Unexplained Commanded AFR, need help!

[schpenxel]

Yeah, but at least it still had a VE table....:-)

True... although now that virtual VE tuning is built into HPT, I'd honestly prefer an E38. The logging rates are sooo much faster. I get 7-8 times as much data on an E38 vs. E40.

[mowton]

True... although now that virtual VE tuning is built into HPT, I'd honestly prefer an E38. The logging rates are sooo much faster. I get 7-8 times as much data on an E38 vs. E40.

E67 as well and has much higher IFR/Airflow hardcoded limits due to their forced induction applications.

Datalog a 97/98 and you can see the sampling steps :-) They have come a long way....

Ed M

[will15]

haha feeling the love for the e40.

[majorpayne317641]

Ok, so I did adjust all the OP FA gains to 1 to command the 1.0 ratio desired. Yes the VE is adjusted to account for the incorrect injector flow numbers, I think this is what Doug's intent was, he sent me a whole starter hptuners file with the base information so I could get the car running and then tune from there.

So I am left with two problems. These two were not present prior to my engine rebuild to a built forged ly6 (steel version of ls2), same cam and same specs except for a drop in compression from 11 to 10.6 and cylinder head port work mainly on the exhaust side but also blending both seats and cutting down the stem areas.

The first and most annoying is at light load, steady throttle, the car shudders/bucks, it can be at any speed and around 2400rpm and down, if I increase the throttle any form of acceleration no matter how slight will remove this sensation. If anyone wants to see an example I have a very long log file but it contains this occurrence. Pre engine build the rpm fluctuation with this cam looks to be around 25 to 30, and post build the fluctuation seems to be around 30 to 35 rpm. No amount of spark retard cures this issue.

Second is a stall religiously after engine breaking to clutch disengagement, I realize this is a throttle cracker table, however I cannot relate that to this ecm, I have only been familiar with the older LS1 ecm's. If someone can point me in the right direction to make this go away that would be great. I think during hot starts it also has a harder time starting, again with the base idle airflows... Is this the only table for idle? If so do I log as shown on the sticky or is it different for an ls2?

Also Crank Position learn, can it not be done on an ls2 with hptuners? I had to install the crank reluctor ring due to damage during shipment as UPS always has done to all in the past. I built a tool to align it properly based off of an OEM crank I had laying around.

For those that want the cam specs this is what ECS has on their site: The ECS Big Blower camshaft features lift in the .600" range and mid 230's duration on a 115LSA

To be sure any of this is not the build I will do a compression test when the tool comes in and I can post this as well.

Thanks for all your help guys, its slowly getting better.

[majorpayne317641]

Post 85 here is the ticket for the idle fix?

http://www.hptuners.com/forum/showth...ht=idle+tuning

[schpenxel]

Airflow final minimum (idle section). Increase it. That sets how far the throttle can close.

Yours is a good 20% under mine in the 1200 RPM and under areas, and mine is a stock engine. If you go too high the RPM's won't want to come down all the way and ignition timing will be super low (way under your base idle timing table)

[schpenxel]

Where in the log does it do the light bucking? What frame #'s or range?

And is your wideband connected through the AC pressure wiring? What's the formula for it so I can set it up? The .cfg files don't contain that info.

[majorpayne317641]

I fixed the idle/hot start/stall clutch in problem by following the post in that link above along with the config file he included. I have always gotten great results from him on here and another forum for a lot of tech stuff. A great resource. If you go on the log, any of the cruise runs in 2k and below gear. The ac log is a custom histogram I think I can post that right? I will try.

VE - AFR Error.hst

If you go on the log and look at any cruise area, zoom in on the mph from 60 to 80 for a 70mph cruise, do the same with the spark from 30 to 40, that jitter is the best way I can describe it. Looking at my logs from the same set up on the old engine however, I still see that jitter, but I didn't feel it before. I did change out the engine mounts to racing style polyurethane, I feel engine vibrations through the power train a lot more now at certain rpm and that is no big deal, but I don't think you would get that jitter from stiffer mounts would you? Like a go cart kind of feel, very hard to explain. I did a cylinder test on hp tuners 3 times and all times the engine was within 1% on all cylinders about 100 to 101 or 99 to 100 most cylinders matching no patterns either.

[majorpayne317641]

I drove the car to work to try to better replicate what it is doing. I can exaggerate the bucking in first gear at about 1200 to 1400 rpm, it will do it less at higher but will still do it slightly about 2k in 1st gear constant throttle. I can do a scan of this later tonight. This leads me to believe it is the cam and not a misfire. Just unsure why it's this bad if the same set up on my old engine did not do this.

[majorpayne317641]

Ok attached are two log files, second one is me manipulating spark by 10 degrees during cruise... still bucking, does it less with spark smoothing, and you can see the sharp decrease in spark, that is where I feel the bucking, I used a VE table from the tune that did not buck with the old engine but same cam.

log.hpl

log2.hpl

Disregard the afr wide band as I was running the AC

[schpenxel]

Ok attached are two log files, second one is me manipulating spark by 10 degrees during cruise... still bucking, does it less with spark smoothing, and you can see the sharp decrease in spark, that is where I feel the bucking, I used a VE table from the tune that did not buck with the old engine but same cam.

log.hpl

log2.hpl

Disregard the afr wide band as I was running the AC

Good to hear you got the idle part figured out

Please include tune you were running at the time also to match up with these two logs..

What wideband are you using?

Unfortunately sharing histograms or .cfg files will not include the setup info for custom PID's such as your wideband connected to the AC pressure wiring. So, what wideband are you using? I'll find the formula real quick and set it up so I can see what you're seeing on some of your other logs where the wideband was connected

Have you tried running it in closed loop at all?

[majorpayne317641]

I will post in a minute. I had it in closed loop for this.

[majorpayne317641]

Here it is...

[majorpayne317641]

Good to hear you got the idle part figured out

Please include tune you were running at the time also to match up with these two logs..

What wideband are you using?

Unfortunately sharing histograms or .cfg files will not include the setup info for custom PID's such as your wideband connected to the AC pressure wiring. So, what wideband are you using? I'll find the formula real quick and set it up so I can see what you're seeing on some of your other logs where the wideband was connected

Have you tried running it in closed loop at all?

I'm using the innovate lc2 it was calibrated a few days ago for the new engine logging. It's the formula you see all over the net on ac wide band reference.

[schpenxel]

Not trying to be difficult but I really need some frame #'s to even try to look at what's going on in those specific areas though

Also, add VE Airflow to what you're logging next time you take a log

You've got to get the VE/boost VE tables smoothed out though.. they should be smooth. If it's jumping between cells with very different values, spark will jump around (since it's based on airflow). I suspect that's part of the problem.

The amount of knock retard also isn't helping. Possible both are from the same issue.

Here's what mine looks like as an example.

Capture.JPG

Yours:
Capture 2.JPG

[mowton]

Not trying to be difficult but I really need some frame #'s to even try to look at what's going on in those specific areas though

Also, add VE Airflow to what you're logging next time you take a log

You've got to get the VE/boost VE tables smoothed out though.. they should be smooth. If it's jumping between cells with very different values, spark will jump around (since it's based on airflow). I suspect that's part of the problem.

The amount of knock retard also isn't helping. Possible both are from the same issue.

Here's what mine looks like as an example.



Yours:

Can I use those thumbnails in my book to start the VE section :-)....sorry a bit of sarcasm!

Ed M

[schpenxel]

Might as well!

[majorpayne317641]

Can I use those thumbnails in my book to start the VE section :-)....sorry a bit of sarcasm!

Ed M

I will work on a better looking one, I'm guessing your afr is off in places with this table?

[schpenxel]

Mine? Nope.

[majorpayne317641]

Mine? Nope.

How do you get a table so smooth and have zero error under all conditions? I understand creating something to match it as close as possible but every time I have ever logged a VE table there is always some sort of error, if you get it this smooth there must be error in some areas, light load/high load with variations in RPM's and gears sometimes overlap cells and they don't always agree on fueling.