SR20 randomly shuts off, CAS?

Hey all, having difficulties diagnosing an S15 with an SR20 running an ECUmaster Classic ecu. Attached is the log file from startup to when the engine randomly shuts off. It seems to do this for no reason, always at idle. Though the customer has complained that its breaking up at certain engine loads, I am not sure that is related. It also only seems to do this when it warms up to 60*C.

I did a full wiring harness last year, was tuned elsewhere and drove fine all last year. Owner parked it for the winter and just pulled it out of storage to find this issue.

I have checked all my work, power is maintained through the engine dying on the ECU, CAS, ignition coils, Injectors, etc. There is no protection strategy operating, unplugging the coolant temp sensor or wideband makes no difference. I cannot see an issue timing wise in the log, the ecu doesnt lose sync or cam sync. All the other active sensors are working and read nicely. I have put an oscilloscope on the primary trigger and cant see anything wrong there, but I dont have a known good example to compare it to either. My scope only allows screen grabs (as far as i can figure out lol) so a screen grab is attached as well.

I dont have a known good CAS to swap to and I'd like to be more sure in my diagnosis before I recommend an aftermarket trigger setup of some kind. I have tried all the go to stuff and am kinda stumped here without changing trigger setups to rule it out. Any suggestions?

hello that screen capture is a pretty messy signal from trigger one and does your scope have a 2nd channel?

here is a typical trigger pattern altho it is a vg30 cas sensor it is similar in looks, you will notice trigger one has 360 even-sized pulses per cam revolution and the trigger two has six, and possibly in your case it will have 4 pulses ranging from small to large,

hello that screen capture is a pretty messy signal from trigger one and does your scope have a 2nd channel?

here is a typical trigger pattern altho it is a vg30 cas sensor it is similar in looks, you will notice trigger one has 360 even-sized pulses per cam revolution and the trigger two has six, and possibly in your case it will have 4 pulses ranging from small to large,

Hey Ross,

Yeah that looks familiar from the last Link I did. Wish I had saved photos of it like that.

The customer is going to bring me a CAS today so hopefully I'll be able to rule that out. If that doesn't work I'll bust out the scope again and try it with both channels and grounded to the block and see if that cleans up the wave.

The ECUmaster Classic doesn't really like the 360 optical CAS, mostly RBs, but I did experience some breakup at higher RPM under load with some SRs so I switched to only installing ECUmaster black now, and it's been better. I am not sure it is related to your problem though.

I also install a fuse and relay so the ECU 12v+ is improved as the classic only have 1. S13 shares their 12V ignition feed with the fuel pump wire, and it create a lot of noise in the wire that feed the ECU. No idea about the S15. But I noticed that your battery voltage had a couple of random drop in your log.

@ Frank @ Elegant Drift Shop - Appreciate the feedback, I hadnt really caught that voltage dip. However it seems random and doesnt correlate to the shutting off. S15's are wired differently thankfully, I've powered the ECU off connector F4 pin 9 which feeds the injectors and turns the ECU on from factory. Fuel pump is fed completely separately. I've attached diagrams if you're curious. I tried running power straight off the battery to the ecu with no change, so that rules that out.

Also good to know regarding the Classic and CAS's. I loath the Nissan CAS as is for the being 30-years-old-unreliable, expensive to replace and inaccurate in the best of cases.

Now i dont even know where to start, much less what to try next. I spent my entire day off on this car and am still without a definitive culprit.

I took Frank's advice and powered the ECU straight to the battery, no change.

I powered the CAS directly from the battery, no change.

I ran new shielded wire from the CAS outside the car to the ECU with the sheild grounded, no change.

I swapped the CAS with another the customer brought me and it seemed to get slightly worse buuut, no change.

I wiped the ecu and put the basemap back on it, no change.

I checked power everywhere (again), was all good, no change.

I locked the timing to 10* and let it run, no change.

I logged everything today and have a lot to share that confuses me, and for simplicity I'm going to break this into a few posts.

For one test it was being held at rpm and it died, usually these logs are at idle. Whats most interesting is the Executed Sparks Count log, usually this is linear and consistent with RPM but it spikes and dips even as RPM is steady. The engine doesn't misbehave at all until a few seconds later when it shuts off. I have logs of it doing this at idle as well. It has also done this and not died so not sure what it means.

The strangest thing that happened to today is in this next log. Around the 15:30 mark the engine dies but the ECU kept the fuel pump running and when I looked back at the laptop it was holding 274 RPM like rock steady. Spark count also stuck rock steady at 12, which is odd given the engine isnt moving anymore and the engine stayed 'in synch' through this. I also chanced upon getting a trigger scope log of this, the forum wont let me upload the file type but I've attached a screenshot. Keep in mind the engine is not moving for this scope.

I am more and more leaning to this being a software problem of how the ECU understands the CAS's signals. Even a clean scope trace from the EMU software doesn't have 360 triggers before a home signal. Theres another rabbit hole here about the trigger setup of the CAS in the EMU software, I've messed with it a lot in the past and it only works with ECUmaster's basemap settings so thats what this car is using.

The kicker is I can't sell any of this to the customer. The car ran fine for like 6 months last year, I've ruled out a ton of stuff but telling him he needs to buy a $600 trigger kit isn't really appealing when the "known good CAS" has the exact same issue. This either makes me look incompetent or look like a liar and if I convince him to buy a trigger kit and it doesn't solve the problem I'm going to eat my hat.

The known good CAS worked on a stock ECU (not sure how long ago) and the trigger wheel appears wiggly compared to his but both have the same issue in his car. Switching him to BLACK is also not that straight forward...

Does the engine dies straight away like you’d turn the ignition off, or does it sputters and slowly dies like it’s running out of fuel? Are you monitoring fuel pressure?

I don’t like to suspect the ECU but did you try to swap it to another Classic just in case?

Another test you can do is put a timing light and check if the ECU is still firing the coils when it dies.

I should also mention that I tried replacing the battery, running it with a charger on, 2 batteries, and removing the alternator power feed all to see if I could figure out where that dip was coming from. None of that changed anything.

Voltage drop test from the battery to the ECU and CAS was like .2 Volts..

I also took off and cleaned all the ground with no change

Try fitting a aem trigger disc, this has far less tooth count. 24-1 these are around $30 and easy to get hold of.

@ Frank - It dies like the ignition is shut off. I have tried killing the fuel pump and its faster then that. If i kill the relay that powers the ignition coils its the same. There is a fuel pressure gauge on the FPR in the engine, and it stays at 40 psi through a shut off.

I have swapped the ECU out for another with no change.

I will have to do that with the timing light again, I'm pretty sure I've tried that but I'll do it again to be sure. I think ill also put the O scope on an injector and see if they stop firing at the same time, I suspect the ignition coil stops firing because the ECU stops telling it to. I know it has 12V and ground the whole time.

Thats a good idea Ross, I think I'll do that.

Okay wow sounds like you got your hands full on this one. There is a big list for what it could be. Before i start with bombarding you with a text wall can you get me specs on the following:

Psi/bar of compression test of all 4 cylinders

Fuel rail pressure (might neef aftermarket gauge for this)

Coil resistence of each coil in ohm.

Was any sound systems installed (y/n)

Are any imobilisers anti theft or programed key present)

Have you star point ground all the things?

I had a very similar issue with my GTO. Pretty sure i can help you out if you want but it is a lot of labour and probing and testing. Just heads up.

Hey Enzo,

Engine runs fine and healthy, I have no reason to believe there is anything wrong with compression.

Fuel pressure is 40psi

Coil resistance I don't know, but they are all brand new. If one was bad it should just drop a cylinder, or pop a fuse.

There is a subwoofer and aftermarket headunit, both of which I installed in the past. They are turned off for these tests and on separate circuits. I had thought they could explain the voltage dip.

I removed the alarm back when I installed this harness in the car.

Yes star point grounding. Battery -> chassis -> engine block

Is your GTO EFI and using an optical trigger? Curious to hear the story there but I am not sure how relevant it will be.

Okay that gives me some info.

For me the first thing i always do is compression check. I dont like guessing. Everything on a car is just specifications. If the specs are right,it works. If not it doesnt work. The trick is finding the one spec that is off.

Doing a compression check dies 2 things for me

1 it ensures there is pressure for combistion. Wet and dry tests are important.

2 taking out the spark plugs tell you a lot about the engine.

Even if it is right. Check it anyway. Tick off system as working. The process you are using is to eliminate systems that are good. Iv made a lot of mistakes in the past thinking something was good then it was actually not.

Right. Fuel at 40 psi is about 2.7 bar. Personally i feel 3.5 to 4 is good. But varies from car to car so check manual on factory pressure. Could be bad ground on fuel pump. It could cause issue you having. Take the ground to the star point to be sure. We will check flow rate later.

Coil resistence you can check with multimeter. You have 2 circuits. Primary and secondary. Primary is the one where it plugs into the harness and secondary is the spark plug. (The number of ohm isnt important, just as long thr ohm reading is more or less the same as the others)

These get pretty hectic to test some times but there are lots of youtube vids. You have an osciloscope (i call em silly scopes) so thats gonna be a big help later on. However if you say they are new they are probably good but test them anyway. Out of box failures are a thing.

Remember the goal here is eliminating systems that work not fixing the problem. What remains is the system which we will dive into. Ill expand on coil tests a bit later when we get to backprobing ECU pins.

Okay so on the the sound system. A very common thing with sound systems is ground loops. This was an issue on my car. i also had the start then dies issue identical to yours. Long story short is when you plop grounds all over the chassis and it creates a loop what happens is one of those actually carry a charge and causes the ECU to go haywire. My coils were firing irratically and 2,5 and 1 ,4 fired at the same time. So what happened was the ground wire actually carried a positive charge because of the loop. Solution is pull wires from all over the car to the point of ground especially fuel pump ecu engine block head and gearbox. Injectors ground internally. I turned the key and the car started like it never had an issue. I believe your problem is here but i could be wrong as these kinds if issue are VERY rare. There is almost no info in ground loops in cars. Usually this happens when audio is installed. So you can ground all infotainment to chassis but then everything else must be off the chassis to the star point and one thick lead that goes from star point to chasiis. If you go battery to chassis to block but the gearbox is grounded to chassis thats a ground loop cause its in contact with engine also.

And to clarify its a 91 Mitsubishi GTO. Not the Pontiac. OBD 2 becamse standaed after 1996 so there is no way to scan codes. Gotta do it the old school way.

In the GTO/3000GT you have bat+ then that goes to am MFI relay that powers up the whole EFI system seperate to the ecu. You said you have S15 right? I dont think they ahve those. But esentially when ignitiom goes on the systems power up. The checks we are doing is to determine which system doesnt power up right. Then we go from there. I can tell you now they probably will start up right but lets hope there is an issue somewhere around there. Get the basics sorted then we know for sure and dont end up chasing our tails.

Also sorry for all the typos. Im dyslexic so i cant see all the typoes sometimes. My eyes fly right over them

Ok so not exactly an update, as I don't have much hard data to contribute but I did get the AEM CAS wheel working and that didn't seem to solve the issue.

I had a phone call with ECUmaster's tech support guy, we have been emailing back and forth to find a solution. I had suspected there to be some sort of issue with the way the software interprets the 360 wheel; it only looks at 60 of the teeth and I'm not sure how it decodes it down from 360...

But anyway this doesn't seem to be the culprit, as with the AEM wheel it was still shutting off, granted a bit less often and still only when CLT was around or above 60C. What we found was that the CAS was falsely triggering, sometimes on the primary trigger, sometimes on the secondary. When this happens the ECU loses the position of the engine and its no longer firing the coils at the correct time, which seems to fit how it shuts off. Not sure whats could be causing this, I have re run wiring outside the car, tried different ECU's and CAS's, and the car ran fine before.

From the phone call I had a few ideas to try to see about trying to mitigate this false triggering, but the car had other ideas. The #4 injector seized open and filled the crankcase, intake manifold, intercooler etc with fuel. So this is on hold until the injectors are replaced/cleaned.

Small update;

Got the injectors back and tossed them in. Some fighting a dead battery later and we're back to where I was in May.

Noticed many of my problems seemed to be at low battery voltages, have to say the Classic does not handle or tolerate voltage fluctuations as well as the Black does. I threw my O scope on the ECU's power input and got funny readings, thinking that voltage dip on the log may have been something. However, I scoped the injectors and CAS (since they are all fed from the same fuse source) and they were clean, so I ran a new power wire to the ECU straight off the battery and it didn't solve the issue (2nd time I've tried that now). Swapped it back and the scope is clean, not sure what if anything changed.

I will probably run a relay dedicated for the ECU power as Frank suggested, just to shorten start up times etc.

The Classic (and assuming the Black) decode the 360 wheel down to 60, but still not sure how. I swapped the AEM disc into a different CAS just to rule out all possibilities and lol it seemed to get worse. Shuts off at any temp now.

I did a timing drift test and while timing stays pretty locked on, every so often it just skips a bunch of spark events. Hard to say how long or how many but I can see the timing light go dark for a few engine cycles and find its way back. If at idle it usually doesn't recover.

So I've ordered a trigger kit from Ignite 5-0 in the hopes that all this is the solve.

Slightly off topic but maybe not, this car has melted 2 turbos and everything around them. Before I wired it and then again after. Could this misfiring or break up as the owner describes be causing extreme EGT's? I do not have logs or anything from before to compare to but I highly suspect this issue has been present longer then I knew and the owner being somewhat naive and unsympathetic mechanically didn't get it sorted until now.

sorry i missed your first reply.

"What we found was that the CAS was falsely triggering, sometimes on the primary trigger, sometimes on the secondary. When this happens the ECU loses the position of the engine and its no longer firing the coils at the correct time, which seems to fit how it shuts off."

sounds like ground loops are your problem. very few people seem to know about these. its a common thing to consider in audio in audio setups. I had the exact same issue on my GTO. ECU was grounded to the chassis on the inside of the cabin. and i had coils 2 and 3 fire at the same time and coil1 doing nothing (its a V6 with 36-1 magnetic pickup). had to run a ground wire all the way to my battery negative and star point all electronics there. then it worked perfectly fine after that. if you dig around on the forum you'll see it there. "weird spark issue" is what i think i called it.

Then i discovered my rotor phasing was off (teh home pulse) i was sitting at 5 or 7 teeth when it was actually tooth 27. i confused which way the engine turns. then after that i got it started to discover breach in harness on (can you guess where) coil 1. and after all that jazz she started up and drove her up the driveway the past weekend for the first time since 2019. it was at 1500rpm but ill take thew win.

my suggestion is ditch the CAS and get a crank trigger on. CAS modules are known for their problems.

also on the EGT yea you need to get an EGT sensor. strange enough i also have that problem with my car EGT it too damn high. gotta get it lower. the AFR is lean and the intermittent misfire you mentioned could be throwing your reading off.

Update: SOLVED

To summarize the last bit, got the 5-0 kit on and running. Quite a nice kit. It has actual instructions which I can't say for the last 2 trigger kits I've installed... Reasonably easy to install, looks cool, and has a very clean signal, and 0 timing drift from idle to 6000 rpm. I'll share trigger settings if anyone asks.

However it did not solve my issue, it did improve it by eliminating the triggering issues but the car was still randomly dying.

More testing I found the alternator was draining the battery and the root cause of the voltage dips we see in the log. Swapped a KA24DE alternator in and cleaned that up, however it also did not solve the problem.

By this point the issue is hard to replicate, was taking around 1 hour to get the car to misbehave. Using a timing light I watched as the ignition coils would stop firing when the engine would die. I now knew this was not a triggering issue as the scope trace was 100% clean. Coils maintained power and ground through it shutting off. All 4 coils were stopping firing, so that only left 1 common point. I am not sure if I've mentioned it before but this car is running 4 pin VW "R8" coils that have a sensor ground return, which I used.

The sub harness had a bad crimp on the pin for the sensor ground, and after some time warming up and rather inconsistently it would micro arc and no longer be making a connection and all 4 coils would stop firing at the same time. I usually do not crimp a 22 awg into a size 16 DT terminal, that is not enough wire to ensure a good crimp. The other 4 in the connector were done correctly and yet I somehow missed this 1 crimp. Fixed that and all the issues have gone away. The other symptom that I was not sure was related until it disappeared was an intermittent breakout under boost/load.

Mixed feelings finding it, I'm glad it's resolved but mad at my self for making the error and how much time it took to get here. Lesson learned. I usually find that the most complex issues have the simplest explanations and this was no exception.

Oh and because this car hates me, as soon as we pulled it in for the night the diff started leaking gear oil. I'll just never be rid of this car lol.