When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Just spitballing at this point.
--Can you feel the misfire?
--How does the oem system sense a misfire?
----I noticed that there is a code per cylinder so there's something tying ignition timing to it
----Knock sensor? Seems like the O2 sensors wouldn't be able to respond fast enough to detect a per cylinder event.
----Maybe that system is down?
Now THAT'S a loaded question! There's so much that isn't stock, but here is a data point...I've had to go through this foolishness since moving to ATL 4 years ago. Last year I had to go through this same crap and did pass. Since then, there haven't been any substantive changes made to the engine, or the control systems (other than those outlined above). This P0300 code isn't new, it just is proving more difficult than before to eradicate.
Originally Posted by leboeuf
Just spitballing at this point.
--Can you feel the misfire?
Nope, there are no symptoms that I can feel at all - aside from the lopey idle. No hesitation, no stumbles, no...nothing.
--How does the oem system sense a misfire?
From what I can determine, the OEM system "figures" out the time between pulses from the crank sensor based on the RPM's and then compares that time to what it reads from the crank sensor. If there's a difference (that would cause the "emissions" to be more than 1.5 times normal - whatever that turns out to be) then the system records that as a misfire and stores it. If the misfire event occurs on a particular cylinder (there must be some algorithm tied to ignition signal that determines this) for more than x concecutive times then the system calls that a P030n error. If the system can't fix the misfire to a specific cylinder then it set a P0300 "random misfire" code.
----Knock sensor? Seems like the O2 sensors wouldn't be able to respond fast enough to detect a per cylinder event.
----Maybe that system is down?
AFAIK the O2 sensor has nothing to do with the P030n codes, so that's off the table. However I know that the O2 sensors are ok; a) because I spent too much time getting that circuit right and b) because I monitor that with the Torque app and they are doing their thing just fine. In fact, the readiness monitors associated with the O2 sensors turn green just fine.
What I did find is that certain manufacturers (GM and Ford as far as I can tell) have a "crank sensor learning procedure" that has to be done when/if the crank sensor is changed. The procedure "calibrates" the new sensor to the ECU. So far, I can't find anything online that says the Mazda has done this for the Miata, but the search is on.
Now THAT'S a loaded question! There's so much that isn't stock, but .... [doesn't answer question]
......
Keep the suggestions coming, cause I'm stuck.
So what isn't stock right now? It's almost a guarantee that something you did, that isn't stock, is causing the problem. If the car was completely stock right now, I bet it would pass, because you'd remove/undo something you have changed that is CAUSING this misfire.
You're not going to permanently fix this problem until you find out what is causing it and then correct it.
Possible examples: "I have a huge fuel pump that overpressurizes the oem regulator", or "I run aftermarket coils that the factory ecu isn't controlling correctly" or "I have a PCV/Breather system that is leaking air that the MAF isn't measuring", or "I have something regarding the factory control system that the ecu is not currently controlling", etc.
Well it certainly sounds like you have your thinking cap on and know more about this than I do.
That sounds like a pretty cool solution from mazda. If I understand correctly, a properly running engine will produce some duty cycle and a deviation from that would indicate one of the 4 strokes is wonky during a full cycle.
Are the things attached to the crank in good shape? Any oddity on the harmonic balancer? Maybe a messed up keyway or something? I'm not sure if any valuable information could come from the cam sensor; they're generally kind of noisy do to the belt/gear lash.
First, before going any further, let me reiterate that the car has passed the GA OBD check for the past 3 years - not without issues, but none that were related to the misfire code.
Originally Posted by patsmx5
So what isn't stock right now? It's almost a guarantee that something you did, that isn't stock, is causing the problem. If the car was completely stock right now, I bet it would pass, because you'd remove/undo something you have changed that is CAUSING this misfire.
You're not going to permanently fix this problem until you find out what is causing it and then correct it.
So, what's not stock? I'll only list the stuff that has changed since it last passed the OBD check;
- Exhintake mod (this may well be the culprit)
- Port-matched intake and exhaust ports (???? but not easily un-doable)
- Accusump (can't believe this is affecting anything)
- BE oil pump (ditto)
- Oil cooler (ditto)
Other control system things;
- stock injectors put back in
- stock ECU in charge (obviously)
- swapped coil packs
- new crank sensor
- new stock plugs (NGK BKR5E-11's gapped to .040)
- new NBO2 sensors
Other non-stock things (also tested and passed the OBD check last year and before);
- Magnecor plug wires
- no charcoal canister in the engine bay (lines are looped)
- EGR module connected, but "hanging in the breeze"
- turbo wastegate diaphram disconnected and wired open
- all turbo system-related vacuum lines capped
Everything related to the stock ECU is connected and operational.
Originally Posted by leboeuf
Well it certainly sounds like you have your thinking cap on and know more about this than I do.
That sounds like a pretty cool solution from mazda. If I understand correctly, a properly running engine will produce some duty cycle and a deviation from that would indicate one of the 4 strokes is wonky during a full cycle.
Are the things attached to the crank in good shape? Any oddity on the harmonic balancer? Maybe a messed up keyway or something? I'm not sure if any valuable information could come from the cam sensor; they're generally kind of noisy do to the belt/gear lash.
This OBD stuff has taught me more about the stock control systems than I ever wanted to know. The only upside is that it forces me to go through the car with a fine-toothed comb every year. Last year I fought a code related to the VSS for almost 5 weeks! My speedo was working just fine, but the ECU wasn't getting a signal from the VSS so it threw a code.
The balancer is stock and (I think) is in good shape - the rubber isn't cracked or anything. The keyway is in good shape with no obvious wobbles. I haven't checked the nibs on the trigger wheel, but that's next on the list.
I wish there was some kind of "conditioner" I could place inline with the crank sensor that would send a perfectly clean trigger wheel signal to the ECU so that it can't trigger a P030x code.
This is the only thing that is holding up the show!
First, before going any further, let me reiterate that the car has passed the GA OBD check for the past 3 years - not without issues, but none that were related to the misfire code.
So, what's not stock? I'll only list the stuff that has changed since it last passed the OBD check;
- Exhintake mod (this may well be the culprit)
- Port-matched intake and exhaust ports (???? but not easily un-doable)
- Accusump (can't believe this is affecting anything)
- BE oil pump (ditto)
- Oil cooler (ditto)
Other control system things;
- stock injectors put back in
- stock ECU in charge (obviously)
- swapped coil packs
- new crank sensor
- new stock plugs (NGK BKR5E-11's gapped to .040)
- new NBO2 sensors
Other non-stock things (also tested and passed the OBD check last year and before);
- Magnecor plug wires
- no charcoal canister in the engine bay (lines are looped)
- EGR module connected, but "hanging in the breeze"
- turbo wastegate diaphram disconnected and wired open
- all turbo system-related vacuum lines capped
Everything related to the stock ECU is connected and operational.
Based on this, I'd say it's the exintake cam causing your problems. How did you time the cam? I would recommend either retiming it such that the overlap stays the same as stock, or just swapping to the factory cam to pass emissions. Retiming it more retarded ( to bring overlap back to stock) will get the overlap back to stock and still help your top end power, without the downsides of more overlap at low speeds.
Originally Posted by rwyatt365
This OBD stuff has taught me more about the stock control systems than I ever wanted to know.
Words can not describe how much I've learned about the stock ecu in the last few months. I feel your pain!
How did you time the cam? I would recommend either retiming it such that the overlap stays the same as stock, or just swapping to the factory cam to pass emissions. Retiming it more retarded ( to bring overlap back to stock) will get the overlap back to stock and still help your top end power, without the downsides of more overlap at low speeds.
I was so deflated last night I didn't respond to the rest of your post, Pat.
My exhintake cam is timed by re-drilling the cam gear. I had considered an adjustable cam gear to fine tune the overlap but after reading horror stories about slipping, or broken adjustables I stuck with the fix gear.
And just to hammer the point home, I found this article on m.net this morning...
The P0300 or any P030x misfire code comes when a "misfire" is detected, which is not necessarily ignition related. It just detects when a pulse does not come at the time it was expected to come. Anything that causes this pulse to not occur, incomplete combustion for whatever reason, will potentially cause a P030x...
With the cam timing off a tooth or adjusted such that there is enough overlap that it causes a rough idle, then you will get a P0300 as a side effect. It has nothing to do with ignition...So my theory is that adjustable cam gears may allow you to make finer adjustments than 1 whole tooth to the cam timing in order to get some of that lower-rpm torque shift while not adjusting it so far as to reduce idle quality and invite a P0300. The idle quality may be something we can live with but in TX you can't pass inspection with a P0300.
So far, no P0300! (and now that I said it, it'll come back)
HOWEVER, now that the stock intake cam is back in I'm down on power.
My super-accurate highly sophisticated "Butt-Dyno 5000" (guaranteed 0.01% accuracy up to 5000 HP) says that I'm down 14.27 WHP across the entire RPM range. Add that to the fact that I've been NA for almost 3 weeks now and I'm in real pain here guys. I'm just waiting for the readiness monitors to go green so I can get out of this boostless hell hole.
Try timing the exintake came a bit more retarded to reduce overlap. You may be able to run it and not throw a code, and still see the improvement in performance.
The exhintake goes back in as-is tomorrow along with re-connecting the turbo-stuff and ridding myself of the hateful stock ECU for the blessed Megasquirt. I ordered a Toda adjustable gear yesterday - that just gives me something new to fiddle with.
What was supposed to be a 2 hour job turned into a 7 hour nightmare.
So, I went to de-OBD my car, and return to the world of boost yesterday at lunchtime. I popped the hood, reconnected all of the vacuum lines, took out the stock injectors and refitted the FIC 650's, swapped the stock intake cam for the exhintake, rewired the IAT sensor (that is now properly located in the IC outlet) and even pulled off the Maxxis tires 'cause it's supposed to dip into the 30's this weekend.
I was done in just over 2 1/2 hours and went to start the car and...nothing. A lot of coughs and sputters, but no running. WTF! So I checked everything and found that I forgot to hook up the ground wires for the injectors. Silly Me! I connected that up and...nothing.
So, I opened up the valve cover and - lo and behold - the cam timing was off! How did THAT happen? I used the "zip-tie the sprockets to the belt" method. It worked when I swapped out the exhintake cam. Oh well, I reset the timing. Double-checked it and triple-checked it before buttoning everything up. Went to start it and...it runs, but it idles at 2000 RPM and there are horrible noises coming from the front of the engine.
I shut it down and check the engine. I had forgotten to tighten the bolts on the water pump pulley. So I rescue the belt (that was the noise) and tighten down the pulley. But why was the engine running so weird? I decided to hook up the stock ECU and see what happens. It idles (poorly) at 2000 RPM too.
W...T...F...!!
So, I decide to go back to the stock injectors. I pull out the FIC's and put the stockers back in. I did notice that the metal gasket that sits between the upper and lower IM wasn't properly in place, but I didn't pay it much attention. I start it back up and...there's that NOISE again...this time the IC hose was rubbing on the crank pulley (how did THAT happen). But at least the car was idling properly on the stock ECU with the stock injectors.
Now I take out the stock injectors and put the FIC's back in (this time, properly seating the metal IM gasket). This time the car won't start with the stock ECU and the FIC's - which is as it should be. So I connect up the MS and...wait for it...THE NOISE (I know I was getting loopey now, between the time spent and the gasoline fumes from swapping out the injectors so many times, I've made the same stupid mistake 3 times). But the car is running properly now. I button everything up and head for home (all of this happened at the garage where all my car-stuff "lives").
Moral of the story;
Don't be over-confident and think you can do stuff blindfolded and half-asleep.
Check your work ("measure twice, cut once") - rework is a bitch.
Gasoline fumes are not a substitute for oxygen.
Boost is good.
Theory as to why the car ran weird after the cam timing was fixed; I think the metal gasket was not properly in place and trapped between the two halves of the IM, allowing a huge air gap all around the IM. This was acting like a "super idle air valve" allowing mucho air into the mix and pushing the idle to 2000-ish RPM. Of course the car couldn't run properly with this huge air leak. When I re-did the injectors, I closed the leak and - voila - a running engine! So, learn from Uncle Roger, please check your work.
Nothing much to report. Got the helper springs, Torrington bearings and separators installed on all 4 corners. Started ordering stuff to replace the POS NB1 coils with GM coils; I've got this bracket in-hand, and the pigtails from Ballenger (don't feel like crimping wires this week). Next on the list are the coils themselves (again, EBay to the rescue). After that comes the task of rewiring for the new coils. I've read the coil mega-thread about 100 times, so I think I'm marginally ready to tackle this. I'm going to wire a connector into the coil harness so that I can put the stock coils back if I need to (for the state-sponsored annual OBD2 check/rape).
Wish me luck!
Also on the horizon is a change from the current MBC to EBC - time to join the 20th century (wait...that's over, right...damn).
So...after months of absence (from this thread) because there was nothing significant to post (unless anyone is interested in reading about my horrific attempts to create a stable cold-start through all of the crazy temp extremes we enjoy here in northern Georgia - 100* afternoons in the summer to 20* mornings in the winter).
Anyway, I've now "graduated" into the Big-Boys School of GM Coil-ness - after months of researching, gathering parts, more researching, procrastinating and fretting over my terrible lack of electrical skills. I finally got up enough courage to put my conglomeration of pieces together into the car and get it running on the GM coils.
I installed the brackets (thanks, TheBandit) and coils (thanks EBay), hacked up a spare harness (thanks, Tom from The Parts Group), used the wiring diagrams from the LSx coil thread and put it all together. Then, I re-gapped the plugs to .035 (up from the .020 I've been using) and using the dwell settings from psyber_0ptix, I turned the key and fully expected to release the special "electrical smoke", but...it started! ...and RAN!
I was so happy, I shut off the car and ran up from the garage and fixed a large plate of spaghetti that my wife had fixed for dinner - I didn't want to push my luck!
I'll go back later (maybe tomorrow...just to keep the high for as long as I can before the inevitable disappointments) and buttone everything up, and try to optimize settings. But for now...I declare VICTORY!!!