the cam wheels though

But if you swap any engine, any vehicle, the crank/cam must correspond to the ecu in the car.