No dice. the interger speed does limit the cycle of surge but if I added the idle angle from the log it wanted even more angle. if I inverted that, it still wanted more angle. Intuition says that the surge is throwing off the calculation. ran this 6 tiimes and still basically does the exact same thing. somehow need to know what the initial error angle was before it starts surging as all it does is capture how much the feedback loop was. I also adjusted out my idle air flow a bit til it was reading +/- .01.
What angle vs airmass did you end up with? The stock table says you should idle at 3* 4* (3.5%-5%)once the dust settles. Your last log it started out at 16-17%(13*-14*) and slowly drops to just below 15%(12.5*) and then right when the dust settles, the car is going from emissions mode to stability mode, you see the final drop in rpm the MAF period starts to raise then you get a huge ETC error almost 12*. The car opens the throttle because it thinks it needs 3.8 lb/min when the ETC it is at 12.5* but it is only seeing .55lb/min. Once it opens it the RPMS climb above 700 and it closes it tring to maintain that RPM. the last Log had throttle % not angle.
I'll take a look at all this tonight from this perspective. I think I'm still misinterpreting what you're describing somehow. The part that I think will help the most is this... does the car want 12* of throttle at idle or 4? does it want .55lb/min or 3.8lb/min. 3.8lb/min sounds pretty high for idle, yes or is .55lbs to low? so I'm lost on which needs to move. I tried moving each independently as a control and got no change in behavior... tho it sounds like I wasn't doing something right. I'll post up the air mass after work tonight and check that I'm still logging the right pids. I appreciate you continued efforts to make this sensible. I can see what you say that the throttle angle map is fighting with the air mass map but not how to bring the two into harmony.
My guess is adding 10* of angle to the entire table so that you get .3- 11.28 and .8- 13.4. This should get the expected angle close to the airmass your seeing. 10* sounds like a lot, but you did increase the volume of the intake tubing significantly. So at idle the engine is working quite a bit harder and probably needs the help.
I figured it out. If I add 10 to each of the etc idle expected angle I get the numbers. .3 = 11.26. I saw that as .3 thru 11.26 and it just didn't click. Ok... off to take a drive.
I think there's something odd going on here. with the larger angle, yes, it has less of a surge cycle but it's very unresponsive while it surges. I tried something on the 2nd log to see if it was short on air and that made it worse. What do you think? Kinda nervous to leave that much throttle angle at idle if it's actually restricting it back down to 4-7*
Here's what I think is happening. By commanding such a large idle angle, the driver input has to be even larger to over come it if the car is in surge. I.e. now it keeps surging even with some throttle input because the table says it can use that much angle to get what it thinks it needs. so here's my new theory: The idle tables may be a little off but not much. What's off is the throttle body model. That actually seemed to be making progress. Or did you expect this to get worse before it got better?
So change improves idle but causes drivability issues. If you full the table with 0s the ecu should ignore the table. That way ecu ignores it and you won't have surging or drivability problems. Other than that leave the +10* and modify driver demand. I'm curious about what these logs show, I should be able to look at them in an hour or so.
I think the change you made would have helps but you went the wrong way. Your ETC vaccum at idle was lost after that change. Log Idle desired RPM and graph it against airflow. Get 1200 and under sorted first. Then work on the higher RPMs of that table.
One thing I noticed is your spark jumps up first out of everything. I checked your MP 0 borderline table and noticed you had it modified from 42/45 to 17 in the idle area. In fact all MPs are like that.What made you do that?
Im also not sure why under Idle spark advance when the AC is off it has 55* for all ECT*. Its like this in the stock file I got out of the repository but I would modify that to correct a high idle spark before the MP tables.
the spark table being 17* in the upper left 4 cells, That is a carry over from the procharger base tune. I've moved it around to no effect on previous log sessions.
I'm confused on why the changes you recommend are opposite of the error shown. The log error was asking for more idle air - which when that was increased by a small amount it made things worse. So the result and your advice are in sync but why is it opposite of the error shown? That doesn't make sense for the log tool to say "give me more air" when the advice on this is the engine is choking on too much. something is not right there. Got to be doing something wrong. I don't disagree with the idea to work one area and move to others but can we focus on finding the root cause and then adjust the other areas to suit? I'm not sure I see the root cause in all this.
Nope. stock headers, nothing changed there. I had some success tonight achieving a result I wanted. I got the idle adapt to be very close to 0. I bumped the air mass by .21 and undid some of the other changes, then added a little bit to the idle expected angle. Still seeing this data that it consistently thinks it's short on throttle angle of about 2* in the idle area. None of the items we've discussed tweaking puts a dent in that number, it's not a useful measurement at all since it won't respond to even gross changes. What's another way to graph out this issue? I'm also pretty convinced that the idle area is dialed in for now. Where else can we look: The surging happens only in torque-N which is not idle. It hangs in torque-N too long I think. How do we convince it to knock that off sooner?
Target N is a torque limit source. I believe its limiting torque to achieve idle RPM. Decel torque decay you can add torque to it to allow faster decel if possible. The other option is tuning MP 0 and 1 in the idle RPM and load ranges. Oh and driver demand table in the same area.
What it's telling me is it's seeing too much load(air) and it trying to limit it. This could be a torque/load value issue but because target N happens after the throttle error and opening I kinda expect it to be there. It lasting too long I would say is because of the extra volume in the intake/intercooler piping.
I agree with you that the target-n is dealing with the load issue as it returns to idle. For tonight I think I'm going to leave the throttle body alone and focus on the Torque/torque inverse in the low rpm area to reduce expected load til it lines up with the throttle body tweaks made to this point.
What keeps sticking out in my mind is that the log of desired air mass during the surge is about .8-1.3lb/min. the maf is consistently measuring .55lbs-.8lbs. I see that same logic in your description - but from the opposite direction.
basically the calculated load of 1.3 is higher than that of .55lbs and thus it enacts torque control on the throttle body to bring the load in line. which then trips the underspeed and it kicks in 4-5* of extra throttle body to prevent stalling. If I give the car extra throttle in the throttle body map it (to match the 1.3) it still reads 1.3lb/min desired vs .55lb maf which is my theory on why the etc error doesn't change. the car idles just fine with .55lb/min but it keeps expecting 1.3lb/min. at idle it ignores this but under torque control it can't. so it fiddles with it until some learn time limit is hit and it ignores the discrepancy. So I'll decrease the torque values and lets see if the desired air mass drops down closer to .55lb/min that the car idles at.
reducing the torque on MP0, or raising it, did not quell the issue only change it's character. going lower made it more stall prone. going higher made it hunt endlessly. So I put it back to stock. reloaded it back a couple version and calling it good for dyno day tomorrow. I have no idea what to try next! I'm curious if the throttle closure rate is something to try?
There is a switch for desired air mass at idle, you can try disabling that. you could also try playing with the values in the speed density, mainly MP0 offset and slopes. Lowering the offset and raising the slopes should tell it less airmass. Looks like the base tune gives you a general idea of what to do with the tables when compared to the stock tables of other NA engines and when comparing MP 0 to MP OP.
Well MAF/RPM= IBm/rev. Offset is pressure that didn't make it into the cylinder. Slope is pressure that made it in. Blow through is pressure that was lost to over lap. Quadratic is pressure that was not accounted for in slope due to compression of air. When you add all these up you get the manifold pressure before intake valve opened. If you multiply slope and blow through by (MAF/RPM) and Quadratic by (MAF/RPM)(MAF/RPM) then add up all the values you get your MAP. So (.55/700)=.000785 and (.55/700)(.55/700)= .0000000616. I'll finish when I get home to get the values of the speed density.
As you can see it is hitting the upper limit of calculated MAP. I think by changing the calculated Map Max value your speed density is thrown off. I would return that to 31.88 let the clyair multiplier do its thing and see what it does.
I follow your math there. So the concept is that if we lower the map max back down to 31.88 and leave the cylair multiplier at 1.99/1.99 it will accomodate the boost without having to recalculate the entire speed density table... which I support since none of the files I have as examples touched any of the SD files.
now here's some interesting info. both of my reference files both bumped maximum MAP to 60 & 64 respectively. And both have cyl air multiplier at 1.99/1.99. hmmm. does it make more sense to raise the ceiling a lot or try back at stock?
be careful with the procharger_bad_idle file if you snag it for your own uses... the MAF transfer was horrid among other issues. it would rev to the moon and not come back down.
By ceiling i assume you mean the 1.99 multiplier. YOu can leave it at 1.99 or lower it slowly until you see a difference in logged Map. the multiplier is the easy way to tune the SD. The draw back is you lose transient performance as the ECU figures it out. Most people can live with that and have no complaints.
Reply With Quote