SpeedBump

Hey Guys,

I've been trying to debug a misfire that turned into no spark at all and I'm at my wits end. This has ceased to be fun, and I've eaten too many weekends with this ****.

General setup:
- '90 Miata w/ ~204k miles, stock internals
- Toyota ITBs
- Toyota COPs still run in wasted spark, no 10,000 uF cap in place
- DIYPNP which was running well about 8 weeks ago

The current situation is that I seem to not have spark out of the ECU. I pulled the MS today, and wired in some LEDs and resistors to the spark and injector outputs to see if I could get some blinking action on the bench. As far as I know, I'm using the output test function correctly, as I was able to get blinking LEDs on both injector channels. However, I'm not getting anything on the ignition channels.

I understand that the injectors run at 12V and the coils run at 5V, and I have chosen resistors appropriately. I also understand that the MS provides the positive voltage signal to the coils. In other words, I don't think I've wired something incorrectly, but I'm certainly open to the idea that I did something retarded.

As for the output test function, here's what I'm doing:
- enter 12345 into the appropriate field
- press TAB
- press the "burn" button
- select coil test from the drop-down menu
- select coil A from the drop-down menu (I've tried coil B through D with the same lack of result)
- set test interval and dwell to 20 ms (I've played with many values here)
- set number of injections to 1000 (not needed with coil testing, right?)
- press TAB
- have a desire to throw MS as the Ign. LED doesn't light

Is this right, or am I doing something stupid?

For giggles, and switched the spark output to non-inverted, and suddenly I got a light. It still wouldn't blink, but I'm less suspicious that my uS module is fucked.

Does anybody have any thoughts on what the hell is going on? I don't want to buy another uS module to test, but I'm not sure what else to do.

Here's some background, in case it's relevant:

I had a light but persistent misfire under light load that would clear up with more load. I could floor the throttle, and everything was fine. I have another set of COPs, and spent an afternoon switching coils around convinced myself that the coils weren't the issue. This makes sense, as I would expect a coil problem to manifest itself in the exact opposite way of what I was seeing - I would expect the misfire to show up with high load. In any case, I read online that a leaking injector may cause this, so I sent my injectors out for cleaning. Sure enough, the test report came back that one was leaking and had a a crappy pattern.

After reinstalling the injectors, the car wouldn't start at all. It would almost catch, but wouldn't quite do it. It was acting like it wasn't getting enough fuel. I increased the req' fuel to quickly see if I could get it started, and sure enough, it started but was only running on 2 cylinders. It was obvious by looking a the spark plugs that the 1st and 3rd coils weren't firing. The plugs were obviously fouled. I didn't touch the car for a while, and the next time I tried to get it started, it wouldn't even try to catch.

I reloaded firmware a few times in desperation, and still nothing. I didn't touch the car for a while again, and today I finally pulled the MS to bench test as I described above.

I've soldiered my fair share of shenanigans with the MS, but this one is kicking my ***...

Thanks in advance!

Braineack

iTrader: (62)

doing what you are doing is begging for issues to spring from your diagnosis. It's pretty easy to listen for spark/fuel without having to solder **** into the MS.

did you even first do a composite log to make sure the MS was still getting a correct signal back from the CAS?

hector

I dont have an answer for you but Im pretty sure the coils run at 12v. They may be controlled by a 5v logic source but thats not what drives them, its what guides them.

And I have a diypnp also which I ran spark test on. Did it in the car though. Why did you take it out to do a bench test? It would be way more accurate on the car.

Have you checked your wiring in the car to the coils?

ChrisSTR

Old fashioned spark test. Remove plugs from car, use wire to ground them, turn over engine and look for spark.

SpeedBump

Thanks for the replies.

It didn't occur to me to take a composite logger file. I was getting the "Ready", "Sync", and "Cranking" lights on my TS dash while cranking throughout this process. I was initially suspicious of the CAS as well, but figured it was fine if I was getting those lights. I suppose I could have an intermittent problem in the time span between the TS serial data polls? In any case, I took a composite log this afternoon, and it seems fine. It's attached, along with my current .msq.

I wouldn't have guessed that you could hear the spark plugs fire using the output test in the car - that's good to know. Since I was working alone, a traditional spark test was difficult. The best in-situ test I could come up with was to wire an LED/resistor into a coil connector so I could see it while cranking. I didn't get any blinking in any of the four connectors when I did this, so I pulled the MS for a bench test.

In any case, am I using the output test correctly?

My wife is home today, so I'll have her crank the car while I look for spark. I'll try the output test in the car, as well.

Thanks again.

hector

I haven't done it in quite some time but you don't need to crank the engine over to do a spark test. IIRC, go to advanced in TS and output test mode and spark test, take out the plugs, do the test and listen/look for spark. Same for injectors but I wouldn't take those out, just listen.

SpeedBump

Since I reinstalled the MS in the car to grab a composite log, I tried the tests that were suggested above. I didn't get spark with either a traditional spark test or via the output test mode. I'm still not sure I'm using the output test mode correctly, though.

Can anyone confirm that I'm using the output test function correctly? I described what I was doing in the original post.

About ready to part it and buy a motorcycle...

hector

I dont know what screen you are looking at or what firmware you are using. I said advanced before but its under Canbus/testmode.

From the help section from TS for spark test mode:

First you enter fuel pump on or off
You would then enter milliseconds between spark
Coil test mode set to one
Coil to test
Dwell

Hit start and you are in test mode.

Note if your coils are powered via the fuel pump relay, set fuel pump to on.

Hope that helps.

BTW, have you checked to see if you have 12v to your coils yet?

SpeedBump

Thanks for the reply. I think what your describing is only for MS3? I don't really know, as the TS reference manual for MS2 suggest the same thing. See page 162 in this link:
http://www.msextra.com/doc/pdf/Megas...erence-3.3.pdf

My output test mode is under the "Advanced" tab and the screen looks completely different. There isn't a enable or disable button, just an elaborate sequence to make sure you don't do something stupid.

I do have 12V at the coils, by the way.

hector

I have a diypnp like you so we both run ms2/e. Im on firmware 3.3.2, the latest I believe. If you're not on it, get it. Make sure to make a back up of your current tune so that you can load it. Updating the firmware makes changes to too many things you wouldn't think of.

Ok, at least you have 12v at the coils.

If all else fails with the output test, have the wife crank it over and check for spark.

I saw the composite log but it was in excel so I dont know what to make of it. I only kinda understand them in megalogviewer.

Braineack

iTrader: (62)

he appears to still be syncing just fine. You'd normally want to open this within TS, but you can clearly see that as he's getting the sync flag.

Basically the PriLevel and SecLevel flags are the readings from the crank and cam sensors respectively.

if the MS wasn't syncing, there would be no spark.

Now the case is either:

1. the ignition outputs on the microsquirt module somehow failed (very unlikely).
2. the wiring between the module and coils failed.
3. the coils failed.
4. the firmware corrupted / software changed.


I'd hope he'd have 12v at the coils, because if he didn't, then he wouldn't have 12v at anything behind the METER and ENGINE fuse as well--basically anything powered on the white/red 12v, including: the CAS, fuel pump relay, injectors, and ECU.

hector

The reason for verifying 12V to the coils is number 2 on your list. Since he has Toyota coils, I ASSume he had to do some kind of wiring from the factory harness to the new coils that maybe went kaput.

Just trying to cover the basics. Which begs the question: is there injector pulse? Does the fuel pump prime?

Suggestion 4 is worth a look as he mentioned that he did reload the firmware several times. I know every time I reload firmware I lose all my settings, except the MAT and CLT calibration, go figure!

Maybe he has overlooked something. I know I goofed up a couple of times and selected the wrong module, which in the diypnp case is, microsquirt module or microsquirt non-cased, or something to that.

Oh and thanks for the comp log lesson. A little more knowledge, a little less dangerous....

SpeedBump

Well, I put my big girl panties on today, and got determined to fix this.

I updated to the new firmware and updated TS, and now I understand the output test mode. This is way more intuitive than the old way... In any case, this fixed the no-spark-at-all issue. Not sure why reloading the firmware I was using several times didn't fix the issue, but whatever. The new FM did the trick.

So, with the new firmware loaded, I was back to the original issue of running on 2 cylinders. For ****'s and giggles, I bought a brand new set of plugs in the very off chance that I was doing something stupid or wasn't cleaning the used plugs well enough to get a reliable spark. I verified that each coil/plug/connector was okay using the output test mode, and everything was perfect!

I also looked at the report sheet that came back with the injectors. They seem to be flowing 10% more fuel than they were, so I backed the req. fuel off 10% figuring that all of my fuel settings were ~10% high to make up for the low flowing injectors... I tried and tried, and worked the req. fuel back up to it's original setting, and couldn't get it to turn over. It really wanted to, though.

I checked the plugs, and sure enough, number one was fouled. Number two, looked brand ******* new. Number 3 was fouled. Number four was also clean... Since I was using used plugs previously, I misinterpreted what I was seeing earlier. I thought I wasn't getting spark to cyls. 1 & 3, when in fact, I wasn't getting fuel to 2 & 4!

Using the output test, I confirmed the problem 2 & 4 injectors. I moved them around, and the problem followed the injectors. The ECU is fine, the ignition system is fine, and the harness is fine - Somehow, something got f'ed between the test bench and me installing them in the car. I checked the coil resistances, and they all check out at 14.7 ohms. I just happen to put the bum injectors in cyl. 2 & 4, which are on the same output.

Sorry for the novella. I'm beyond excited. Now I have to talk with RC Engineering and see what's up...

Thanks again for the help, everyone!

hector

The panties, were they black, maybe a thong...... Uh, nevermind.

So you did injector test with bad injectors and they did not "click" or they did "click" but not inject fuel?

SpeedBump

One didn't click, and the other was the saddest little injector you've ever heard. I almost didn't catch that it was clicking over the ambient noise of my neighborhood. I held it against my ear to confirm...

Since they weren't clicking, and the plugs in the corresponding cylinders were clean and dry, there is no way they were injecting fuel. I dunno. I'm really curious to see what RC says is the root cause, though.