Messed up firing order, random excessive knock, vibration?

[_Madman_]

Hi!

I finally have my car going again, but after a long while it behaves a little weird.

Basically the car seems to vibrate as if one of the cylinders is not working correctly, but it goes and goes pretty fast and with no problems except this vibration.

Recent mods involved all sorts of intake piping, Apexi intake, fidanza lightweight flywheel and swapped ECU (96 JDM -> 99 AWD USDM).

To change the ECU I swapped injector pins and spark wires at the coil like this: http://www.dsmtuners.com/forums/tun...07561-new-ecuflash-1-38-2g.html#post151631922

References:
pinouts: http://www.dsmtuners.com/forums/articles-tuning-ecu/281763-2g-turbo-ecu-pinout-w-wire-colors.html
swaps: swapplugwires [ECMTuning - wiki]

Does anyone know how to diagnose if I have messed something up or not? I have no idea how to verify if injectors are doing everything in correct order, and I have no idea if spark plugs are firing in correct order.

Also the EvoScan goes mad with some random 20ish knock counts somewhere. It's all quiet in 0-5 all around and then 23 in one single spot.

Most knocking occurs in 1st gear, ~0.6 bar boost. Not at 3rd gear pulls ~1.1 bar.

I can do actuator test with EvoScan, but I don't know how to proceed.

I hope it's correct forum to ask this question.

I can gather more information including logs if that can help, it's just that I don't know where to start and how to verify injectors and spark plugs when engine is turning at 700 rpms

 

[sb7775]

Do you have Ecmlink and if so post a log.

Steveb

 

[_Madman_]

No, sorry, I only have EvoScan logs. They are in csv. Or I can post a graph. Which one do you prefer and which parameters are needed?

 

[the_mork]

When does the vibration happen? At idle?

Is the high knock count repeatable? What RPM and load range?


I thought my car was really running like crap for a while and it would shake and vibrate at idle. I ended up solving the problem in 10 minutes be replacing the ISC with one off of an RVR.

 

[_Madman_]

Here are the "RPM/Load vs Knock(Max)" charts from 4 logs, each of them is around 7 to 30 minutes of logging.

 

[the_mork]

The green cells in your 3rd graph look like phantom knock. Minimal load at low rpms doesn't make a lot of sense to me when it is such a high number, I'd be more concerned if you numbers made more sense but those are just bonkers.

You first graph looks exactly like the graph that I get with some partial throttle knock on an evo8 ecu. I'm not sure if this is phantom knock or not, I've heard opinions both ways and I'm not sure what to do about it. Mine was helped by turning off the cold start and egr ignition advance options but you don't have that luxury. Despite changing those options mine still persists.

Looking at your second and fourth graphs if we put them together a little bit we can see that you are probably fine once you hit the 6k rpm range and before the 4k rpm range. I'm not sure what you are using for timing control (are you? I'll double check your post and profile in a second). If you can control timing then take a look at your peak timing advance and your base timing advance for the load range that is giving you trouble. Keep the base and the peak the same as they are now and interpolate the column between them and see if that reduces timing in your trouble range. After that start bumping up your peak timing advance until you start to see knock again (interpolating the rest of the column as you go) then back off one degree. Obviously smooth your graphs horizontally a bit so that you don't get drastic swings while you're testing.

I think you can look at what is happening in your first graph and the rest of your graphs as different issues.

As for getting the injectors to fire in the right order use evoscan to fire each injector and make sure that it is mated up to the right cylinder.

 

[_Madman_]

Thank you for insightful reply, unfortunately I'm still a beginner so I'm still lost.

Concerning timing, all of it is done with the stock ECU, it's a 2G car, so as far as I know I cannot do any adjustment with the internals I have right now.

I did more logging today and here are some more graphs that could give more information about what's happening during excessive knocking.

Also, I tried to measure what happens when I activate injector actuators in EvoScan, but nothing seemed to change so I'm a little bit lost.

From what I gathered I must verify the resistance between 2 pins on injectors, they should be more or less equal on all 4.

Then I have to verify that pins on the timing belt side of each plug is +12v versus ground.

If the actuator is on, this pin should read +0v, right?

But what is the second pin? And what should it read?

I guess I'm asking stupid questions, but I'm new to this data logging stuff and I don't know where to begin. Is there any "datalogging for dummies" somewhere I could study?

Also the ECU throws out following codes, but I guess that they are OK.

Updated: 18:11:12
7 Diagnostic Trouble Codes Found:
----------------------------------------------
P0135 Heated Oxygen Sensor Heater circuit Bank 1 Sensor 1.
P0141 Heated Oxygen Sensor Heater circuit Bank 1 Sensor 2.
P1400 Manifold Differential Pressure Sensor circuit.
P0443 EVAP System Purge Control Valve circuit.
P0446 EVAP System Vent Control.
P0403 EGR Solenoid.
P1104 Turbocharger Wastegate Solenoid.

(P0135 & P0141) - because Japanese 2Gs don't have O2 heaters.
(P1400 & P0443 & P0446 & P0403) - because Japanese 2Gs don't have emission equipment.
(P1104) - because I have MBC and the stock solenoid is removed.

 

[the_mork]

I'm not sure what to tell you then, I'm certainly no expert but I replied with what I've experienced.

The only think I have to add is that it is much easier IMO (though less accurate) to listen for injector clicking.

You could pull all the injector plugs except the one that you are trying to verify and have a friend push the button to activate that injector (with the car off but ignition on). If you hear a click then you're in the right place. Try it with all of them connected so that you know what you are listening for first.

 

[_Madman_]

Ok, I measured the resistance on injectors and it's 3.8, 4, 4, 4 ohms on cylinders 1, 2, 3, 4 respectively. Are these values ok? I think I saw 2-3 ohms somewhere.

Also, could it be that knocking occurs because boost pressure is rising too quickly at partial throttle? Because I have MBC and it's possible that the boost is too high for stock ECU maps in the troubled range

I will try to turn the MBC down completely to verify this, but that's tomorrow.

 

[_Madman_]

Ok, more updates. I verified impedance on another 2Gb DSM today and it turns out it also had 3.8/4 ohms.

But then the interesting part...

Since the car was behaving so badly, we tested if it would work better with the old ECU. After the swap the engine started and there was no vibration at 2500 RPMs when I pushed the throttle.

Before the swap we switched the injector wires back as they were before I swapped the ECUs. Also the plug wires were swapped.

After this experiment I put back the 98/99 ECU, but I only swapped the spark wires. I left the injector wires as they were with 96 ECU.

The car started to behave better, the vibration was still present around 2500-3000 RPMs, but it was less noticeable, the logger also logged only 2-3 counts of knock, and in smaller area during some 15 min testing city+highway.

Still, the car was vibrating at the troubled range, so I decided to try what the old ECU does. I quickly plugged it in without bolting, swapped the wires. And voila! No vibration and gosh, suddenly the power went up tremendously. I pushed the throttle to metal in 2nd gear and the tires just spun all the way to the redline

So I guess there is something terribly wrong with the rewire I did for 99 ECU.

The strangest thing though is that I can swap plugs on one coil pack around and the car works as expected, if I do the swap between coil packs I can't start the engine. Does that mean we always have a combustion at 2 cylinders at the same time? EDIT: ( Yes -> http://www.dsmtuners.com/forums/newbie-forum/329520-97-ecu-cas-inversion-question.html#post151796434 )

And does anyone know what I did wrong when I tried to compensate the CAS difference for the 99 ECU? You can see the wrong way in the picture I linked above.

I really want the 99 ECU because of it's data logging abilities

EDIT: is the swap injector procedure mentioned here correct for 96->99 swap, or should it be done the other way around? ( RRE Instructions )