When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
So I have battled this issue for 2 years with 2 different cams. Please help.
When I come to a stop, the RPMs dip and usually oscillates pretty bad, but once I come to a complete stop, adaptive spark takes over, and everything is fine. It idles well. But I wouldn't say the idle is so stable that as a load is placed on the engine it will take it with ease (this is in SD or MAF/VE Blended - MAF only is rock stable, so I know there's an issue with the VE table).
How did I tune idle? Turned adaptive spark off, set my timing at 20 degrees, and played with VCM Bidirection controls until it more or less idled on it's own (which pegs the wideband and the narrowbands end up reading like 20-30mv). It idles well at 850 with a 60kPa or so. But that's very lean. Cam has 5.5 degrees of overlap. It does not need to idle so lean. So when I turn on CL, the car adds fuel and it becomes unstable. Running with the MAF, I don't get that problem. I can run lean or rich in CL or OL and it idles fine. But as soon as I blend the VE table back in, the issues come back. But SD-OL is fine. Anything else with the VE table is flaky garbage. I have driven at low speeds and added 2% in the 400-800 columns until it becomes unstable while also taking 2% out. If I add 2-4% it goes unstable, but I can take 6-8% out and it doesn't seem to make a whole lot of difference (mainly cuz it's already pegging the wideband). I don't mess with the 1200 column a lot because it cruises there under 50mph. But there's a big delta between 800 and 1200 and that may not be helping in the VE table.
Injectors are from the L9H 6.2L Flex Fuel. Data was converted from Gen IV PCM to Gen III. The injectors are not flow-matched and you can tell in the log. It idles lean no problem. So it's not a PW issue. Also, I've changed the EOIC and PW data in the fueling tab. Both consistent with how I did it with the bigger cam (used the same injectors there).
As far as airflow, I actually plotted MAF air against IAC counts and Idle Desired Air (base airflow) against IAC counts as well. I've also plotted MAF Air and Dynamic Air (and they are typically within .5g of each other - so the VE table is close to the MAF transfer function or is that because it's running blended and using each other?). I've also logged Idle Desired against IAC counts and adjusted to where it's within .1g of reality. So the IAC Effective Area seems to be in line (also 22 steps ~ 1g sec airflow using the MAF as a check). I did drill a 3/16" hole in my 102 TB but have closed the blade enough to where the IAC steps are correct (or at least I believe them to be right enough).
So I've been trying to play with Cracker on coastdown and it helps pad the air a bit. But the issue is when I log Dynamic or MAF air against Desired Airflow... it's off by 4g/sec. And worse yet, when I run the RussK idle configure it wants 4g/sec at temp. But the car will not run with that. It likes 7.4g/sec like what I have in there. Why such a huge discrepancy? I didn't have that with my other cam. Usually, I just put what was in the config and added .5g and it was dialed in. So there is something else going on with that too, and I feel like it's contributing to my issues.
I've upped cracker enable/disable to 10/9 and that helps. But I feel like that's masking an issue. When I go back to stock 2/1 it really has problems.
I'm running 2.2.4 on my tuning laptop, but have 3.4 on my Macbook and Windows 10. But I haven't tried to tune with it yet. So, I've attached my 2.2.4 tune. I have one log attached. I need to get a log in a parking lot with the oscillation, but I'd rather it work without adaptive nannies and then I can enable to really make it rock solid.
I guess I could turn off the O2 sensor low voltage DTCs and go OL at idle... like I did with the bigger cam by setting PE to 1 in the idle cells. But I shouldn't have to.
In the log, you'll notice every time I come to a stop... I overshoot the idle (850) and end up in the 700s and then it surges back to 1300 to catch it. This is very annoying and noticeable when coming to a stop. Panic stops worry me because they could pull the car all the way down to a stall.
I notice that your Idle Spark Overspeed and Underspeed tables are very aggressive, adding or subtracting 11* when under/over by 150 RPM. Try cutting them in half.
Also in your Idle Spark Advance in Drive, you drop the advance by 4* to 20* when going from the 1200 to the 800 column. Try duplicating the 1200 RPM column for both 800 and 400 RPM. I sent you my tune - you will see that my Overspeed and Underspeed tables only go up to 6* and that my idle advance is 24* between 400 and 1200 RPM. Reason is that my 19* cam has the best vacuum (kPa) at 30* advance at 950 RPM. Therefore with a base 24* having the underspeed table add more than 6* would be useless and big values will cause oscillation.
Hey - no guarantee that will work, just something I would try myself.
I'll try. I just updated my overspeed/underspeed tables. Trying to get them more aggressive. I had capped them at 9 degrees. It doesn't really make much difference.
I've had it as much as 24 all the way across as well. I went down to 20 to get the NOx smell down. And to give me more headway in the idle/overspeed correction.
The bigger issue is it oscillates before I ever get to idle. So it's the return that's screwed up.
I don't like how much difference there is bank to bank. I'm reading my wideband off Bank 2... and it's not correcting much. But the driver's side is screwed up. I have some GT500 injectors I may throw on there. They are flow-matched. The data I have for them is pretty good from Banish.
Jake I can't look at your tune or anything, but a simple thing I've found is often the follower delay is set to zero. On mine, simply adding a 2-sec delay stopped it. It has a momentary hang at 1300 and then settles down. It seems to me on larger throttles that delay is helpful to prevent the dip and surge, by keeping the IAC open and allowing airflow to settle a bit before closing it off and dropping to idle.
Yeah. I opened my TB some more. Trying to balance between getting it to hang and not enough air. That helped some too.
I also increase the cracker to like 2.5g at 1000rpm at 16 and 12mph. It holds the IAC open. It doesn't really dip now. But it does hang open a little. I can play with it some more.
MAF CL or OL doesn't do it. Blended or VE it does. So the transition in the VE table is still off I think for fueling, but it's as close as I can get it on the street.
So you don't normally run this tune as a MAF CL tune? I noticed this same issue with my tune, but when the MAF is on, it's fine. I also cut the overspeed/Underspeed tables in 1/2 and it helped a little, but not anywhere near perfect. I did lower my timing by 3* in those areas and that helped more, so I guess overall it still was partially timing/torque related, but I can't see what the MAF is doing different, because airflow #'s on mine, are almost identical.
I have seen similar symptoms from throttle decay table. However with the maf info you gave I would check that the timing isn't changing based on airflow difference.
Return to idle is controlled by air and spark. Too much cracker can cause as much problems as not enough. Watch your MAP readings... they tell you what the engine is doing.
Okay. I think I have spark okay. It's pretty clean there.
I'll watch MAP to see what it's doing. But what I've seen there typically is it coasts down under 40 and it's around 1400 RPM and the MAP is in the low 40s. Then as I slow down it comes down to 1300 and 1200 and the MAP continues to go up toward 46-48. Then Around 1000 RPM or so and under 20mph it continues on 52-53-54 and when I hit about 17mph, it dips down to 900 RPM or less. And my MAP readings jump up to 56-57 then back down to 48 (all in line with the RPM swings which usually go back up to 1300 or so). It idles at 850 around 59-60kPa. So it doesn't cleanly fall from 52kPa to 56-57 and then down to idle. It seems to dip prematurely and then it oscillates to catch itself.
One thing I observe is that your VE table has a lot
more "MAP influence" than mine. That is, VE follows
MAP strongly. Now to the extent that this is true,
fine. But if you have actually more "modeled" MAP
effect than real, this makes a positive feedback
possible (more MAP gives more calculated air,
more fuel, then need more MAP to run...). When
I was bedding in the VE I got into some what I'd
call "MAP surge". And fueling to me is the main
cause of idle surging. Add to this that it's tough
to get true data from idle range RPM, across the
MAP axis so whatever you used to fit the idle range
VE in that dimension, may be suspect or may be
nothing at all.
I think with the bigger cam your low-end VE should
be lower than mine and it isn't really - high-MAP you
have about the same and low-MAP is less, so probably
more realistic.
Pic - yours on left, mine on right. Can't say mine's
anything but "drives nice, idles stable, reads right"
with a 224 on 116 cam. But that's what you want,
right? I'd try tamping down the higher-MAP lines so it
looks more like mine in terms of span / step, in the
400RPM and 800RPM cells, keep the bottom as your
point of reference.
Hmm interesting. So what do you suggest I do? Obviously, I'm using a wideband and adjusting based on error. But I'm not completely sure my wideband is worth a **** anymore.
As far as the 800 cells... I tuned it without the wideband. The feedback there is unreliable. I just do it based on sound and MAP reading and narrowbands. Try to get the narrowbands to read close to 0, get MAP readings as low as they will go, and get it to chop. Play with bi-directional controls to adjust fueling. From there I use the wideband.
But I notice large differences cell to cell as I go lower in the RPM range. I see in yours the VE table remains tightly bunched so there are no large swings in fueling.
How do I fix that? When if I were to smooth out my graph to look more like yours I'd be significantly rich or lean in many places.
Here's screenshots of the log. MAF stays steady around 12-13g as I approach idle. Dynamic airflow swings. At idle, it's in line. Above 2000 they generally stay within 1g of each other. But at say 600 RPM or 1400 RPM there is a significant difference. Which is where my VE table is broken. Of course in HPT 2.2.4 you had to adjust by percentage and not g/sec. I need to check 3.4 and see if that's different. In EFILive, you can view VE by g/sec vs whatever the hell the % values mean. I think I would have seen this much earlier looking at the VE table.
What this tells me is I'm off in the 400 RPM cells low by up-to 4g/sec at 55-60kPa. And I'm off 4-6g/sec high in the same MAP region for the 1200 cells (and probably 1600). I guess I could try something there. At least in those idle MAP regions 55-65kPa. And multiply the 400 cells by 1.5 to try and bring them up. While reducing the 1200 cells by .64 and then blend around them. Well no - that gives me the same numbers as the 800 cell. Go figure. So I need to adjust. Stock table and even most of my VE table is only off 3-5 VE % points off from each other. So 800 looks okay. Maybe I'll copy it to 400 and subtract 4 and add 5 to 1200. And then blend. And see what happens...
Those logs include me forcing the cracker way up. Just to keep the motor from falling too fast. It really doesn't matter what I've done there. It doesn't help.
My idle timing comes down 28-25-24-20 in the spark idle table (at 0% TPS) and the main table mirrors it. It's when it starts oscillating that it jumps back up from 20-26 or whatever. I've had it set at 24 from 400-2000 and it didn't matter. That's what I'm saying. It doesn't matter. I've taken all Cracker airflow out. Added 2.5g in. Held timing the same.
The only thing I haven't done is play with the VE as extensively as I'm about to do today. Trying to let it warm up a little. 55 deg and the drag radials are angry. I'll probably set cracker and hold .5g until 12mph and reduce it to 0. Spark I'll hold at 22 from 400-1600. Take those out of the equation.
So I guess I don't have any idea what you are saying. Cryptic messages don't help. It's a mess. Why. What would you do that I haven't done?
Last edited by JakeFusion; 12-03-2016 at 10:16 AM.
There's nothing cryptic about my message. Look at your spark and look at your IAC counts, they both bouncing around... do you really think your engine is gonna take that **** and give you a smile?
Your MAP is a rollercoaster ride... you need to smooth it out. You can try messing with the VE table but it's not going to make much difference when you're in closed loop... the trims will clean up any errors in the VE table. A nice VE table just means your trims won't have to work so hard.
11-26-2016, 05:17 PM
Still get oscillation coming to a stop 
