Transient tip in and tip out tuning

[PJBURGES]

Hi everyone,

Trying to get rid of a transient lean tip-in and rich tip-out effect. I don't really care about the rich tip out but the lean on tip in can be felt as kind of an initial dead spot. I attached a log file and my current tune. There are a few tip in events in there- some were worse than others.

QUESTION IS:
What tables in the transient tab would I alter, and in which direction, for this particular scenario? I understand only the basic concepts of the fuel wall "puddle" effect - and I am trying to understand what the best thing to try to adjust would be before I just go turning knobs.

Background info: Its a work-in-progress tune so don't beat me up! I don't have ignition dialed in yet, so yes occasionally I'm picking up a few degrees of knock retard. Its my next thing to fix once I get this transient figured out and then turn the narrowbands back on. My VE is like 70% there and the MAF is currently pretty good. So you'll notice in the log I am running about 5% error in commanded vs achieved EQ in steady state, that just is what it is for now. I may go back and dial it in closer this winter. I just got sick of fighting zone boundaries and everything else VVE.

Engine is a 6.0L running on an E38 ECM no VVT or DOD. Injectors are out of a L96 flex-fuel truck (12613412 50lb/hr)@ 58psi on the rail with a 450lph walbro pump. I think I have those injector tables right in my tune but honestly not sure! Wondering if some of this tip in/out is due to a discrepancy in injector timing vs valve event from the different cam. Has CNC ported 799 heads, Gen 4 intake with K&N and straight pipe, cam is a mild 219/228, 112 LSA. CR is ~11.3:1 Narrowbands are turned off, DFCO is off, MAF and VE and Dynamic Air are on. Trans is locked in 2nd gear, pulls are from a stop, or near stop. Converter stall is around 2500rpm-ish.

Thanks for any insight! There are lots of posts I've searched and read - but mostly theory related, hoping for advice specific to whats observed in my log.

[sgod1100]

In my experience and reading a lot of what other guys have said on here, you want to get your steady state fueling 100% complete before messing with any of the transients. So dial in your VVE/MAF as good as you can then start messing with transients. Impact and Evap factor are going to be the two main tables you will be adjusting based on what I've read.

[PJBURGES]

snap_of_leanspike.jpg

I know the VE isn't perfect but this spot in the log is not just EQ moving around from a bumpy VE map - this is pretty clearly transient lean on tip-in. Just curious which direction to adjust the impact factor and gain to fix it.

EDIT: I really think part of my problem here is that the factory cam would have had around -23.5 deg overlap and my cam has -1.5 deg overlap. My wideband will read 3-4% lean but even with the factory cat system in place if I sniff the tailpipe while idling it smells like gas (rich!). Something must not be right?? This change in valve timing has to throw off the EOI (End of Injection Target) and that has to be a substantial effect on everything. I may try to bump this down by ~20 deg in the Normal EOI Target adder vs ECT table. I've deduced only from forum posts that this would be the table to try and shift the injection timing in. Can anyone confirm?

[rabbs88]

Setting your EOI target correctly will help with the smell at idle substantially. There are a few calculators around the forum to help determine your new valve events and set your values accordingly. I havent looked at your tune yet but as far as your lean tip in, set your PE ramp in rate faster. Max it out if you need to. That will help for now. Transient fueling may come in to play later on but like stated previously, fueling needs to be accurate first. To smooth out your vve table (Again, havent seen your tune yet) copy your whole table, move your zones around, paste it back in and use the smoothing function on the entire graph (once or twice) until 3d graph looks OK. Then go out and dial it in. You will be chasing your tail if the zones don't have somewhat smooth transitions before you apply any changes.

[sgod1100]

snap_of_leanspike.jpg

I know the VE isn't perfect but this spot in the log is not just EQ moving around from a bumpy VE map - this is pretty clearly transient lean on tip-in. Just curious which direction to adjust the impact factor and gain to fix it.

EDIT: I really think part of my problem here is that the factory cam would have had around -23.5 deg overlap and my cam has -1.5 deg overlap. My wideband will read 3-4% lean but even with the factory cat system in place if I sniff the tailpipe while idling it smells like gas (rich!). Something must not be right?? This change in valve timing has to throw off the EOI (End of Injection Target) and that has to be a substantial effect on everything. I may try to bump this down by ~20 deg in the Normal EOI Target adder vs ECT table. I've deduced only from forum posts that this would be the table to try and shift the injection timing in. Can anyone confirm?

I've always seen that same time of ridge just prior to getting into PE and just left mine alone....judging by how thin that is it's barely any time has passed and you are in PE. Can you even feel that?

[PJBURGES]

Yea I'm at 1.2 lambda and the engine is commanding .88. in that screen shot for about ~.3-.4 seconds? It just feels "mushy" when you're behind the wheel. Like the power doesn't really snap in but rolls in a brief time after I mash the pedal. In neutral if you just pop the throttle real fast, then instead of it revving up clean it will very briefly stumble and then catch and rev up. But thats doing it from 750RPM with no load. I think I have my rolling idle at like 950 - which a lot of the log pulls are from 950 and with load and the stumble isn't felt, it just feels mushy. I like the idea of turning up the ramp in speed on my PE as suggested by rabbs88, and maybe even setting the target PE richer a bit near the stall speed of the converter to make it shoot harder that way with the existing transient parameters left as they are - rather than messing with the transient settings themselves. I'll give that a shot tonight, and I'll also look into the EOIT - my hunch is the transient parameters may be just fine but the new cam has made EOIT so far out of whack that all of these transient modifiers are no longer operating quite as intended?

[PGA2B]

Are you sure you are not feeling the dip in timing? In that shot you are at 6* of timing.

[sgod1100]

Yea I'm at 1.2 lambda and the engine is commanding .88. in that screen shot for about ~.3-.4 seconds? It just feels "mushy" when you're behind the wheel. Like the power doesn't really snap in but rolls in a brief time after I mash the pedal. In neutral if you just pop the throttle real fast, then instead of it revving up clean it will very briefly stumble and then catch and rev up. But thats doing it from 750RPM with no load. I think I have my rolling idle at like 950 - which a lot of the log pulls are from 950 and with load and the stumble isn't felt, it just feels mushy. I like the idea of turning up the ramp in speed on my PE as suggested by rabbs88, and maybe even setting the target PE richer a bit near the stall speed of the converter to make it shoot harder that way with the existing transient parameters left as they are - rather than messing with the transient settings themselves. I'll give that a shot tonight, and I'll also look into the EOIT - my hunch is the transient parameters may be just fine but the new cam has made EOIT so far out of whack that all of these transient modifiers are no longer operating quite as intended?

I've always used MAP trigger for PE. I know it's slower than the TPS trigger, but with having a ctsv a small pedal movement can result into getting into boost rather quickly. Try lowering your tps trigger to all 5's and using the map setting and see if it feels any better.

[PJBURGES]

Are you sure you are not feeling the dip in timing? In that shot you are at 6* of timing.

This is the ignition table: ignition_timing.JPG

Should the advance ramp in faster at those low rpm's? Is the OEM just trying to be super safe/conservative about encountering knock? It does feel and sound much better as the advance hits 21 degrees but its taking it about a second to build to that advance. Maybe that IS the source of the mushy feeling and sgod1100 is correct that the brief flick in the EQ in the chart is not what I'm feeling at all. This is a 5500 lb tahoe with a 4L80 and not a turbo'd CTSV, maybe I'm being to picky comparing a big truck to what I'm used to in sports cars hahaha

[PJBURGES]

I've always used MAP trigger for PE. I know it's slower than the TPS trigger, but with having a ctsv a small pedal movement can result into getting into boost rather quickly. Try lowering your tps trigger to all 5's and using the map setting and see if it feels any better.

These are my current PE settings:pe_current settings.JPG
pe_eq_current.JPG

[PGA2B]

This is the ignition table: ignition_timing.JPG

Should the advance ramp in faster at those low rpm's? Is the OEM just trying to be super safe/conservative about encountering knock? It does feel and sound much better as the advance hits 21 degrees but its taking it about a second to build to that advance. Maybe that IS the source of the mushy feeling and sgod1100 is correct that the brief flick in the EQ in the chart is not what I'm feeling at all. This is a 5500 lb tahoe with a 4L80 and not a turbo'd CTSV, maybe I'm being to picky comparing a big truck to what I'm used to in sports cars hahaha

I'm not a professional tuner, but I think the dip in timing is a Torque Management thing. As far as your timing table, I'm of no use there as I have a different platform with much more aggressive timing. I was just pointing out at the spot you showed, you had very little timing advance.

[rabbs88]

Just looked at the tune file you posted. You said your running L96 (12613412) injectors but your settings are a little different from an L96 file I looked at? They are very close but double check that. Your VVE table hasn't been touched from stock either from what I can see. Your EOI timing has been changed already. If it's good/correct I did not check. Are you running e85 or 93/91? Also are you still using a stock LS2 MAP sensor? I've made a few changes for you that should help get you started to dial in your fueling and possibly fix your stumbling issue. Use the compare function so you know what I did.

[PJBURGES]

Just looked at the tune file you posted. You said your running L96 (12613412) injectors but your settings are a little different from an L96 file I looked at? They are very close but double check that.

Ok - will do! Was it the offset vs pressure vs IGN voltage table that looked different to you? I used a 2010 3/4 ton Yukon VIN 1GKZREG0AR189018 Operating System 12636005 to pull my numbers for the injectors. RockAuto shows the Delco part # 2173412 for that year/model which cross references to the OEM 12613412?

Your VVE table hasn't been touched from stock either from what I can see.

Here is a screenshot of my VVE in my editor - I've done about 20 data runs/ iterations on it and I know it changed quite a bit from where I started! haha Maybe its showing up different in your editor?
vve_table.JPG

Your EOI timing has been changed already. If it's good/correct I did not check. - Its the same as the 2010 Yukon I pulled the rest of the injector data from. I guess this could actually be a source of my issue because that truck would have had rec-port and I'm running ported 799's (catherdral port) - By different I assume you were looking at like a 2007 corvette that is a true "LS2" and comparing? Hmm maybe I should toss in the transient tables and then the EOIT tables one at a time off the LS2 and see if the response looks better?!

Are you running e85 or 93/91? 93 octane. It has all the flexfuel hardware - I might run that at some point after things are dialed in better.

Also are you still using a stock LS2 MAP sensor? I'm using a MAP sensor off a 2009 Tahoe 5.3L. AC/DELCO 12711681.

I'll check out all your changes! Going thru the comparison now - you did exactly what I was about to try next regarding the PE tables.

[GHuggins]

This is what I recommend doing for dialing in the injection timing.

Start with your stock or even current settings - Now using this and not changing anything else in the tune make 3 different tunes. First tune - your current / Second tune - add 30 to boundary / third - subtract 30 from boundary from the original current table

Now flash and let idle for 20 minutes. At the 20 minute mark snap the throttle with a good wot quick stomp. Do this for all 3. Go back through the logs and see which one was running with the most negative fuel trim in your MAF table and the lowest injector p/w.

Which ever one was the lowest go that way with the whole table. Now see which one had the best throttle snap and showed the best transient fueling and richest on the wideband as rpms jumped up. Go that way with the boundary table off idle.

You can further use this spreadsheet to help visualize your cam events in relation to injector spray time.

[hjtrbo]

This is what I recommend doing for dialing in the injection timing.

Start with your stock or even current settings - Now using this and not changing anything else in the tune make 3 different tunes. First tune - your current / Second tune - add 30 to boundary / third - subtract 30 from boundary from the original current table

Now flash and let idle for 20 minutes. At the 20 minute mark snap the throttle with a good wot quick stomp. Do this for all 3. Go back through the logs and see which one was running with the most negative fuel trim in your MAF table and the lowest injector p/w.

Which ever one was the lowest go that way with the whole table. Now see which one had the best throttle snap and showed the best transient fueling and richest on the wideband as rpms jumped up. Go that way with the boundary table off idle.

You can further use this spreadsheet to help visualize your cam events in relation to injector spray time.

In all my years on this forum, this would have to be the best explanation I've read on how to get started with EOIT.

[GHuggins]

In all my years on this forum, this would have to be the best explanation I've read on how to get started with EOIT.

This is the quickest and easiest way to set it and get the jist of which way to shift the tables that I've found. I have all of my customers do this and then I basically have 3 different ways I slope the boundary table from that point depending on the cam and all around engine build/setup. I've also started using the rpm table a lot for advancing timing when need be for idle and up through 2600ish rpms. I wish the gen4's had offered different timing settings for E. Would have made a better all around setup.

You have to make sure and let it idle for the 20 minutes to keep the variables the same for the wonderful rich after flash - hopefully by this point that's mostly dialed in anyway.

Takes a lot of idling but it simplify's things a lot.

[rabbs88]

Here is a screenshot of my VVE in my editor - I've done about 20 data runs/ iterations on it and I know it changed quite a bit from where I started! haha Maybe its showing up different in your editor?
vve_table.JPG

My apologies, forgot when you compare VVE tables with the compare function it's the same table on both tunes unless you open the other tune in another editor window. I did already make zone adjustments for you in the tune I changed so your VVE table is much smoother. Not sure if you saw that or not. Also your right about the EOI targets being stock. Missed that too (It was like 3am). Otherwise I would do exactly what GHuggins stated to get your injection timing right with your cam. Should be done first before fueling is dialed in. He also made an excellent write up on getting your Offset vs Injector tip temp dialed in which would help the rich after flash you'll see

[NotSure]

This is the quickest and easiest way to set it and get the jist of which way to shift the tables that I've found. I have all of my customers do this and then I basically have 3 different ways I slope the boundary table from that point depending on the cam and all around engine build/setup. I've also started using the rpm table a lot for advancing timing when need be for idle and up through 2600ish rpms. I wish the gen4's had offered different timing settings for E. Would have made a better all around setup.

You have to make sure and let it idle for the 20 minutes to keep the variables the same for the wonderful rich after flash - hopefully by this point that's mostly dialed in anyway.

Takes a lot of idling but it simplify's things a lot.

Speaking of rich after flash, I am under the impression from volumes of posts that Injector Tip Temp is the root cause. I see the tip temp jump to around 950F sometimes after a reflash, but not always. It may be dependent on what was changed in a reflash, I dunno. Right or wrong, I put my tip temp chart like the following (zero'd out at the peak), since I never see tip temp over, say 150F, in normal operation.

injector tip temp.jpg

Hoping that this could resolve rich after flash? Or, at least pay attention to the tip temp when logging, and when the 950F stuff goes away in a few minutes, you're good to go.

[rabbs88]

Speaking of rich after flash, I am under the impression from volumes of posts that Injector Tip Temp is the root cause. I see the tip temp jump to around 950F sometimes after a reflash, but not always. It may be dependent on what was changed in a reflash, I dunno. Right or wrong, I put my tip temp chart like the following (zero'd out at the peak), since I never see tip temp over, say 150F, in normal operation.

injector tip temp.jpg

Hoping that this could resolve rich after flash? Or, at least pay attention to the tip temp when logging, and when the 950F stuff goes away in a few minutes, you're good to go.

Yes it is the root cause and you quoted the guy who explained it in great detail years ago

https://forum.hptuners.com/showthrea...ip-Temperature

[GHuggins]

Unless 0ing it out actually puts trims back in line, you don't want to 0 it out. Everything will change it including injection timing. It needs to be roughed in then dialed in after injection timing is corrected. Sometimes 0ing it out works correctly. Just all depends.

To my knowledge Jslic was the one who originally discovered this. I was working with him at the time on it. There was another guy toward the end of that thread that jumped on me for posting it and taking credit, which I never have. I also never knew he had anything to do with it. This was the time Scott and I were hunting down and having hpt add a bunch of new parameters into the cals using SCT and EFI... Wish they had added in more parameters that we had sent them, but guess they thought it wasn't worth it.