LQ4 single turbo e85 cannot make it idle... Help..

[carlsonauto]

Ok I have a 2004 lQ4 6 litre that I put in my 94 OBS 4x4 truck. I have it all working happily (well its getting there) except no matter what I do I cant seem to make the dumb thing idle. I have searched and read and searched some more and I have done everything I can find and think of to make the thing play nice. Its just crabby about life and nothing I do seems to make it happy... I have done lots of things that have made huge improvements but its just not right. The cam isnt even that crazy, I cant believe its giving me this much trouble. A little info about the combo im running.. The cam is a comp 54-444-11 (224/230 duration on a 114 lsa) It has an aluminum aftermarket ls1 style intake, 102mm throttle body, the pcm and harness were originally out of a 2003 silverado 5.3 4x4 its converted to DBC and SD using 04110003 OS and a base tune from a 2004 GTO. Another thing, the 102mm throttle body that I am running was not a name brand unit and it originally only had about a 1/4" passage from the IAC to the manifold vacuum side of the throttle blade, :doh2: needless to say that made the effectiveness of the IAC stop at about 40 steps so I looked at a factory throttle body and its passages everywhere are huge as I suspected. I took it upon myself to put it in the mill and rectify the problem. I did measure all the other dimentions of the IAC opening (depth, diameter, main passage opening etc) just to make sure that all else is correct and It is all within .010". I put it all back together and the IAC actually does something in the higher counts now. I also have the "flex fuel" sensor wired up and plumbed and for some reason that doesnt seem to be functioning either.. If anybody has any ideas or a "miracle fix" for either issue I am all ears. Its probably something obvious so here is my current tune that I am using to "auto-tune" my VE tables with. 13103 thank you in advance...

[bmax]

B4603 looks low. Have you done the RAFIG process. I would start there.



Brad

[carlsonauto]

I have not, I cant get it to idle long enough to log anything...

[carlsonauto]

this may be a dumb question but if I set the idle (in the tune not on the tb) to say 1000 rpm and datalog to get the RAFIG will the numbers be off be off because of the different rpm?? I am assuming so, thats why I have not done that either......... although it may get me close enough to make progress...... any thoughts??

[carlsonauto]

ok I reverted everything I had changed back to stock and got it to idle long enough to do the rafig with hopes that that would make a huge difference... not so much.. This thing is making me feel dumb, I made a tbi 383 idle with a way more radical cam and a much dumber pcm, I should certainly be able to get this thing to idle... I kinda have a gut feeling that something is hurt in the engine so I am going to do a compression test in the morning just to rule it out... Assuming its not hurt what should I change first to help it idle better?? I have my own ideas but im very much open to suggestions...

[macca_779]

Add 5 grams to desired air across the board to get it running and go from there. It's difficult to tune something that won't run


Sent from my iPhone 4 using Tapatalk

[carlsonauto]

ok I am pretty sure I figured it out... it wants to idle in the extremely lean end of the spectrum... I had tried playing with the V/E table before but stayed in the low 12 to high 13 range at idle because I have never seen an engine that wants to idle higher or lower than that it actually seemed happiest toward low 12's.... just for the heck of it I decided to make the V/E table flat from 400-1200 rpm and from about 30 to 75 kpa range and kind of a lot lower than I thought it should be.... (wend off gut and burning eyeballs) wideband read 14.5-high 15 even low 16s at times :Eyecrazy: but the stupid thing idles about 98% better.. I tried recalibrating the wideband because that cant be right (right??) but no change.. must have something to do with the wideband 02 placement in the downpipe or omething crazy like that.. I cant imagine its actually correct... :ermm:

[joecar]

B3647 and B3649: all those cells with value 1.00 EQR will cause the COS to STFT trim to stoich... i.e. the PCM will trim ahead of you regardless of your VE edits

( this is a feature of the COS's, in OL any cells in B3647/B3649 that are at stoich will cause STFT trimming... it's called semi open loop (SOL) ).


In your case it looks like you're using only B3647.

[joecar]

Also: between B3647 and B3618, fueling is not sufficiently rich for any significant load.

[carlsonauto]

guys thanks for the replies, joe as far as the B3647 and B3649 go I am a bit confused as to what they do.... :Eyecrazy: Is that where the GM.AFR pid gets its input from?? I have been watching that on my logs and wondering if i can and how to change that. Are you saying that no matter what I do with the v/e table its going to keep going rich if I dont change B3647??? It seems (in my head) that regardless of what B3647 says I should be able to manipulate it however I want with the V/E table, even though thats becoming apparent that that is not the proper way to go about doing it.. That brings me to my next question... What is the proper method?? If im seeing and understanding this correctly (and I very well may not be) table B3647 is basically just a mutiplier of sorts to the V/E table correct?? Also I would love to be using the B3649 table, I have the "flex fuel" sensor all wired in and plumbed and unless I am missing something in the tune I believe it should function, however the pid for ethanol shows 0% all the time and im at least 70% am I missing something here too?? sorry for all the dumb questions but it has been 7 or 8 years since I drove around in this thing pushing buttons on the laptop and that was with tunerproR/T on a TBI 383.... EFI Live on an Ls pcm is a bit more complex but I know if I could make that dinousaur pcm play nice I can certainly handle this.. in time..

[carlsonauto]

Also if there is one thing that I remember from tuning before its dont screw with stuff that you dont understand and dont change a ton of stuff at a time.. That being said I probably shouldnt change anything :joke:

[joecar]

Your tune file is setup for doing tuning the VE table (MAF is disabled, CL/trimming is disabled)...

you have a COS installed...

in OL the PCM uses B3647 as the commanded fuel;

a feature of the LS1 COS's is to do STFT trimming on any cell in B3647 that is set to stoich (i.e. same AFR as B3601, or EQR 1.00 if you're using EQR units); this feature is called SOL;

so while you're trying to correct the VE table using a wideband, SOL is trimming ahead of you, so any VE adjustments you make seem to have no effect;

change all the stoich cells to values that are slightly off-stoich, this avoids SOL trimming.

[joecar]

B3647: PCM reads this to know what air/fuel ratio to command in OL;
B3618: when enabled (via B3616 triggering) this overrides B3647 as the commanded air/fuel;

PCM uses VE to calculate cylinder airmass just ingested;
PCM uses B3647 or B3618 (if triggered) and B3601 to determine what air/fuel ratio it will command (AFR = B3601 / B3647 in EQR units);
PCM uses cylinder airmass and commanded air/fuel ratio to calculate required fuelmass (i.e. fuelmass = airmass / AFR);

so yes it is a multiplier/divider... you can think of it as such... or you can think of it as I described.

[joecar]

Also, make sure you know what units you're viewing the various tunetool tables and scantool maps in...

:)

[joecar]

B3647, B3649, B3618 are really defined in EQR units... but you may be using the tunetool's AFR display convenience (which uses B3601 to modify the display of the other tables).

Same with VE table being defined in units of g*K/kPa... the tunetool can display in % units by using the cylinder volume;


I like using EQR fuel units and g*K/kPa VE units;

also, it is easier to use Metric units for MAP and MAF.

:) :)

[carlsonauto]

wow thats a lot of info but i think I understand it for the most part and yeah the different units can get confusing but i think i finally have all my pids and maps and everything in the tuning tool all the same finally. I have always used the afr method for tuning so yes that is the method im using. and yes the first thing i did was change everything to imperial units in the tune tool then started tuning and about 5 minutes later i realized that was not at all what i was used to so i changed everything back to metric except the all the speed ones, those are still mph... so if i want the B3647 and B3649 to not do anything while doing the auto V/E any value other than 1.0 is good and basically disables the table?? then when i am done with the auto V/E I should change all the cells to the desired EQR for each respective cell and it will function as intended??

[joecar]

Yes, during AutoVE tuning, set B3647/B3649 to values other than EQR 1.00... if you need to be close to 1.00 then use 0.99 or 1.01.

Yes, then when finished tuning, set the desired values (including 1.00).

[joecar]

More info: Summary-Notes (http://forum.efilive.com/showthread.php?14188-Summary-Notes)

[macca_779]

Suggest reading the auto VE tutorial. It will help you alot


Sent from my iPhone 4 using Tapatalk

[carlsonauto]

Suggest reading the auto VE tutorial. It will help you alot


Sent from my iPhone 4 using Tapatalk

I did but I didnt understand that part while i was setting it up so i skipped over it intending to go back and research it and apparently fogot all about it... now it makes perfect sense... and the way the tutorial says to set it up looks pretty darn close to what it should look like.

[carlsonauto]

so does the 105 kpa cell of the B3647/B3649 tables just carry over to what would be 105+ kpa??? So does it just use the values of the 105 kpa cells column under boost is what im asking if that makes any sense?

[joecar]

Your file contains a COS which has a Boost VE table which starts at 105 kPa;


you tune the Main VE table (make sure the 105 kPa column is hit sufficiently), and then copy this 105 kPa column over to the Boost VE table;

then as you tune the Boost VE table you should get confirmation that the 105 kPa column is correct;

( if you make any changes to the 105 kPa column you should copy the column to the other VE table; i.e. the joint between the two VE tables should be seamless )

[carlsonauto]

ok well with table B3647 in line with what it should be everything seems to work a whole bunch better... It idles pretty good, could be better and it will be as soon as I find more time to mess with it. I still dont understand why the ethanol sensor reads 0% anyone have any ideas?? Am I missing something obvious again?? Here is a copy of my new tune 13114 feel free to look at it and tell me if I have something else messed up...... or if someone needs a base tune or some ideas feel free to use it...

[carlsonauto]

Your file contains a COS which has a Boost VE table which starts at 105 kPa;


you tune the Main VE table (make sure the 105 kPa column is hit sufficiently), and then copy this 105 kPa column over to the Boost VE table;

then as you tune the Boost VE table you should get confirmation that the 105 kPa column is correct;

( if you make any changes to the 105 kPa column you should copy the column to the other VE table; i.e. the joint between the two VE tables should be seamless )

I have the V/E tables about as good as I can until the bigger wastegate shows up (maxes out the 2 bar map at about 3500rpm and 45% throttle) but what I was wondering about is B3647not the V/E tables does the 105kpa column carry over just like the V/E table would in a normal OS when the table runs out?? I assume so.. As in at 110 kpa or 115 kpa it uses the 105kpa column and so on.. it would make sense as the desired afr shouldnt need to change after that point...

[joecar]

Oh :doh: B3647, when you go beyond 105 kPa, the last column in B3647 is used beyond 105 kPa

( this is generally true of most tables, the edge from which you fall off will be the value(s) used )

[carlsonauto]

Oh :doh: B3647, when you go beyond 105 kPa, the last column in B3647 is used beyond 105 kPa

( this is generally true of most tables, the edge from which you fall off will be the value(s) used )

Thats exactly what I figured but I just wanted to make sure I wasnt missing another table since I completely missed B3647 in the first place