Random Idle Throttle Blip

[GHuggins]

To begin with this only seems to happen on trucks. Don't believe I've ever seen this happen on a car I've done. Basically whilst the truck is idling for no apparent reason (AC doesn't seem to be cycling or anything like that) the throttle will "blip" open and timing will drastically dip. This can be seen at 4:20 in the attached log. Sometimes you can see the transmission going through some sort of self test energizing the solenoids when this happens and other times like this one there isn't a rhyme or reason for it. I can usually tweak the torque model and idle controls and get the large majority of this out, but would really like to know what's actually causing it even if it's something we currently don't have access to if anyone knows.

[5FDP]

I've seen my truck do this too, don't know why either.

[GHuggins]

I even thought about maybe an O2 test as it will drop timing to 0 and it only seems to do it on startups primarily. I've worked and worked torque models to get the most of it if not all of it out on several along with tweaking idle throttle corrections and accessory limit settings, but if I knew what was actually commanding it to do it or at least what was actually going on then it would probably be something as simple as changing another table elsewhere even if it's not available at the moment.

[TriPinTaZ]

Not that this helps much, but I had a cammed Camaro ZL1 that would do this. My BEST GUESS was that virtual torque had too much of a difference in values between RPM and Airmass and MAP in the idle area and whenever it would move just slightly out of the normal idle range, it would kick off a chain reaction which required the ECU to add throttle angle and then take it away to stabilize idle.

In this particular scenario, the fix was to adjust Virtual Torque, the MAF curve and tighten up the speed control reserve table.

[MORLOK]

My truck used to do this and it would go away if I failed the MAF. I think it happens when the MAF and VVE don't agree. I think it could also be caused by the MAF and throttle airflow tables not matching. Is there any possible source of un-metered air getting past the MAF? Like through the PCV system?
I have since vented my PCV to atmosphere and dialed in the MAF and VVE and now my truck idles dead smooth.

[GHuggins]

My truck used to do this and it would go away if I failed the MAF. I think it happens when the MAF and VVE don't agree. I think it could also be caused by the MAF and throttle airflow tables not matching. Is there any possible source of un-metered air getting past the MAF? Like through the PCV system?
I have since vented my PCV to atmosphere and dialed in the MAF and VVE and now my truck idles dead smooth.

This is interesting. Trucks do have completely different pcv systems and they did re-engineer them a few times over the last few years.

Even though it's not entirely accurate I normally fail the MAF and dial each model into one another - actually works decently, but must be done after fuel corrections, which I'm still making on this one - making torque and timing along with the slopes of the models the same so when engine load increases torque will stay close just so nothing odd like this happens is how I normally do them. This particular one hasn't had that done 100% yet as like I said fueling isn't 100% perfect like I like them and I've changed parts of the models to fix some factory decel issues - again only the trucks seem to have those problems and even at one time had a tsb for just that. I've just been trying to keep the models on this one close for now and it will get better then get worse. I've just noticed even though torque and timing are the same even on the ones where you've balanced them out between failed and theoretical 100% maf it'll still do this from time to time. It's not till I massage the idle controls that things usually smooth out and even then some will want limit increased substantially whilst others need it lowered and the adaptives upped a little. Like I said, I've never had a zl1 or any car that I know of do this even when the models aren't 100% balanced cause I'll dial them in to keep timing and throttle right for take offs, idle and everything like that. It's always just the trucks for me that do this and it's usually every one of them unless they've had some serious mods done and then they won't do it at all.

In fact the more I'm thinking about it, the trucks that don't do this have usually had blowers added or something done that changes the pcv system... Those will idle glass smooth... There may actually be a true 100% connection there with that throwing the MAF off??? Early trucks have 3 pcv lines whereas the newer ones only have 2. Then some trucks have actual old style pcv valves while others are running the fixed orifice. I know the ones running the 3 have to keep them otherwise they'll build crankcase pressure which I also don't understand why 2 aren't enough on those like the newer ones.

I had to go back and check, but I had tuned a 6.2 cammed swap a couple of months ago that had breathers instead of the normal pcv setup. It didn't do this either...

So now the question is, does one skew the airmass side to the positive to counter this? I can't say that I've paid attention to which way I had to shift the models to get rid of it. I may have to shift both to keep them balanced. We'll see where I wind up...

Guess I'm just overthinking things again

[GHuggins]

Not that this helps much, but I had a cammed Camaro ZL1 that would do this. My BEST GUESS was that virtual torque had too much of a difference in values between RPM and Airmass and MAP in the idle area and whenever it would move just slightly out of the normal idle range, it would kick off a chain reaction which required the ECU to add throttle angle and then take it away to stabilize idle.

In this particular scenario, the fix was to adjust Virtual Torque, the MAF curve and tighten up the speed control reserve table.

You normally just adjust the airmass side. Did you lower or raise it to counter this? Really thinking more and more about that pcv thing on the trucks though...

[TriPinTaZ]

In this instance it was increase the MAF in the idle area and then decrease the VT in the idle area. Also to tighten up the speed control reserve table. But I think the important part was that I made the Airmass/MAP VT areas for where it idles, and the adjacent cells, have less of a variance. I guess sort of a mild flat spot with a mild slope. Since the cam sorta kicks around the Airmass and MAP values more than a stock cam would, I needed the torque value to stay more steady. Now on the MAP stuff, you can actually alter VVE instead of lowering VT, but I actually like to do a moderate MAP VT change and then use Dynamic Air vs MAF air vs VE Air values to adjust the VVE get them more aligned. Typically I only do this on cammed cars that don't behave. Mild cammed cars don't usually need this.

Anyhow, yeah I understand trucks play dirty lol. They just don't respond the same way as The Camaro's and Corvette's in many ways. It's so bizarre! You'll figure it out though, I have faith in you. (PS thanks for hooking me up with Will, he's going to do some work for us. Much appreciated, sir!)

[GHuggins]

In this instance it was increase the MAF in the idle area and then decrease the VT in the idle area. Also to tighten up the speed control reserve table. But I think the important part was that I made the Airmass/MAP VT areas for where it idles, and the adjacent cells, have less of a variance. I guess sort of a mild flat spot with a mild slope. Since the cam sorta kicks around the Airmass and MAP values more than a stock cam would, I needed the torque value to stay more steady. Now on the MAP stuff, you can actually alter VVE instead of lowering VT, but I actually like to do a moderate MAP VT change and then use Dynamic Air vs MAF air vs VE Air values to adjust the VVE get them more aligned. Typically I only do this on cammed cars that don't behave. Mild cammed cars don't usually need this.

Anyhow, yeah I understand trucks play dirty lol. They just don't respond the same way as The Camaro's and Corvette's in many ways. It's so bizarre! You'll figure it out though, I have faith in you. (PS thanks for hooking me up with Will, he's going to do some work for us. Much appreciated, sir!)

Yes, I'll eventually be posting up how to modify those tables and what does what. I already sent some rough edited forms of the tables to Will showing what parts of the tables control what. I can't believe with how those tables affect things they haven't already been "discovered" and added in. Guess hpt hasn't bothered. That's fine for me though... The one table alone influences idle torque and timing, deceleration control, rev matching by controlling throttle closure rate, boost to throttle plate function control, free revving/throttle opening rate and lots more yet to be discovered I'm sure. Will just found some more for take off control to fix torque management stumbles on the big builds launching hard that I haven't had a chance to test yet. Probably won't for awhile as I've still got some more tuning to do on the drag truck build for one customer.

As for the VVE fixing the map side of the torque model... It DOES NOT ... I can tell you how to log and correct the torque models off idle though if you want. You can just shoot me an email on that. It's actually very easy especially since you already understand what does what. I just got through fixing a 4th gen having tcc issues because of it and his VVE was dialed in to a +/- 3ish percent error and even had cam movement dialed in that closely. It is funny though. Nearly every single time the map side of the torque model will look like the new VE table after it's corrected to what the ecms using.

Your actually doing the same thing in the idle area that I do. I heavily modify the map side as that's what changes the most and then modify the airmass side to favor the slopes if need be.

[GHuggins]

Ok, so all I did was add 8 lb/ft to peak airmass because as previously stated I knew it was pretty close and then interpolate down to negative load. I then lowered the external load from roughly 1600 rpms down and all of the throttle blips are gone at the moment. The customer did fail the MAF and got a driving/idle log on that showing the VVE was still +/- 0 to 3 % error at the most pretty much even after dialing that in a month ago, so was happy to see that, however it did reveal the airmass side of the model is going too positive in the higher load cells at idle. This is causing torque to jump up in the 100+ range and timing to jump up rather high at idle so I've still got to lower the peak loads on the airmass side some more and then raise the lower load areas back up.

I'll get it all sorted. Was really hoping it was just another setting somewhere since it is primarily on trucks, but perhaps it really is just the airmass side of the torque model due to air bypassing the MAF from the excessive pcv hookups... Thanks again Morlok for that insight...