Diagnostics Fun... (problem solved)

[crazycanadian]

I figured for fun I would post this up... Its not everyday you see this failure...

My little DD/autocross/TSD rally/ Track day car developed an odd miss fire back in December

The miss fire has been extreamly random and proven difficult to track down... I could drive the car all day long all over the city and never have a problem... Idled smooth, has lots of power.... After a lot of driving I finally narrowed down when I could get it to act up... I had to drive at a constant speed for at least 2 - 3 min, with the engine over 3000rpms... It would start out small and get worse but if I did all this I could feel the miss start and eventually get pretty bad to where the car was jerking pretty good.. It went away as soon as I romped on the throttle or if I started varying my rpms/throttle input... I have no trouble codes, and no check engine light at all...

So where would you start if this was your car??

I finally had time this weekend to bring home the lab scope from work and hook it up to my car...


Here is what I found...
(Red = CMP sensor, Green = CKP sensor from the distributor, Blue = ignition control signal from the computer, yellow = random injector current ramp)


Here is another pattern...
(Green = CKP sensor from the distributor, Blue = ignition control from the computer, Yellow = Primary ignition patter/Tach signal)



This is a zoomed in look...
(Green = CKP sensor from the distributor, Blue = ignition control signal, Yellow = primary ignition pattern)


So whats the diagnosis???... There is a slim chance my mass air flow sensor is dropping out causing the irregularities with my ignition control signal... But look closely at the dip in the ignition control signal in the zoomed in look... This dip is most likely caused by the driver in my PCM crapping out... :-\ :'( Damn... hopefully I can find a PCM for a 93-94 V6 mx6/626/Ford probe gt before friday... I am signed up to race at in the Thunderbird TSD rally next weekend...

While I was at it all I hooked up to my gf's car... It uses the same distributor and computer controls as my car and has no driavablity problems...
(Green = CKP sensor, Blue = CMP sensor, Yellow = ignition control signal)


One thing that isn't posted up that I did test was the CKP sensor down on the crank... I forgot to save the capture I had gotten... There were no irregularities with this sensor either when the motor was missing... Just for fun I also disconnected this sensor to see if it would change anything... The CKP sensor down on the crank is only used at idle or low speed for engine stability/smoothness... If I spent more time and looked at injector pulse width this is where I probably would have seen a change... Besides a slight missfire at idle the car had no problems with this sensor disconnected...

[SuperK]

Blame it on the EGR!
not serious about that one.

My air flow sensor went bad in my V6. It would work great the first day. Second day, jerkajerkajerk while on lower speeds. Then 3rd day, it was on all speeds, and completely unbearable. Then I would let the car sit for a week and it'd restart the cycle.

I think ryan's got a buncha extra ones, I would holler at him for a spare to try.

[Ryan]

NEAT TOOL! I WANT!!!!


Aside from that, I'd think less fancy. Start with O2 sensors. you could hook your dealie up to them too. They are disrearded until the motor is warm and the RPM's have exceeded 3000 RPM.

Next up is VAF. Easy to check. Dealie could look at that too.

[crazycanadian]

Well wouldn't you know it... I installed a new PCM and I still have the same problem... So since the ignition control signal is an output we all know "Garbage in=Garbage out".... I had a feeling this was the going to be the case... Now we get to discover what else Mazda decided is important in controlling ignition dwell time...

NEAT TOOL! I WANT!!!!


Aside from that, I'd think less fancy. Start with O2 sensors. you could hook your dealie up to them too. They are disrearded until the motor is warm and the RPM's have exceeded 3000 RPM.

Next up is VAF. Easy to check. Dealie could look at that too.

You are definatly right that I needed to look at these thing... My next look is at the air flow meter... I ran out of time on sunday to look at these things.. Since I am good friends with the place where I can get used parts from I'll update this tomorrow after I have tryed out a different air flow meter...

Just a bit of back ground info... My KL swap has about 35,000km on it since I bought the car, finished the swap and got it on the road... Plugs/wire/cap/rotor/fuel filter are all a week old... I replaced them all due to fact I run the s--- out of the car and at this mileage I just replace the parts for maintenance purposes... Distributor I replaced 6 months ago when I had a random stalling/no start problem..... It was a new, aftermarket replacement distributor... O2 sensors were both brand new when the car first got up and running... EGR valve has been both active and blocked off during this problem...

keep tuned... tomorrow night is dodge ball night so I might not get around to working on the car to much.. But I should have an update if the mass air flow sensor helps or not...

[crazycanadian]

Well picked a few more things off the list and might possibly have an idea where the problem lies...

I revisited the CKP sensor on crank shaft today.. I wanted to have a closer look at its actual pattern... There were no variation or odd drop out when my car was acting up but I did notice irregularity thats presents all the time... I'll post pictures tomorrow...

While I was out driving around I also took the time to hook up to my TPS and air flow meter signals... No problems or irregularities here so I moved on...

After that I tryed disconnecting the knock sensor... all though you could feel the car lacking in power under load(the computer is in limp mode and pulling timing) I still got the same miss fire problem... So we cross this one off the list...

On my last good stretch home I unplugged the CKP sensor on the crank shaft... All though the car doesn't idle that great like this, I could not get the car to act up... So I think tomorrow after work I'll look at this a little more closely and possibly replace this sensor...

[crazycanadian]

Tonight I put in some more time on the car... Still no luck... I installed a new CKP sensor with no change...
Since the car is a home built with harnesses mixed and matched all threw out the car I didn't want to try and find all the actual ground locations.. So I took the computer grounds, spliced into them and ran a couple extra grounds... No change...I also voltage dropped them and saw no problems with resistance on the grounds.. On the road I disconnect my alternator, double checked that it wasn't charging anymore and drove the car... No change...

I did have some success... I disconnected the CKP sensor down by the crankshaft and drove the car a lot longer then I have in the past... It never did act up.... Once I reconnected it the car started acting up again but took a little bit... I have a feeling it has to do with putting the computer into a limp mode though... I had changed the sensor and it didn't solve the problem...

While I was at it I took a couple more scope pictures... At this point I am going back to what a lot mechanics do.. replace parts based on pattern failures... I'll send my distributor back and get warranty on it....

Scope pattern from last night... (Yellow= ignition control sig, Green =TPS, Blue = CMP, Red = Air flow meter)... No dips drops or anythere here that I can see...


Scope patterns from tonights run...
A close up look of the pattern from the CKP sensor down on the crank shaft... This pattern didn't make sense to me last night... The data on the Verus has the wrong wave form picture in there data base... I didn't know this until tonight when I had a look at the reluctor ring on the crank shaft...
(yellow = Ignition control signal, Green = CKP sensor down on the crank, Blue = CMP)


The reason the pattern looks this way is because the reluctor ring is set up like a saw blade... The pattern show on my verus would have come from a reluctor ring that has normal square teeth...


Can't see any problems with this CKP position sensor...
(yellow = ignition control signal, Green = crank shaft CKP sensor, Blue CMP)


Here is an interesting one... I current ramped all 6 injectors to see if they were being affected by the problem as well.. turns out its not just coil timing thats being affected... Injectors are being missfired as well.. The spike in the current pattern is where the problems occuring... You see this spike because multiple injectors are being fired really close to one another... When this happens the current signatures from each injector start over lapping one another...
(Yellow = ignition control signal, Green = CMP sensor, Blue = injector current ramp)

[Ryan]

The sensor down by the crank is totally unnecessary for normal operation. The ECM ONLY uses that signal over 4000 RPM to help with ignition resolution, to add a few ponies. Without it, your timing is just a wee bit less aggressive. The only possible way it could cause issues is if the teeth were misaligned, like on some cheap Ebay UDP's, in which case the ECM would get very confused.

That is all. Quit looking at it, it is not causing your problem.



If you think both the coil and injectors are both misfiring, that is absolutely no surprise, because they both run from the same signal; the NE1 signal from the distributor, using the G signal to find C1 TDC.

Also, have you looked at your O2 sensors? There is a procedure to check the readings in the manual.

[crazycanadian]

The sensor down by the crank is totally unnecessary for normal operation. The ECM ONLY uses that signal over 4000 RPM to help with ignition resolution, to add a few ponies. Without it, your timing is just a wee bit less aggressive. The only possible way it could cause issues is if the teeth were misaligned, like on some cheap Ebay UDP's, in which case the ECM would get very confused.

That is all. Quit looking at it, it is not causing your problem.



If you think both the coil and injectors are both misfiring, that is absolutely no surprise, because they both run from the same signal; the NE1 signal from the distributor, using the G signal to find C1 TDC.

Also, have you looked at your O2 sensors? There is a procedure to check the readings in the manual.

The ECU uses it for more then just ignition resolution over 4000rpm... When I disconnect it and run the car with out it connected I pick up a slight missfire at idle.... Also my problem goes away... My problem is happening at a cruse between 3 - 4K... It might not be my cause, but it is related to my problem...

I know how to test O2's and scoped them out last night just didn't post up the pictures...Nothing to report there...


Today I discovered that the distributor I bought last year was actually on jan 19th... The company I bought it from wouldn't warranty it anymore....

[crazycanadian]

Hey Ryan... I gotta thank you for your insight about the cheap Ebay under drive pully.... I wasn't aware of this problem.... Just so happens I have an Ebay pully on my car..

So that leads me to my fix...

Distributor timing... To far retarded was the actual problem... So why didn't I check this earler?? Well its simple.. I haven't changed the timing on the distributor since Jan 19th 2011... Why do I know the date? cause I just found out this morning I didn't have warranty on my distributor and thats when I bought and installed it... So why did it act up now?? Well back at the end of Nov I had gone out to help a friend get his K8 swapped BG up and running... While at it we had swapped the distributor out of my car and stuck it in his... When I did this all I did was mark where I had it and pulled it out... When I installed it I just lined my marks back up and went on my way...

Looks like mazda uses the distributors crank sensor as its primary signal to decide ignition and injector timing.. All though at certain times and under different load conditions it'll reference the crank shaft crank sensor and use it to determine ignition and injector timing... I managed to get mine distributor set at a spot where under the right conditions the computer got confused and threw a fit causing my missfire...

I think there is more to all this but I am going to get my hands on some parts first and I'll continue all this later...

[Ryan]

you're not telling my anything new about the timing sensors .

The sharpie mark trick only works if you replace it with the exact same unit you took out, (like during disassembly for maintenance). A different unit won't be lined up just right due to factory tolerances, and just a degree or two is a giant mix up as far as the ECM is concerned.


Take out your cheap pulley, and hold it back to back with the stocker, and line up the keyways. I think your problem will become apparent.

[crazycanadian]

you're not telling my anything new about the timing sensors .

The sharpie mark trick only works if you replace it with the exact same unit you took out, (like during disassembly for maintenance). A different unit won't be lined up just right due to factory tolerances, and just a degree or two is a giant mix up as far as the ECM is concerned.


Take out your cheap pulley, and hold it back to back with the stocker, and line up the keyways. I think your problem will become apparent.

I changed my distributor back in Jan of 2011... No problems with it until november... When I did the sharpy thing I just borrowed the distributor off my car to try it on my friends car... I installed my own distributor back on my car...

I will be taking the pully out and checking... I was told by the previous owner its a corksport pully though..

[Daninski]

Interesting read, thanks for posting.

[kira_kira]

thanks guys very informative. this post has nothing to do what so ever with what i was searching for but i found myself reading this thread top to bottom =D

[crazycanadian]

Since someone posted here.. I do have a small update... I have been to busy with house reno's to go get a new crank pulley.... I did how ever attempt to use a timing light to get the distributor timed properly.. Either something is messed up in my wiring or something it defiantly wrong with crank pulley.. From the looks of everything the distributor was actually to far advanced causing my problem.. I cranked it back and the closest I could get was 14 deg base timing...

Alldata was telling me its suposed to be 10... So something still isn't right.. Hopefully when I get house reno's finished, my house back in order and get my garage back I can play with it some more...

Either way the car has been running.. Just wont pass emissions right now.. I've toasted the coverter with all the autocross and track days I do..

[Ryan]

Then your belt jumped a tooth. It happens. pieces of the timing cover gaskets just get old and fall apart into there, and bam, jumps a tooth.