NefMoto

I just modified my tune to finally take advantage of my 52lb injectors and turn up the wick with regards to boost.  I adjusted all of the boost maps that I felt were necessary (simplifying corrections such as IAT, broading LDR limits, etc) and yet boost spikes to maybe 18psi falling almost immediately to 7-9psi and then holding to redline.  I am at a loss as to what could be causing it. 

Please take a look and see if you can catch anything that may be causing this. From everything I can gather boost should be hitting around 23psi as dictated by LDRXN, and KFMIOP should be modified so that there are no unnecessary torque interventions. 

Mods:
K04's
3" Downpipes with no Cats
SSAC catback
034 motorsports 85mm maf housing
Hitachi sensor
K&N Panel filter
710N DV's
52lb EV14 injectors


Any help?

Logs please...

I sent you that ecu. lol.

Haha, did you really?  It just showed up today and fired up great just as I had hoped my tune would.  I tweaked fueling a bit to get trims in check, but boost ultimately is not where it needs to be.  SUCKS, as I was desperately trying to take it to the track tonight for a big alumni weekend and show up a lot of friends. 

I don't have any logs currently other than VCDS, which has an ungodly terrible sample rate.

Don't bother doing any tuning until you have a proper logging setup.

Much easier to see what is going on than to dig through our bin and guess

Yes, of course! Digging a bin file maybe possibly useless unless you see whats going on with the variables while driving.

I will get on it today.  What parameters would you suggest would best help capture what may be going on?

I hope you don't mind, but I looked at your file. KFLDHBN needs to be increased. I'd bet that's holding you back.

FYI, I don't think that that file's safe at all. If you ask me to critique it I will.

Feel free to critique, I'll never learn unless I get feedback from those who have done this before.

BTW, you may be on to something with KFLDHBN, those values are completely stock, which would be extremely limiting for sure.

I have tuned MANY ecu's over the years, the ME7 just is another creature in and of itself for sure.  It's been a steep learning curve trying to wrap my head around a different way of thinking.  Please feel free to let me know your thoughts, I have fairly thick skin and understand that I left many things more widened than they could be when optimized.

I'd pull about 6 degrees of timing at high load/rpm points, I forsee MASSIVE cf's if not. Start conservative, dial it in. That, and unless the maf is underscaled I forsee limp. I'll look the rest over shortly!

Logs please.

I'll try to take it out and get some actual ME7 logger logs after work today.  FWIW KFLDHBN was the trick, I need to dial it back a fair bit and begin slowly making it more aggressive. 

Took a quick log today.  Boost is hitting just as I had requested, no limp mode, MAF maxes around 344g/s, with knock values reaching around -6 degrees at 5000rpm.

Take a look and tell me what you think.

requesting too much boost too soon; you might throw positive deviation

you are over boosting past MAP limit .. probably need to adjust i-limit

your requested AFR is ridiculously rich .. adjust BTS

you are requesting too much timing up top. as boost comes down, that might fix itself.

Interesting, so ME7 will throw positive deviations during low boost if I was flat out requesting full boost too early?  I didn't realize that.  Adjust KFMIRL downwards from 4k and below?  Thanks for the insight regarding AFR as well. I haven't had a chance to convert it over myself to AFR so it's interesting to see just how rich it was requesting.

Yea. just request peak boost a bit later. 3250-3500 should be fine. You don't have to go nuts.

Oh, also, add wastegate duty to your logs.

and might want to log spool->peak boost in 6th as well... you might have to adjust KFMIRL at lower rpms.. but for obvious reasons you might want to avoid redline in 6th :)

This is a TIP car so 6th isn't happening.

Before I blindly play with KFLDIMX, what do you suggest doing to the values:

4000=79.8
4500=82.3
5000=88.120
5500=93.72
6000=96.98
6500=97.8

All in the 1000 column. I don't know what the units are, nor what to expect with changes (do I go up or down?  Hard to tell since I don't know the units).

5th. whatever. A tall gear. btw you should do a 6sp swap, your tranny isn't going to like this tune.


Units are in %wgdc

Which is why you need to log wgdc :P

Well I adjusted a few things and took it back out.  I am really overboosting, so even though I adjusted the I-limit by +5 I believe I am still way off. WGDC is logged, take a look. I dropped timing pretty heavily and knock really seems to be under control now.

Something else that I am noticing.  A VERY audible loud whistle (not the turbo) at say 5k+. It doesn't happen everytime, and it is very reminiscent of a slightly leaking coupling.  I'm wondering if it's possible that the overboosting is blowing open my 710N's somehow? I'm seeing spikes of 26-27psi (!!) on the boost gauge, so I need to get the boost PID under control ASAP.

Bring down your I-limit.

A lot :)

Also, your wgdc should be 95% up until JUST before request meets actual.

Something is going very wrong in your PID right now.

Haha, see I told you that I had no reference with which to judge how to adjust it :).  FWIW, as I'm sure you know, depending on the way the valve is setup bleed off can happen at 100%, or at 0%. Therefore I wasn't sure how the S4 N75 valve is setup.  Rough guess, how much does it need to be brought down?  Also are we talking about only during spool conditions in the 1000 column?  What units are the columns in? mBar deviation from target?

5-10% probably. Hard to tell w/o knowing what your wastegates are doing... and you are hitting peak boost VERY late for K04s (you are running k04s, right?)


I believe it is plsolr, which is requested *gauge* boost (plsol - plgrus), i.e. requested absolute pressure - ambient pressure.

Really, there is something screwy with your wgdc... how many PID changes did you already make?

There are two things to that comment.  #1 I am a TIP car, and have a slightly higher stall torque converter installed.  This tends to trick people who are used to looking at logs of manuals as the converter will slip the rpm higher during turbo spool, making it look like it's more laggy.  The other possibility is that I really do have some sort of leak taking place (see my comment about a very audible noise), but I will verify that with a boost leak test asap.

Since the last log the only PID change I made was to the 1000 column of the Imax, and even then I only upped it by 5. Other changes are:

KFLDRL
KFLDRQ2

Attached is my file along with a stock L box BIN for comparison.

I am chasing down the same type of issue, except that mine is not loud (I'm only boosting to 19-20 psig).  Found a couple of small leaks, including the MAP sensor where it seals to the mounting bracket, but it persists off and on (you might check there).  One possibility I thought of is a slipping accessory drive belt; sometimes it sounds like that.  I'd love to know if you figure it out.

I would stick with chasing down small leaks, the accessory belt shouldn't be prone to slipping at high RPM, but at low RPM

Having worked with quite a few turbo'd setups I would say it is definitely boost related somehow and only shows up at 25ish psi, as though something is finally loosening up. 

Nyet, care to give me some insight into your comment regarding the odd WGDC curve?

Attached is a log from tonight that is 10% reduced I-limit from 3k and up.

I think i steered you wrong about requested boost. Might want to bring it up earlier... the guidance I gave you will only cause your wgdc to start to go low too soon, which will cost you spool. Move it back to hit peak rlmax at 2500 rpm.

Your wgdc should be 95% until right before peak boost... something like this...

Funny, I think you came to the same conclusion that I did.  I was looking at WGDC and noticed some slight fluctuations early on, and just changed my map at low boost to target 95% accordingly.  Since we're dealing with the I term of the PID feedback, it really only needs to be adjusted DURING spool, as that is where the maximum deviation takes place. 

How does everything else look?

Before boost hits req, the biggest term is P, by far. The I contribution is the steady state term, and moves very slowly. It forms more or less your target wgdc, since D and P should be near zero when lde is near zero.

You can log each of them btw....

lditv_w           ;{BoostPID-I-Result}                ; {LDR, Tastverhältnis vom I-Regler (word)}
ldptv             ;{BoostPID-P-Result}                ; {LDR, Tastverhältnis vom P-Regler}
ldrdtv            ;{BoostPID-D-Result}                ; {LDR Tastverhältnis vom D-Regler}

That will help dial in lag, thanks for the insight, I'll see where it stands currently.