Failed MAF = Intermittent Stumble

[Cringer]

I thought I read this on the forum long ago, but now I cannot find the thread...when you fail the MAF (not sure on E38 only, or all Gen 4's), the engine will randomly stumble or hiccup while steady state cruising. As I recall this is "normal" and there is nothing that can be done about it, and it won't hurt the car.

Still, I thought I would reach out about this. I recorded this log the other day. There were two stumbles 4 seconds apart. What appears to be happening is that the calculated cylinder airmass spikes which then causes the injector pulse width to spike.

My only guess this that the zero g/s being reported by the MAF is throwing off the Engine > Airflow > Dynamic > Prediction Coefficients (or at least the formula using the prediction coefficients and the MAF input). Is there a work around to this?

stumble.jpg


E38 MAF Failed Stumble.hpl

[GHuggins]

No workaround without doing a 2bar patch and some have said it did it on a few even then.

As far as no damage. Depends on how big of injectors and how high of fuel pressure. There was one guy several years ago that had this happen and supposedly damaged his motor. Don't know. Didn't see it myself, but his if memory serves did it while wot in boost and bent a rod or rods from fluid and boost in the cylinders???

Gen V's actually do this too. On the gen V's it's directly tied to 0ing out the prediction coefficients as I discovered on them. They actually will do it more often when it happens then not do it at all for a while which is why it was a little easier to test for cause and solution on them. Not sure on the gen 4's. More than likely a multiplier or something coming into play at a larger level when the problem occurs on the gen 4's.

Verlon would be the man to talk to on this. He's dug through code on the gen 4's more than anyone on here and might know what to kill or correct? I might even ask him myself when I email him next time.

[Cringer]

No workaround without doing a 2bar patch and some have said it did it on a few even then.

As far as no damage. Depends on how big of injectors and how high of fuel pressure. There was one guy several years ago that had this happen and supposedly damaged his motor. Don't know. Didn't see it myself, but his if memory serves did it while wot in boost and bent a rod or rods from fluid and boost in the cylinders???

Gen V's actually do this too. On the gen V's it's directly tied to 0ing out the prediction coefficients as I discovered on them. They actually will do it more often when it happens then not do it at all for a while which is why it was a little easier to test for cause and solution on them. Not sure on the gen 4's. More than likely a multiplier or something coming into play at a larger level when the problem occurs on the gen 4's.

Verlon would be the man to talk to on this. He's dug through code on the gen 4's more than anyone on here and might know what to kill or correct? I might even ask him myself when I email him next time.

OK interesting, I appreciate the reply!

I did apply the 2 bar OS and so far have not had any issues with this yet. I was also reading this mega-thread and the consensus was to zero them all out. However, I guess that only applies when the MAF is not failed though.

[GHuggins]

0ing them out doesn't change anything on the gen 4's. May be tied to something, but I've tried with them populated and 0ed and it made no difference on them for this issue.

On the gen V's however, it's almost like it bogs down the processing of the ecu and that is what causes the dynamic airflow to go crazy. It's like they can't process things correctly when they are 0ed out.

I still think it's a comprehensive or rationality check taking place retesting the MAF. Verlon has taught me many things just in the last couple of days with how things are tied together via test and rationality checks on the gen V's. I did not know there were so many things that had to be disabled or changed just for one component removal. I would be willing to bet that failing the MAF isn't all that's required to take it out of the equation, at least not completely for every drive cycle causing this "test/check/glitch" to take place.... Just like post O2 test take place all the way to 5000 rpms, but pre cat O2 test only take place up to 3200ish or something like that rpms or the fact that they'll use post O2's to verify the operation of the pre cat O2's. Seems like all of this should be the opposite to me.

[Cringer]

0ing them out doesn't change anything on the gen 4's. May be tied to something, but I've tried with them populated and 0ed and it made no difference on them for this issue.

On the gen V's however, it's almost like it bogs down the processing of the ecu and that is what causes the dynamic airflow to go crazy. It's like they can't process things correctly when they are 0ed out.

I still think it's a comprehensive or rationality check taking place retesting the MAF. Verlon has taught me many things just in the last couple of days with how things are tied together via test and rationality checks on the gen V's. I did not know there were so many things that had to be disabled or changed just for one component removal. I would be willing to bet that failing the MAF isn't all that's required to take it out of the equation, at least not completely for every drive cycle causing this "test/check/glitch" to take place.... Just like post O2 test take place all the way to 5000 rpms, but pre cat O2 test only take place up to 3200ish or something like that rpms or the fact that they'll use post O2's to verify the operation of the pre cat O2's. Seems like all of this should be the opposite to me.

So it sounds like you're selling me on the stand alone HPT ECM? LOL

[GHuggins]

So it sounds like you're selling me on the stand alone HPT ECM? LOL

Didn't know hpt was coming out with such a thing. Hopefully better than the same alternative from other companies... Hopefully more OEish with the other companies stuff mixed in it? Although I imagine it might be the same with background items you can't change without someone being able to get into the coding?

[edcmat-l1]

https://www.hptuners.com/core-ecu/

[GHuggins]

https://www.hptuners.com/core-ecu/

Yepper's. That's what Cringer's link took you to as well. I read through everything it was supposed to offer before posting my last Just hoping they give you access to everything to adjust - or at least the most of it, without having to use coders... Makes a big difference when you can adjust things in the background with the current OE ecm's. Of course since it's "offroad" obviously some things won't need tweaking.

[verlon]

I thought I read this on the forum long ago, but now I cannot find the thread...when you fail the MAF (not sure on E38 only, or all Gen 4's), the engine will randomly stumble or hiccup while steady state cruising. As I recall this is "normal" and there is nothing that can be done about it, and it won't hurt the car.

Still, I thought I would reach out about this. I recorded this log the other day. There were two stumbles 4 seconds apart. What appears to be happening is that the calculated cylinder airmass spikes which then causes the injector pulse width to spike.

Hi, Cringer!

That's very interesting and I would like to investigate this. As I can see on the log there are spikes only on the airmass channel and dynamic airflow has no spikes.
If those stumbles are easily acheavable can you make a new log with a highest frequency polling of airmass, dynamic airflow and added Reference Period if it's accessible for logging on your OS?

[Cringer]

Hi, Cringer!

That's very interesting and I would like to investigate this. As I can see on the log there are spikes only on the airmass channel and dynamic airflow has no spikes.
If those stumbles are easily acheavable can you make a new log with a highest frequency polling of airmass, dynamic airflow and added Reference Period if it's accessible for logging on your OS?

Verlon, I would be really interested and grateful for what you can find!

I have hacked the scanner config to reduce the polling time to 0ms...I never achieve that, but it ensures the lowest latency possible...I think.
As you can see polling delays are going to be an issue here. The RPM's (for which I am using the broadcast PID) is grabbing data from the CANBUS at an average of 12.5 ms. The Cylinder Airmass is a request parameter and it is refreshing at an average rate of 22.7 ms. I will slim down the log channels to bare min though. I will grab the airflow channels and RPMs...anything else?

I do not see "Reference Period", but I do see these, which I assume will not help:
-Low Res Period
-Low Res Period Prev
What exactly does this channel report on?

I will grab a log when the snow melts...hopefully in less than a week.

fast poll.jpg

[verlon]

Verlon, I would be really interested and grateful for what you can find!

I have hacked the scanner config to reduce the polling time to 0ms...I never achieve that, but it ensures the lowest latency possible...I think.
As you can see polling delays are going to be an issue here. The RPM's (for which I am using the broadcast PID) is grabbing data from the CANBUS at an average of 12.5 ms. The Cylinder Airmass is a request parameter and it is refreshing at an average rate of 22.7 ms. I will slim down the log channels to bare min though. I will grab the airflow channels and RPMs...anything else?

I think that will be enough.

I do not see "Reference Period", but I do see these, which I assume will not help:
-Low Res Period
-Low Res Period Prev
What exactly does this channel report on?

May be that's it. Reference period is time between cylinder events. It is mainly used to calculate engine speed, airmass from airflow and vice vera.

[Alvin]

There is no fix that I'm aware of.

For OS's that can take a 2 Bar OS that resolves it. I've yet to run into a 2 bar OS do this.

I've never seen or heard of any damage coming from the VVE bug/hiccup.

[Cringer]

There is no fix that I'm aware of.

For OS's that can take a 2 Bar OS that resolves it. I've yet to run into a 2 bar OS do this.

I've never seen or heard of any damage coming from the VVE bug/hiccup.

OK good to know, thanks for confirming!

[JayRolla]

I'm glad to know this is normal because I always thought it was a slightly fueling issue with the VVE table. Never happened with MAF enabled so just ignored it.

[squeeler642]

I thought I read this on the forum long ago, but now I cannot find the thread...when you fail the MAF (not sure on E38 only, or all Gen 4's), the engine will randomly stumble or hiccup while steady state cruising. As I recall this is "normal" and there is nothing that can be done about it, and it won't hurt the car.

Still, I thought I would reach out about this. I recorded this log the other day. There were two stumbles 4 seconds apart. What appears to be happening is that the calculated cylinder airmass spikes which then causes the injector pulse width to spike.

My only guess this that the zero g/s being reported by the MAF is throwing off the Engine > Airflow > Dynamic > Prediction Coefficients (or at least the formula using the prediction coefficients and the MAF input). Is there a work around to this?

stumble.jpg


E38 MAF Failed Stumble.hpl

OK so I can't help but comment here. I stumbled into this post searching for "polling rate" of the MPVI3 in standalone vs over USB with VCMsuite. Still looking for an answer to this, although you guys kinda touched on/indicated a limitation with the HPT CORE discussion.

I have been having nearly this EXACT issue with my Gen III truck and have scoured the forum for hours in search of someone else having this issue ... to no avail. Here's the link to my thread and logs. I had some TPS codes thrown in the midst of my MAF codes so replaced the TB/TPS and am still seeing the issue occasionally when cruising, but it's better now and the TPS codes are solved. Before this, I tried replacing the MAF with a parts store brand and it made the issue worse, so I swithed back to my original MAF and double-checked everything for leaks. Still getting P0101 MAF performance/range codes.

My issue with all this is that my truck just started doing this about 3 weeks ago, and it's been in the same state (same tune, no hardware changes to the truck) since January of this year when I swapped a 6.0L into it. Seems like there was some step change, where now I'm having (what I called) "stuttering" while cruising like you describe, but have yet to see any MAF or related issue at idle. I was talking to a buddy at work today and he and I pretty much agreed after lots of discussion that the next thing to try is cam/crank sensors. I can't see/log enough cam/crank data (or can't figure out how) on my older truck ECU to watch those sensors as the truck hiccups... but the theory we were able to put together it is that it seems like the cam/crank sensor intermittently cuts out, which trips out the ECU which pulls all the timing out in an effort to prevent a lean-under-power condition.

Any thoughts on whether our issues are similar?

[Alvin]

This issue is GEN 4 only. IF you have a stumble in a GEN 3 it will be related to something else.

This VVE bug acts like a variable overflowed or something. For a instant, randomly, and are unable to reproduce. When you look at the scanner when this happens things max out in odd directions. Like a 2000% duty cycle, 0 airmass, etc.

[squeeler642]

This issue is GEN 4 only. IF you have a stumble in a GEN 3 it will be related to something else.

This VVE bug acts like a variable overflowed or something. For a instant, randomly, and are unable to reproduce. When you look at the scanner when this happens things max out in odd directions. Like a 2000% duty cycle, 0 airmass, etc.

Here's my issue, in a screenshot and attached log. Seems like the exact same symptom. Thoughts on how this would happen on a Gen III truck?

[Cringer]

Hi, Cringer!

That's very interesting and I would like to investigate this. As I can see on the log there are spikes only on the airmass channel and dynamic airflow has no spikes.
If those stumbles are easily acheavable can you make a new log with a highest frequency polling of airmass, dynamic airflow and added Reference Period if it's accessible for logging on your OS?

Hi Verlon, it took quite a bit of driving tonight. I didn't think it would happen, but then it started happening every few seconds, it was unstoppable. I removed as many PID's from the channels as I thought responsible and changed the polling interval to 0ms. Good thing too, as most of the logs show that VVE & Dynamic Air & Cylinder Airmass all spike. The previous log only showed Cylinder Airmass.

I actually captured 14 instances of this happening! Don't worry, the logs are all very short and can be access here.
https://drive.google.com/drive/folde...kP?usp=sharing

I would call out your attention to this one, it is another back to back occurrence.
SD Stumble (back to back).hpl

I am curious to know what you can find, although I am not sure there is going to be any real fix we can do with HPT.

[Cringer]

Here's my issue, in a screenshot and attached log. Seems like the exact same symptom. Thoughts on how this would happen on a Gen III truck?

There may be an injector pulse width spike, however, your MAF is not failed (so as to force speed density operation). You are getting some TPS and MAF DTC's apparently, so if the ECM is getting unreliable input from system components, it is logical to conclude this is the cause of your fueling issues.

1) Get your DTC's sorted out and I bet your issue goes away.
2) Then as a test you can fail the MAF in the tune and see if you still get this issue. However, as far as I know (which isn't much) about Gen 3's, they operate just fine with a failed MAF.

[verlon]

Hi Verlon, it took quite a bit of driving tonight. I didn't think it would happen, but then it started happening every few seconds, it was unstoppable. I removed as many PID's from the channels as I thought responsible and changed the polling interval to 0ms. Good thing too, as most of the logs show that VVE & Dynamic Air & Cylinder Airmass all spike. The previous log only showed Cylinder Airmass.

Exactly! And you've caught far more precious thing

ve.png

Thanks, Cringer.