Just another misfire/rough idle thread...

[StkmltS]

Yes, that's pretty normal. Check your swaybar endlinks, even when I think there is no way that could be the cause, it always is.

 

In other news,

 

I get misfires in two conditions 1. cold startup (I will get 1-2 across some random cylinders) 2. On shifting, clutch in misfire will appear (briefly).

 

1. Almost seems like intake manifold gaskets, but it's not that cold (55 degrees) and I can see orange tabs

2. Seems like Vacuum issue (which i think is supported by my LV)

I'd pay good money to only have misses under those circumstance!

 

I wouldn't worry too much about the misses bergen shifts. It may have something to do with the switch from open loop to closed loop?

 

Sent from inner space.

[hadvw]

That's about what I was experiencing as well with the 06. It never really misfires right at cold start and idle. I had to drive it just a bit. And then, when I got to THAT traffic light with the car not fully warmed up yet, misfire count was showing up at idle.

Even with the engine fixed now, every time I get to THAT traffic light, I am always scared of misfires .

 

That's EXACTLY what was happening to me. Traced it down to a non-OEM coilpack (not sure how it got there) that died early. New OEM coil-pack, and it seems much better. Not 100% sure yet that it's fixed, but about 90% sure: 3 days in a row, including cold/wet, and no issues.

 

StkmltS: have you tried swapping coil packs to try to isolate it? I thought it wouldn't do much, but so far, it seems to have helped me isolate the problem and solve it.

[StkmltS]

It started missing again this morning at around 115°. Playing with the throttle at <115° did not induce misfiring.

 

That's EXACTLY what was happening to me. Traced it down to a non-OEM coilpack (not sure how it got there) that died early. New OEM coil-pack, and it seems much better. Not 100% sure yet that it's fixed, but about 90% sure: 3 days in a row, including cold/wet, and no issues.

 

StkmltS: have you tried swapping coil packs to try to isolate it? I thought it wouldn't do much, but so far, it seems to have helped me isolate the problem and solve it.

 

Indeed I have. I jumped on a good deal on 4 new non-OEM coil packs a while ago (please save the scoldings until the end), so those may go on sometime soon just for the love of moving things around and not sitting still. I also swapped injectors, swapped plugs, replaced the IM o-rings, installed a new turbo inlet, installed new vacuum lines, smoke tested, compression tested, leak-down tested, sanity tested (failed), had an oil sample analyzed, did a few other things, aaaaaaand completely rebuilt the motor.

Edited November 1, 2016 by StkmltS

[xt2005bonbon]

How many miles ago did you swap coils, and injectors?

[StkmltS]

How many miles ago did you swap coils, and injectors?

 

Injectors and coil packs were first swapped in early January, 2015 @ approximately 98,162 miles. So about 12k miles ago.

[xt2005bonbon]

ok.

 

how many miles since the rebuild now? If you have added quite a lot by now, I would have a look at the spark plug and compare with #4.

[StkmltS]

I have about 1,300 miles on the rebuild.

Maybe I'll pull the plugs at my next oil change in ~200 miles.

Edited November 1, 2016 by StkmltS

[xt2005bonbon]

+1.

[StkmltS]

How do I find the HEX address to monitor Fuel Pump Duty?

I know the (other?) address is 3ad04, but to monitor it in BtSsm I need the hex address.

ROM: A2WC522N (based on A2WC521N (based on 32BITBASE)).

 

At this point I'm more interested in understanding how to find it, not just finding it.

[Perscitus]

How do I find the HEX address to monitor Fuel Pump Duty?

I know the (other?) address is 3ad04, but to monitor it in BtSsm I need the hex address.

ROM: A2WC522N (based on A2WC521N (based on 32BITBASE)).

 

At this point I'm more interested in understanding how to find it, not just finding it.

Use the standard SSM param for Fuel Pump DC (%). Should work for most ECU/CALID combos irrespective of Model/ModelYear.

 

For pre-defined standard, extended and engineering params - always look to the latest RR logger definition XML file.

Search it for the params you're interested in (standard params - generic - same address for most of our cars; extended/engineering params - memory addresses are ECUID specific).

 

<parameter id="P47" name="Fuel Pump Duty" desc="P47" ecubyteindex="13" ecubit="0" target="1">

<address>0x00003B</address>

<conversions>

<conversion units="%" expr="x*100/255" format="0.00" />

</conversions>

Remember that Fuel Pump DC in our ROMs is typically set to 2-3 target values so when logging you'll typically see associated step values.

Say 0%, 33%, 66%, 100% or 0%, 50%, 75%, 100%.

 

Might as well log Injector PW or DC as well to see if you have any fuel delivery issues. Fuel pump DC alone might not be very revealing.

Edited November 3, 2016 by Perscitus

[StkmltS]

Use the standard SSM param for Fuel Pump DC (%). Should work for most ECU/CALID combos irrespective of Model/ModelYear.

 

For pre-defined standard, extended and engineering params - always look to the latest RR logger definition XML file.

Search it for the params you're interested in (standard params - generic - same address for most of our cars; extended/engineering params - memory addresses are ECUID specific).

 

Remember that Fuel Pump DC in our ROMs is typically set to 2-3 target values so when logging you'll typically see associated step values.

Say 0%, 33%, 66%, 100% or 0%, 50%, 75%, 100%.

 

Might as well log Injector PW or DC as well to see if you have any fuel delivery issues. Fuel pump DC alone might not be very revealing.

Thank-you.

I was only looking in the A2WC521N.xml and ecu_defs.xml files.

The more I learn about this stuff, the more interesting it gets.

 

Worth checking out: Fuel Pressure Explained

Edited November 3, 2016 by StkmltS

[StkmltS]

Im going to add a little fuel at idle and see if that helps or makes it worse. My thinking is: if it gets worse then Im probably dealing with a sparking problem; if it gets better I may have a fueling issue. Now to figure out the correct table to modify... Hopefully Ill have some free time this weekend to play around on the computer and figure it out.

[xt2005bonbon]

what are your current long term fuel trims?

 

fyi, I am constantly viewing injector dc and I am seeing about 1% at warm idle, sometimes going to 2%.

[StkmltS]

All four trims are well within +/- 5%. The A trim fluctuates alot, but it always floats somewhere between -1 and -4. I've rescaled my MAF once since the rebuild, but even before that all four trims were within +/- 5%.

 

I'm wanting to decrease the target AFR (add fuel) only at idle just to see what happens. Maybe I could change a few cells in the MAF scaling table to accomplish this, but that seems like the wrong way to do the right thing.

 

Plus, my brain has recently shifted into "tuning mode" and I'm eager to start really digging into how to tune properly. Don't get too scared, right now I'm more interested in the how and why of tuning, and not so much into trying to add power. That will come eventually, probably, but I want a good foundation of solid understanding first. Right now my misfire problem is a major motivator for me to learn more about tuning my car. I really believe that understsnding how the ECU tells the motor what to do is going to be a key part of figuring out what the root cause its.

 

I'm starting to ramble on, I'll stop now.

 

Sent from inner space.

[xt2005bonbon]

so if the A trim is currently trending a bit negative, and if you plan to add fuel, it will then become more negative right?

 

On a side note, it sucks that we don't have afr sensors for each cylinder, or at the very least for each bank..

[StkmltS]

so if the A trim is currently trending a bit negative, and if you plan to add fuel, it will then become more negative right?

 

On a side note, it sucks that we don't have afr sensors for each cylinder, or at the very least for each bank..

Yes, sort of, i think. Rescaling the MAF table to add fuel (trying to "trick" the ECU into adding more fuel) would just make the trim go more negative. I need to change the target AFR. Then I just sit back and let the computer do all of the short-term fuel corrections it needs to do (also affecting the long-term 'A' learned value) so that I hit my new, lower AFR.

 

I could be 100% wrong, but I'm 93% sure that I'm 99% correct.

So with a personal target CWR (correctness-wrongness-ratio) of 1:0, my short-term certainty is -7%, and my long-term learned correctness is -7.93%. Do the math, I'm pretty sure it checks out.

 

Sent from inner space.

[rhino6303]

 

I'm wanting to decrease the target AFR (add fuel) only at idle just to see what happens. Maybe I could change a few cells in the MAF scaling table to accomplish this, but that seems like the wrong way to do the right thing to do.

 

Target afr is really only changing your OL fueling. You can do this and force your car to always be in open loop. If you want to change cl fueling, you'll want to change the values in the cl fueling compensation tables. The stock values are fairly odd.

 

Keep in mind that zeroing them out won't work correctly (there's a thread on romraider about it). They have to stay slightly negative or positive. The table is in points additive and not percent. So a -0.5 would change cl target from14.7 to 14.2.

 

I've edited my table so that the majority of cruise is -0.001 with no ill effects.

Edited November 5, 2016 by rhino6303

[StkmltS]

I think you're right. There's an obvious table for OL AFR, and only compensation tables for the CL AFR.

 

I'll look into the points/percent that you mentioned. I'll probably have some time late tonight.

 

Sent from inner space.

[rhino6303]

I think you're right. There's an obvious table for OL AFR, and only compensation tables for the CL AFR.

 

I'll look into the points/percent that you mentioned. I'll probably have some time late tonight.

 

Sent from inner space.

It is points additive and not percent. I edited my previous post for clarity.

 

You might also try zeroing out your fuel injector compensations.

[StkmltS]

It looks like I may be finally getting a little closer to figuring this thing out.

 

Image #1, "Prior to AFR Change, Prior to reflashing"

I let the motor warm up to about 180°F before I took this ~18 second log @ idle.

For tonight's test this is my baseline log.

 

Image #2, "Post AFR Change 1, Immediately after reflashing":

I changed a few values (to lower the AFR) in the 800rpm row of the CL Fueling Target Compensation (Load) table and it successfully reduced the number of observed misses! This log was taken within the first minute of the motor running post-reflash. Usually after resetting the ECU my idle is absolutely horrible and it takes a workday's worth of miserable driving before the ECU learns the correct fueling well enough to keep the motor from stalling whenever the RPM drops down to idle. After this reflash the idle was much more tolerable and I didn't have to play with the throttle up to keep it from stalling immediately after start-up.

 

Image #3, "Post AFR Change 2, Immediately after reflashing":

I changed those same values even lower and the misses slowed down again. Could this really be happening? This log was taken within the first minute of the motor running post-reflash #2.

 

Image #4, "Post AFR Change 2, After a few miles of driving":

I needed gas in a bad way so I drove a few miles on the second AFR tune. Believe it or not, the misses improved even more after the ECU had some time to do a little fuel learning.

 

My idle RPM fluctuates between about 750-850 (800 target), so tomorrow I'm going to change the values in the 900rpm row (CL Fueling Target Compensation (Load) table) to match the changed values in the 800rpm row to see if the misses improve even more.

 

Details worth noting:

The axes for the first three plots are all set to the same min/max values so that the charts are easier to visually compare. The logs are all about 18 seconds long so there's hardly any data missing in the X direction. The fourth chart has a longer X-axis because I wanted to see more data. The first three logs took place within a window of about 10 minutes, and the fourth log was about 20 minutes after that.

 

Conclusion:

What should I do with this information? Answer: sleep on it.

Edited November 6, 2016 by StkmltS
Spilling

[rhino6303]

It looks like I may be finally getting a little closer to figuring this thing out.

 

Image #1, "Prior to AFR Change, Prior to reflashing"

I let the motor warm up to about 180°F before I took this ~18 second log @ idle.

For tonight's test this is my baseline log.

 

Image #2, "Post AFR Change 1, Immediately after reflashing":

I changed a few values (to lower the AFR) in the 800rpm row of the CL Fueling Target Compensation (Load) table and it successfully reduced the number of observed misses! This log was taken within the first minute of the motor running post-reflash. Usually after resetting the ECU my idle is absolutely horrible and it takes a workday's worth of miserable driving before the ECU learns the correct fueling well enough to keep the motor from stalling whenever the RPM drops down to idle. After this reflash the idle was much more tolerable and I didn't have to play with the throttle up to keep it from stalling immediately after start-up.

 

Image #3, "Post AFR Change 2, Immediately after reflashing":

I changed those same values even lower and the misses slowed down again. Could this really be happening? This log was taken within the first minute of the motor running post-reflash #2.

 

Image #4, "Post AFR Change 2, After a few miles of driving":

I needed gas in a bad way so I drove a few miles on the second AFR tune. Believe it or not, the misses improved even more after the ECU had some time to do a little fuel learning.

 

My idle RPM fluctuates between about 750-850 (800 target), so tomorrow I'm going to change the values in the 900rpm row (CL Fueling Target Compensation (Load) table) to match the changed values in the 800rpm row to see if the misses improve even more.

 

Details worth noting:

The axes for the first three plots are all set to the same min/max values so that the charts are easier to visually compare. The logs are all about 18 seconds long so there's hardly any data missing in the X direction. The fourth chart has a longer X-axis because I wanted to see more data. The first three logs took place within a window of about 10 minutes, and the fourth log was about 20 minutes after that.

 

Conclusion:

What should I do with this information? Answer: sleep on it.

I just saw in another thread you deleted your tgvs. You can clean up idle issues by copying values from the japanese model tables as they are tuned for tgv deletes.

[fahr_side]

I just saw in another thread you deleted your tgvs. You can clean up idle issues by copying values from the japanese model tables as they are tuned for tgv deletes.

 

 

 

Not necessary. Just copy the values from the non-cruise table to the cruise table.

[hero2zero]

A few weeks ago, I got a P0302 Cylinder #2 misfire code. The misfire only ever happens at idle. I checked the spark plug which looked pretty good. I swapped coils with cylinder #1. Still got the code at idle after I cleared it. Then, I swapped the fuel injector with cylinder #4. The misfire followed (P0304).

 

So, I removed all the injectors, labeled them, and sent them off to be serviced. The tech told me that there were no problems with the injectors ("they all are flowing fine..."). I re-installed the injectors in the same locations, but today I got a P0304--the same location they were installed before I sent them off. Is it possible that the injector is bad and they missed it during testing? Seems like an obvious answer, but has anyone ever had this happen to them? I used these guys: http://accurateis.com/store/index.php

[rhino6303]

Not necessary. Just copy the values from the non-cruise table to the cruise table.

We don't have cruise/non-cruise on a 05 LGT.

[fahr_side]

Ah, those just have the TGVs closed for cold start eh.