Jump to content

ECU no longer firing 2 cylinders...


jigga009

Recommended Posts

Odd title, but let me explain what happened...

 

crusing along a highway in my Subaru when all of a sudden I could detect a misfire in progress. It felt as if the car was no longer running on 4 cylinders.

upon diagnosis on the side of the road, I came to the realization that when I pull all 4 spark plugs, cylinders 1 and 3 do not spark together during an ignition test for cylinders 1 and 3. If on the other hand I test just cylinder 1 or just cylinder 3, the spark plugs spark.

Car runs a pair of 2 channel CDI ignition boxes for each side of the engine. I have tried swapping a known good CDI box in on the side of the engine with cylinders 1 and 3, and the same thing happens. 

I have been able to rule out issues with the ignition coils themselves since I can get cylinders 1 and 3 to fire in individual tests, but not together. I think I have also been able to rule out an issue with the CDI box since I swapped a known good box in and same thing happens.

Does this sound like an issue with the ecu itself?

if anyone has any ideas, I would be grateful. I do apologize if the message is a little disjointed, as I’m posting from the side of a major highway on my cell phone:( 

Link to comment
Share on other sites

I did help Femi as best I could earlier.  The tests that we could come up with using the only basic tools he had access to roadside (straightened out his key ring to use as a "back probe" lol) suggested the ECU was sending out ignition signals on all channels but it was far from conclusive.  

Link to comment
Share on other sites

lol I was channeling my inner MacGyver there!  :D

Thanks for the roadside tips yesterday; much appreciated!

Got the car taken home on a flatbed late last night and performed some more repeated testing today once I got my head back on straight and more relaxed. I was also now able to properly back-probe the ECU using a more appropriate probe repeatedly in order to gain some meaningful information.

Here is what I've been able to find out so far today:

Each of the ignition outputs of the ECU is putting out a frequency of 3.7khz regardless of whether the ECU was commanding the output to fire or not. This was the case with all 4 ignition outputs. As soon as the key is set to the ON position and the ECU boots up, all four ignition outputs in use show 3.7khz. If I turn on individual cylinders through the ignition test on the ECU, each output continues to output at a frequency of 3.7kHz. If I turn the car off (along with the ECU), the frequency from each ignition output drops to zero. This was the case also after disconnecting the ignition boxes from the equation. Each back probe probe sat at 3.7khz as long as the ECU was powered up, regardless of whether it was commanding a spark plug into action.

I was also able to better observe what is going on in th engine bay:

When Ignition output 1 only is commanded on from the ECU, spark plug #1 (bank 1) fires vigorously.

When Ignition output 2 only is commanded on from the ECU, spark plug #2 (bank 2) fires vigorously.

When Ignition output 3 only is commanded on from the ECU, spark plug #3 (bank 1) fires vigorously.

When Ignition output 4 only is commanded on from the ECU, spark plug #4 (bank 2) fires vigorously.

When Ignition output 1 and 3 (cylinder bank 1) are commanded on together, Cylinder 1 does not fire at all, and cylinder 3 fires.

When Ignition output 2 and 4 (cylinder bank 2) are commanded on together, Cylinder 2 fires sporadically, and cylinder 4 fires more strongly.

When all 4 ignition outputs are commanded on at the same time, cylinder 1 [bank 1] spark plug is dead, Cylinder 3 [bank 1] spark plug is firing strongly, Cylinder 4 [bank 2] spark plug is firing strongly, and cylinder 2 [bank 2] spark plug is firing sporadically.

 

Adam mentioned that in his experience, some CDI boxes are not happy firing both channels at the same time, but I have always been able to turn on all 4 ignition outputs at the same time from the ECU and watch all 4 plugs sparking vigorously at the same time in order to diagnose CDI box failures - I would just pull all 4 plugs and ground them against the intake manifold turn on all 4 ignition outputs with the spark plugs out of the cylinders, then hop out of the car and see which side of the engine is asleep.

I switched a known good CDI box that I carry as a spare into the left and then into the right side banks, and the exact same phenomenon as above occurred, so I'm quite sure that this isn't a CDI box issue.

I checked the resistance on the primary and secondary sides of the ignition coils in the engine bay, and all were in spec according to MSD, so I'm quite sure that this isn't a coil issue.

I checked for resistance with my plug leads, and again, all were in spec according to MSD, so I'm quite sure that this isn't a spark plug wire issue.

I tried performing the above experiment with a new set of BKR7e plugs, and the exact same phenomenon occurred, so I'm quite sure that this isn't a spark plug issue.

As for wiring issue -- I don't want to cloud people's thoughts, but since I can turn on individual cylinders at will, and they spark vigorously as long as they are the only cylinders turned on, would that not suggest that the wiring is fine?

Fortunately, I was actually data-logging the car when it decided that it was time to go to sleep on the highway. You would need to fast forward to the end the datalog for the fun bit:

Dropbox link to datalog

 

Link to comment
Share on other sites

In an attempt to take power supply issues out of the equation, I fully charged the battery last night as well as hooked up a power supply to the car, disconnected the relays to my fuel system (since the pumps present an electrical load when key is turned to the ON position) and ran the testing again today. Same results as above :(

Link to comment
Share on other sites

Slight update:

I called the manufacturer of the CDI boxes on the car just to have a thought experiment with them regarding the particular scenario I have, and their lead tech mentioned that he had never seen such a failure before relating directly to a car running 2 CDI boxes in parallel before.  They could see perhaps one CDI box failing (in which case only one bank of cylinders would be affected), but not two going out at the exact same time.

In addition, switching in my spare box to test and coming up with the same results as well as being able to fire each channel on their CDI boxes independently was enough for them to rule out the CDI boxes at this time. For them, it has to be an issue upstream from the CDI boxes that is causing this. 

They mentioned that all the CDI is doing in the car is amplifying whatever signal that it receives directly from the ECU, so if it is that I am able to command each spark plug on individually through the ECU, it would suggest to them that the wiring between the CDI and the ECU is intact and functional, but if I am unable to fire all 4 plugs at the same time, I might want to take a closer look at the source for the ignition signal.

They mentioned that on my particular setup, if a CDI box failed on my car, I would be seeing symptoms on one bank of cylinders only (i.e. the driver's or passenger's side of the engine), since I have separate boxes for each bank, and that switching in a known good CDI box would confirm and fix the issue (which it has done in the past when I've experienced a CDI box fail).

 

 

 

Link to comment
Share on other sites

Ok new day and a quick update  for all who have been advising me so far  -  I've been able to rule out the ECU. Adam was gracious enough to talk me through the process of using a test light on the ignition outputs of the ECU. I purchased one this morning, and sure enough, all ignition outputs were firing when commanded to do so via the software. Even when all were commanded on at the same time, I got a positive flashing LED on my tool when I probed the 4 ignition outputs.

So back to the drawing board I guess... I'm beginning to suspect something more general.. such as a grounding issue of some kind that happens to be manifesting itself in the ignition system and possibly other systems in the car that I did not noticed yet. I'll do some more poking around and check back.

Link to comment
Share on other sites

1 hour ago, Ducie54 said:

Try running an extra earth cable to the cylinder heads. Jumpers leads is a quick way to do it.

Thanks for the suggestion! I tried this just now but still does the same thing, so maybe not a grounding issue after all?  :(

Link to comment
Share on other sites

New update for those watching in the background!

I ran further tests using my new-found toy, the LED test light (Thanks Adam!!). 

 

After back-probing the harness of the CDI boxes that serve the left and right banks of the engine, I was able to determine that the signal to fire the coils is in fact reaching the CDI boxes from the ECU, so given the following bits of info that I have accumulated so far:

- ECU is definitely firing its ignition outputs (confirmed by back-probing at the ECU itself)

- I am able to command on each spark plug by itself (suggesting that the wiring pathway between the ECU and the CDI boxes is fine)

-Signal from the ECU is reaching the CDI boxes on both sides of the engine (I back-probed the connector going into each box for the signal from the ECU, and the light blinked at the roughly 10Hz that it did when I back-probed at the ECU connector inside the car.)

- Coils test in spec

- Spark plug wires test in spec

- I used new spark plugs just to ensure that spark plugs are not the cause of the issues I am experiencing.

- CDI boxes are in fact receiving a full 12.87V from my battery

- I can't actually measure the output from the CDI to the coils since it is somewhere in the 500V range, but I can observe whether a spark is jumping at the plug though.

 

I'd conclude that as unlikely as it seems, both the CDI boxes are at fault.

 

There is a bit of a plot twist to all of this though which made the scenario very confusing from the start:

1) I swapped what I thought was a known good box into the car, and the problem persisted, leading me to prematurely rule out issues with the CDI bo. After back-probing this spare box and observing that the ECU is in fact  commanding it to fire, AND that it is receiving appropriate voltage from the car, it would seem that my spare CDI box was a dud also.

2) Both boxes on the car appeared to have failed in similar manners, but to different extents... 

Box 1 can fire each cylinder individually, but when placed on the driver's side of the car, can only fire "1.5" plugs (fires one strongly, and the other one weakly). When on the passenger side it can only fire one cylinder when both cylinders are commanded to fire.

Box 2 can fire each cylinder individually, but when placed on the driver's side of the car, can only fire 1 plug when both are commanded on. When on the passenger side of the car, it can only fire one plug when both plugs are commanded to fire.

My "spare" box appears to have the same issue - can fire each plug on its own, but only one plug out of two when commanded to fire them both.

 

I suspect that in essence, I had been running on 1.5 cylinders on one side of the engine and the full 2 cylinders on the other side for a long time without noticing much loss in power, and never really noticed it until the box that was running at full strength failed and dropped down to 1 cylinder, dropping me down to 2.5 cylinders out of 4, which I could definitely notice, and caused the car to lose the ability to maintain speed.

If there is anything that anyone thinks I should check in addition that could change my conclusions, I'd love your suggestions. 

 

 

 

 

Link to comment
Share on other sites

Because you've got 2 slightly different failure modes between left and right side, swap the boxes over left to right and confirm the fault follows the box. 

Also try swapping front and back outputs and inputs on each CDI box. This will mean your ecu>spark plug mapping is still correct, but you'll be using the opposite channels within the CDI box. Depending on the wiring effort, you could also swap the wiring around so instead of being left and right, the CDI boxes ran front 2 and back 2 cylinders.

My idea here is to rule out anything (ie grounding, shorts between certain wires) that is only impacting the front 2 cylinders.

Link to comment
Share on other sites

Cj, thanks very much for the suggestion.. I'll try this tomorrow and see what happens. I should be able to switch the wires on each bank around, but won't be able to switch between banks without some major rewiring.

 

Another frustrating and confusing mini update from today:

I was in touch today some more with the manufacturer of the CDI i'm using in order to obtain an RMA#, and a different tech wanted me to describe everything from the top again.  While he was more than willing to give me an RMA to send them in,  he was confident that there is something different about the way the G4+ calls for individual cylinder fire within its test mode compared to what it does when you actually run the car, or when you turn on all channels. I asked why it was they believed this, and they stated that it is because if each channel works in isolation (which they do), the CDI is fine. They mentioned that once I figure that out, I would know why it is that it "appears" that each CDI is not firing off both of their respective channels. 

To test, they had me revert to the stock ECU while keeping the CDI boxes in place and crank the engine to see whether I am still only getting one cylinder on each bank firing, and lo and behold, during cranking on the stock ECU, all 4 plugs were firing. I was stunned.

I then tried the exact same thing on the G4+, actually cranking the engine and observing the plugs, and it also fired all 4 spark plugs sequentially just as on the stock ECU.

Now that I established that both the stock ECU and the G4+ were both capable of firing all individual spark plugs sequentially through the same CDI boxes that were on the car when it stopped, I decided to try the G4+'s ignition mode again in order to see what happens, and sure enough, as the manufacturer of the CDI guessed, something changed. It looked just as it looked before - as if the engine was again not running on 4 spark plugs. I could command on each individual plug in isolation just fine with a nice spark, but could not observe a spark in all 4 plugs if I turned all cylinders on and sat back to watch.

 

 

Link to comment
Share on other sites

3 hours ago, jigga009 said:

I then tried the exact same thing on the G4+, actually cranking the engine and observing the plugs, and it also fired all 4 spark plugs sequentially just as on the stock ECU.

So you have confirmed that you cant fire the plugs together, but they fire sequentially, so then why does the engine not run? 

 

3 hours ago, jigga009 said:

he was confident that there is something different about the way the G4+ calls for individual cylinder fire within its test mode compared to what it does when you actually run the car, or when you turn on all channels. I asked why it was they believed this, and they stated that it is because if each channel works in isolation (which they do), the CDI is fine. They mentioned that once I figure that out, I would know why it is that it "appears" that each CDI is not firing off both of their respective channels. 

Of course it is different when using test mode to when running an engine.  When in test mode if you have all drives in test, all outputs fire together at exactly the same time (as per the video I sent you), that never happens in a running engine.  Did they state that the CDI box can actually fire all channels simultaneously?  I think I mentioned to you earlier that I have heard that many CDI's cant actually do this since either the transformer or capacitor is shared between channels.  

Link to comment
Share on other sites

5 hours ago, Adamw said:

So you have confirmed that you cant fire the plugs together, but they fire sequentially, so then why does the engine not run? 

 

Of course it is different when using test mode to when running an engine.  When in test mode if you have all drives in test, all outputs fire together at exactly the same time (as per the video I sent you), that never happens in a running engine.  Did they state that the CDI box can actually fire all channels simultaneously?  I think I mentioned to you earlier that I have heard that many CDI's cant actually do this since either the transformer or capacitor is shared between channels.  

Thanks for the response Adam,  I'm honestly at a bit of a loss as to why the thing stopped in the first place if the Link is firing each plug sequentially. I definitely recall you making the comment about not thinking that many CDIs can fire all channels simultaneously, but I now better understand what you meant. To partly answer your question (because it definitely came up during the conversation with the manufacturer yesterday), the CDI does in fact have an independent capacitor for each channel. It is for this reason they are sure that their boxes work properly since I can command each channel on in isolation, and each channel definitely fires the plug (sequentially when cranking the engine and through the ECU's test mode). My plan for today includes installing the plugs into the engine heads and actually see if it runs with the G4+. If not, I will try the same on the stock ECU to see what happens.

As far as why the engine stopped in the first place, I am not sure...If you look at the log I included of the event, you would see that I was literally cruising at speed when all of a sudden it seemed to shut down by itself. The only reason why I initially suspected a CDI box issue is because that is the typical failure mode I've experienced with them... You would be driving, when all of a sudden you drop one or more cylinders and the car can no longer carry on at speed.  Luckily, I was data logging all channels directly to my laptop at the time it happened, so I do have a record of what the ECU was doing and seeing on all channels, but I'll be the first to admit that I am not skilled enough to see anything that is glaringly wrong in the log. I have included the log in my second post in case something jumps out at you? - Datalog from when engine stopped running

I'm going to follow up with them today regarding the second part of your question - whether the CDI can actually fire both channels simultaneously. My guess after seeing what I've seen so far with the existing CDI's that were in the car when the issue hit, and when I switched in my spare box is that it cannot (just as you predicted), but I will wait and see what they say about that part. I will say though that they were very curious initially as to how exactly I was triggering the CDI to fire all plugs at the same time. It was when I explained that it was through a standalone ignition test mode and which standalone it was that the gears started to turn in their heads. Even then, they wanted to know exactly how the ignition test mode on the G4+ worked..something I could not answer, except to say that it signals a firing at a frequency of 10Hz to the CDI box which I have confirmed is definitely reaching the box on both channels.

I will let you know what they say with respect to being able to fire both channels simultaneously.

 

 

 

 

 

 

 

 

3 hours ago, neil brown said:

Just a question

 

what Subaru are you running and have you 4 coil on plugs 

 

could you use Subaru newage wrx or sti with igniters built in or simple ver 3-4 wasted spark set up with 1 ignitor box 

Hi Neil, thanks for your response.

It is a new-age Subaru bugeye car, but the coil on plugs are long-gone. They had to be pulled because the stock ignition system was causing some rather large transient voltage drops in the car at elevated power levels. Switching ignition systems was the only way to stop this happening. 

The stock ECU does have the ability to run the engine with the CDI system though, but I have to run the CDI's in their non-restriking mode.

Link to comment
Share on other sites

Weird

 

cause I had no issues running ver 3-4 waste spark Subaru system at 1.7bar of boost pressure using standard Subaru coil and ingnitor on a G4 storm wire in module 

Will be moving to Subaru sti newage coil on plug when I upgrade the turbo on waste spark set up 

Link to comment
Share on other sites

Something that occurred to me today while poking through my utterly normal log of the event was something that I used to experience with my old Vipec V88 during summer months under very similar circumstances to the day the car stopped running with the G4+..

During summer months with my V88, it had a habit of shutting down the car on hot summer days (usually this time of the year) if I drove for long periods of time without cooling the interior of the car with AC or with windows down.

Once I let the car cool sit and off for a period of time, it would start up again like nothing ever happened, and then start shutting cylinders down again if I drove for a while again without AC or without windows down. I was able to deduce back then that the ECU might have been getting too warm underneath the passenger footwell where there is no airflow and the downpipe of my turbo passes nearby in the engine bay.

I had not experienced the same thing with the G4+ since I only picked it up about 6 months ago during winter months, but this is my first summer running this ECU and when I think about it, the conditions seem a bit too similar for comfort, with ambient temps in the 28 degrees C range with clear skies on that particular day. I imagine that it was even hotter in the interior of the car since the windows were not open either.

I checked my log, and I was running the car on the highway for about 99% of my trip without the AC on. I don't drive with windows down on the highway either due to wind noise, and it lets me better hear if there is a change in engine noise so I know to shut things down myself. When the engine stopped running this time, I had felt uncomfortable enough in the car to turn the AC on to the first fan speed, and precisely about 1 minute later, the engine itself started shutting down.

I know that the ECU's are burned in at the factory and the like at much higher temps at the factory, which is probably why everything works like normal after giving everything a chance to cool off,  but this was definitely something I experienced consistently with my old V88 in the passenger footwell during summer months, and perhaps the G4+ is the same given that it is located in the same position in the car? I can't say what temps the old V88 logged when it used to shut down because I was never logging when it happened, but I do remember that it was at ambient temps similar to what they were when I experienced my issue this time also, and the mechanical setup of the car is identical between the ECUs.

It might explain why my logs look utterly normal prior to the engine shut-down event. 

I'll be installing another set of plugs into the car and attempting to start it a little later today if I have time or tomorrow, so it should be interesting to see if it runs, given that there have been zero mechanical changes (aside from plugs) on the car between when it stopped working and when I fire it up again. I'll keep everyone posted.

 

Link to comment
Share on other sites

My wire in ecu is in standard location 

 

been in there 4 years with no issues but I am right hand drive and exhaust comes down the drivers side and ecu on left side

 

have you checked battery voltage when engine running as software has a buttery over volts shutdown in it 

Goto limits

then System voltage limit 

 

what’s it set at  ? 

 

Mines on a g4 storm is set to 18v

Link to comment
Share on other sites

2 minutes ago, neil brown said:

My wire in ecu is in standard location 

 

been in there 4 years with no issues but I am right hand drive and exhaust comes down the drivers side and ecu on left side

 

have you checked battery voltage when engine running as software has a buttery over volts shutdown in it 

Hi Neil, thanks for your response! You must be watching me because I was just starting to look through the manual of the ECU to find any circumstances under which it would shut the engine down aside from rev limiters and the like! Voltage was around 13.5V when things started going off-line.

Yes, on mine, the exhaust comes down right where the ECU lives. 

In addition, the size of my downpipe is such that it comes within very close proximity to the firewall of the car. It is 4 inches in diameter, if that gives you a hint. 

Battery voltage was in the mid 13's V range when it shut down. I have warnings for over-voltage on my AIM dash display to warn me of anything out of the usual with respect to voltage.

Looking through the log of the event, it really does look as if nothing is wrong, and someone just started turning off off the spark plugs while the car was in motion. I was stumped, but then only now just recalled that in previous summers when I first got my V88 years ago, I had in fact experienced something similar, and I quickly learned to use AC when running around for long periods of time in order to sidestep the issue.

Link to comment
Share on other sites

Could your cdi boxes be shutting down on high battery bolts if dc charge alt isn’t regulating properly 

 

have you tried running car without the dc charge alt connected - unplug the multi plug on it

 

mines is a right drive car in the U.K. so lay out different 

What about the dwell time as for mines on waste spark at 14v is 3.5 ms 

 

could the dwell setting on ecu in ecu making the cdi boxes over heat 

Goto ignition

then click on dwell control

 

whats your dwell times

Link to comment
Share on other sites

3 minutes ago, neil brown said:

Could your cdi boxes be shutting down on high battery bolts if dc charge alt isn’t regulating properly 

 

have you tried running car without the dc charge alt connected - unplug the multi plug on it

 

mines is a right drive car in the U.K. so lay out different 

What about the dwell time as for mines on waste spark at 14v is 3.5 ms 

 

could the dwell setting on ecu in ecu making the cdi boxes over heat 

Good point regarding the CDI boxes. I will check the CDI manual to see if there is anything outlined in there regarding input over-voltage protections.

With that said, the log from the ECU shows that system voltage was as high as 13.9V shortly after I fired the car up at the start of the trip, and was at 13.7V  about 48 mins later when things started shutting down. I would know immediately if my system voltage goes above about 14.8V through my dash, and no warnings came in, confirmed by the log I have.

I'm honestly not sure whether the dwell setting on the ECU could be making them overheat. You make a really good point though, and I'm going to have to look into a bit deeper. Thanks very much for this.  Looking at the dwell table on my car, at 14V it is set for 3.0ms. I'll admit that dwell is an aspect that I have not read up on, so I'm unsure as to what this means in the context of my particular setup. I will look into it though.

Link to comment
Share on other sites

Looking through the manual of my CDI, dwell appears to be set properly at a "non-zero value", and without change with different voltage levels within the ECU tables, so I would *guess* that as long as the value is not zero, the CDI could care less what value it is? I hope someone with more knowledge can chime in here and bless me with some knowledge :)

Link to comment
Share on other sites

2 minutes ago, neil brown said:

Says in notes on software

 

if dwell too low a missfire may occur

 

if dwell too high igniter maybe be damaged if too hot

Hmm... But is that with a CDI ignition system though or with the stock COP inductive setup though? I am curious as to whether I could actually allow my CDI boxes to run cooler by running a smaller dwell that is closer to zero without throwing off the map.... hmm...

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...