LS3/E38 crank priming tabel / how to control fueling during cranking

Hello all,

I have to start with a short explanation: Ive done a LSswap to 2012 model year car in Europe what means I have to fullfill Euro5 emmission homologation to get aproval for road use. This means not a simple emmission test as done every two years here. A 11km run from cold start (engine/ECt, etc. @20?C) on a dyno style test setup inside a laboratory has to be performed and the total emmissions are not allowed to exceed the limits of E5 spec.

The engine used is a 2013 production LS3 out of a Gen5 Camaro. Airbox changed due to swap, routing of ignition cables slightly changed, aftermarket clutch with a slightly lighter flywheel. Also I currently run shorty headers and HJS 200cpi E5 catlysers, but I plan to change them for production style headers and catalysers. Belt routing is slightly changed, besides that the engine is bone stock. So far the car is running great with no issues. Also a regular emmission test can be passed with close to perfect numbers. (car is tested hot and stainary at specific rpm?s in that case)

I have done a first test run in the laboratory and it turned out ( as expected) that the cold start is massive issue emmision wise. After further testing, datalogging and emmision measurements I found out that the first 30s seconds directly after cold start are generating the majority of total emmissions. From the calculations Ive done roughly 50-80% (depending on which parameter) is generated during start up.
As well, the fuel priming seems excessive. I loges up to 110ms priming duration directly before cranking. As well AFR cmd is 11ish direct after start and goes up to 14.68 within a few seconds. From the signals I see on the upstream O2 (unfortunately no wideband logs yet) the engine is running very rich in the first 5-10s, since signals stay maxed out. (I know, those are digital sensors) As well a very strong fuel/emmission smell (like a non cated engines running rich) is noticeable in the first 30s after cold start.

EDIT: I should mention that measured HC and CO values are excesively high during that first 30s and are reducing dramitically after 10-15s more. Under idle, catalyst start to fully operate after roughly 7min based on downstream O2 signals. But emmission values are already close to acceptable after 50-60s. Fully in spec after 5-6min.

What I want to do now is lean out the cold start as far as possible, meaning as far as the engine is still able to perform a reasonable startup @20?C. For that prupose I searched for any crank priming/crank AFR/crank Airflow/crank VE table, but I wasnt able to identify anything like that.

Closest I get is fuel/general/FA mult stage 1 (I dont see a stage 2) and fuel/openloop/IVT gain gas. But the second named table seems not to affect the area to consider for cold start and running at low rpm. So Fa mult is pretty much the only strain I could follow.

I didnt realy tried this table yet in depth, but from what I learned yet it seems that this table cannot solve my issue.

Is there any other table I could look at? Is the priming table maybe something I have to request at the support? If yes, whats the code for this table and are there others I shold also request?

I added my current tune (what is stock in matter of most of the relevant bases and just modified to work with the swap car) and a cold start log @idle for reference.

Thanks in advance!

Quote:

---

Originally Posted by TRStech

Closest I get is fuel/general/FA mult stage 1!

---

that is the only table this ecu gets, the time is engine off time so the longer the engine is off the further to the right in the table u go up to 12 hours of being off to adjust verses coolant temp

Tahnks 07GTS, you already helped me alot with understanding this table since the description says "As cranking continues the AFR is selected from lower rows of the table..." and "Cranking FA Mult vs. Cranking Time". So I was assuming the time the engine beeing cranked by the starter is meant. So I also was confused by the long time ranges. But it makes much more sence as you described it. Thanks!

With this information I will give it another try, but Im still wondering if there are any further tables I can change to get rid of the massive priming shot. Sure, it will need a bunch of extra fuel for a good start, but 110ms is about 40-50 times the amount it would take to keep it idling in that rpm range.

Would it make also sence to retune the VE tabel since Im seeing LTFT about -15 when idling. These low LTFTs seems to occure in areas with very low MAP pressure over a good rpm range up to about 3-4k. Unfortunately I have no logs of a factory Camaro or similar to compare. I assume it runs directly of the VE table for the first seconds after cranking. If that assumption is correct it could be at least a part of the issue.

I also stumbled over the intake valve temp tables. Is the IVT a predicted value? How is it calculated? I assume this could also make a difference since fuel vaporization of primed fuel sitting at the walls of the intake, port or valve should be part of the OL startup AFR calculation.

If somebody could provide this, a cold start 15min idle log of a bone stock LS3, in best case EU spec, would be very helpfull for comparison purpose.

u also have [ecm 7977] which is the pulse threshold for a big or small prime i think (havnt actually used it as its a new adjustment) and [ecm20221] u can disable if too rich, yea if u are seeing negative trims then u can remove from the VE and or MAF, also the [ecm12209] is a value for the maf fueling at startup steady state before maf has good enough flow to be usable, few more there for u to play with

GREAT! Thanks 07GTS, those are some good points to tweak. I will go ahead and start testing. Since cold start tuning and testing is time consuming by nature this will for sure take a while. Parallel data logging with the emmission tester doesnt make things easier. I realy could make use of a second OBD port to conect MPVI and tester in parallel. :rolleyes:

Many thanks again, maybe Ill come back for the "few more to play with" later. ;)

First...you didn't do a 20c cold start. It was only 10c, which causes a massive difference in emissions. That's grounds for a redo right there. And if you're not running the catalyst you intend to run in the future, this may be wasted effort.

Sure, the Log is not @20C. But the laboratoty test will. The log was taken under colder conditions, since weather is still pretty cold here. Heating the whole shop to 20C is currently not feasible. This being said, the numbers of older logs Ive done at 20C are slightly better, but the priming puls is still 70ms +. For the emissions result this has pretty much the same desasterous impact.

Concerning the converters: I want to give the current setup a further try, since the values are looking not to bad after 40-50s and are, calculation wise, in a range which could fit in the E5 boundrys.

If the next test run with (own) emission check will not show numbers which are in a reasonable range of total emissions (of course again calculation wise) the next step will be to change the hardware. (converters)

If Im able to reduce the emissions in the first 30-40s far enough I can stick with the current exhaust system, what would be my favorit solution. Since E5 was achieved with compareable systems I want to give it a chance. For sure E6 will not nearly be in range, but fortunately I "only" need E5. But also E6 was achieved with LS3 engines successfully. But in that case with much more advanced converter systems compared to mine or the OEM one.

Further comments with respect to your trims, if you're 15% rich in the trims at idle, what that will do is ruin your open loop for the first 15 - 30 odd seconds (or however long your closed loop delay is set for) after start up. You'll be pig rich as you'll be sitting approx 15% richer than commanded. Definitely get those maf and vve tables locked away.

I'm presuming your tester is just a listening device (passive)? Or is it actively sending requests? If passive consider buying a obd pass thru plug that has an additional obd port hanging off to the side. I'm not sure if it's active how it would play with the MPVI. My industrial canbus knowledge says it'll be fine, but I have no idea about the automotive specific protocol stuff.

Thanks @hjturbo! Yes, the MAF and VVE needs realy a lot of care. The negative numbers are in the low rpm+low load area.

Im working with an MPVI2 for loging and tuning, waiting for a AEM X inline wideband currently. But since my basic issue is in low rpm / load condition, I will start with LTFT based tuning. Meanwhile Im setting up a case to put the complete HPT interface with pro line and AEM controller in a singel box for easier handling.

The emmissions tester is a standard tester as used in most German shops. We set it up to work based on external rpm signals without any connection to the car. I also played with the idea of a OBD splitter, but this wont help, since those testers are made for testing at running engine. So it dont do anything before cranking. Once a rpm signal is present, the tester can be strated but goes into a calibrating routine lasting arround 30-40s. So the most important part is over before the measurement will start. To bypass this, we work with the simulated rpm signal, which is running before the car is cranked up. Unfortunately the tester has the same behavior when operated via OBD, so the splitter wont help.

Maybe a completely stupid question, but since I started playing arround with open loop values and tables, my ambient air temp sensor is reading static 0C. Any idea about that? (No, I dont see any obvious defects or similar...) Is this a value which is calculated/learned and therefore reset when the calibaration is writen?

Not sure on the details of how or why but mine does the same e67. No dtc codes present.

the ambient air temp isnt right in hpt its not actual so best to ignore it

I watched this value over the last tune steps. It seems that it needs at least one cold start after writing a new tune before it starts reporting any values again. It seems that the start value is correct then, but during the drive with warm engine it seems to be an estimation.

As well I can report that the "FA Multiplier Stage1" is actualy more or less directly controling the priming shot.

The other tables have minor impact and might also be used for different calculations where changing them could be negative.

As well, I was surprised to find out, that leaning out the priming shot causes much more issues for the cranking at higher engine temps than lower, whats a little surprising for me, since vaporization and flame point should be much better at higher ECTs. Maybe the intake valve temp estimation is not very precise.

As well, for real cold start, it seems the priming shot can be reduced pretty far before the cranking process actualy reacts critical.

With regards to your last paragraph, and the knowledge you have gained of the FA table do you think you might be able to get it through emissions? There should be a PID Engine Off Time which will tell you what column youre in. And who cares about a shitty hot restart for your test right? If you know what I mean...

I have to do actual measurements. I just can see that it starts with a 40ms prime shot as good as with 100ms when cold. What this means for the total emmissions was not measured yet.

Sure, I will optimize the regions which are important for the test cycle. The cycle doesnt care about emmissions on hot (re)start, so I also dont need to care about that, meaning I will keep those regions as found. (Disclaimer: This doesnt mean I dont care about emmissions, but we are talking about a car passing the regular hot tests with ease and by far better values than most of the latest models but is struggeling with this test cycle...)

I meanwhile have some issues with geting my idle LTFTs/Airflow spot on. I have the phenomenon, that I start with tune A, seeing LTFT 0. Stop engine, save log, start engine -> LTFT +7. Shut off engine, change MAF curve by -5% arround idle Frequency, write tune B, reset LTFT and start engine -> -9 LTFT. Shut off, rewrite tune A again, reset LTFT -> -7 LTFT. I repeated similar steps 2 or 3 times. Im a little confused now. Is it typical to see those deviations at idle?

What is your injector tip temperature reading in the logs after flashing? Search for rich after reflash.

Thanks, this could explane everything from the second I did the first flash today. But Im still confused between the first two logs I mentioned (actualy #2 and 4 of today). The flash wasnt changed. Just the engine started, logged, stopped. Then the MPVI2 reconnected since I was under the impression that something didnt connect correctly (what was not true as it turned out) and restartet the engine.

In the following log, arround 5min later, no flash, same tune, same day and nearly same operating temp, the LTFTs are arround +7. As well the MAF frequency seems to be slightly lower while rpm etc. is identical. I attached the two short logs.

Attachment 109947Attachment 109948

BTW, emmissions on very cold start are already pretty much confirmed to be reduced by 40-70%. (depending on which value you are looking) Now Im focussing to get those improvements spot on to then turn towards the later time sections of the idle warmup run. Once that is done, the driving conditions will be the next step.
Dont know if there is a different Mapping for EU models of this car, but I can confirm that the base calibration was identical to US factory version. Still, one thing is sure, with the settings of the calibration as found, even taking the MAF into consideration, which was away from optimum at low flowrates, passing the emmission test (I mean driving cycle) is absolutely impossible. The car as delivered (under the assumption the EU calibration isnt different) would never fullfill those requirements. Still, I dont feel that it is impossible. But definately not on this calibration.

Heat soaking the engine will effect injector tip temp. Have you got it in your log? It's important.

Not in the historic logs, but I will in the future... :)

I also zeroed out the Offset vs. Inj. tip temp in the tune.