Rough data log

Here's my third gear pull datalog. Obviously that's a lot of knock. Anyone have any thoughts on what might be a hardware cause for this? I've contacted my tuner (Tuning Alliance) to see what their thoughts are, but I'm thinking something's physically wrong with the car. Anyone have any thoughts? I've replaced a ton of my vac lines, none of which had any obvious leaks. I've had the car boost leak tested, and nothing was found. I'm just at a loss at this point.

 

Stage 2 lgt on a VF40, 5 speed manual.

Grimmspeed 3port ebcs, k&n intake, AEM 320lph fuel pump, grimmspeed up pipe, invidia catted down pipe, bulletproofed intercooler, and that's all I can remember for now.

3rd gear.csv

https://log.malonetuning.com/chart/252946#H8KLCAAgworCtlwCA03CjEsOw4MgEEPDrzJrI8KFJsO9w5wFwrHCoMONwogiwqUQw4EkSnvDugQqVV3Duj3Dm8OGw5Ajwr3DriHCssOyOS1zIXQWFU4pwocPHxHDpDlydsOCw7Q1ccOlLErCksOawpTDm0JdQMO3VcK5cVTDrx%2FCoxzDvFPDmsKmwrDDvAnDk8Khw4fCgDMuwrjDogZ9woLDlhbCpFvDlS3CksOKPAVpwrfCg8K1O17Do8OHw6HCogAAAA%3D%3D

 

Link to view it above. Once I get to my computer I'll take a look and comment.

 

Edit:

 

Okay, so there was only one knock event in that pull (time 4.58). Timing isn't very aggressive there, perhaps it needs a touch more fuel there. Hard to tell without a proper wideband O2 sensor (if you don't have one, it would be very useful to near required for tuning here, and I'm sure Mike would appreciate having it). The high A/F learning number also indicates that the ECU wants more fuel. The fact that the ECU has a moderately high positive A/F learning D value is something that is often undesired, as it screws with open-loop fueling. This can be changed by modifying the A/F learning ranges to either change the D range to be elsewhere or to disable the A/F learning D range entirely. The fact that the learning value is moderate means the MAF scale might be off. Since you have a K&N intake (assuming you mean the cone-filter one, not the drop-in panel-filter in the OEM box), that's not surprising. Without a wideband, it's difficult to tune for that, though.

 

To be honest, the log doesn't look that bad at all. That's not "a lot of knock" in my opinion. The learned knock is learning itself away it seems, so I wouldn't be too worried about it. Nothing looks to indicate any hardware issues. I understand that Mike is typically pretty busy, but he's a capable tuner and I'd trust his judgment on your logs as well. Personally I wouldn't log things like "Primary Ignition Timing" at all, and if you're just logging WOT pulls, I'd skip logging throttle position too. Wastegate duty could also be skipped if you're not tuning the boost tables. Of course, follow Mike's guidance for what parameters to log, though.

Thanks for the response. One note to make is that I do have a K and N panel filter, not the intake.

 

Is the high learning indicative of a vacuum leak that I still haven’t found, or is it more likely to be a MAF issue.

 

I’m thinking you’re right on the wide band for sure. AEM or is there another I should look at?

High learning in the A/B range can typically indicate a vacuum leak ASSUMING the MAF calibration is correct, the MAF sensor is good, and the primary O2 sensor is good. Also worth checking how your fuel trims (A/F Correction) change in real-time when wiggling the MAF sensor connector, they are known to give problems (see here). This should be easy enough to just plug in your AP and keep it visible in the engine bay while the car's idling and you're wiggling the MAF sensor connector.

 

At higher load ranges (A/F C and D, as well as open-loop), a minor vacuum leak won't make much difference in your fuel trim, so I'd expect learning values in those ranges to indicate incorrect MAF scaling. Keep in mind, though, that the learning values are indicative but not sufficient to assume tune or mechanical issues. If your MAF calibration is perfect and done on a day with 60degF outdoor temps, but suddenly the temperature is lower and your temperature compensation, atmospheric pressure, etc., tables in the tune aren't perfect, then the learning value is there to account for that. But if the tune was done in similar conditions to what you're driving in, high learning values (either positive or negative) indicates either mechanical changes since the tune was done, or an incorrect tune.

What would you consider High in A/B Range?

What would you consider High in A/B Range?

 

 

Kinda subjective. I personally think anything within +/- 10% is fine, but ideally would like to keep it to +/- 5% or so.

 

 

In steady state, as long as your A/F correction isn't pegged at its limits (by default I believe that's +/- 25%), then your fueling in closed-loop is on point. However, high learning values can affect the transient response of the engine, as the ECU changes between learning ranges smoothly instead of instantaneously.

My learning at idle is within the 5%. So I'm hoping it's a MAF scaling issue. Here are some idle and cuising logs to add context. I really appreciate you taking a look at this.

 

Mike is a busy guy, and I don't really like bothering him if it's just something hardware related that I'm overlooking. I hope no one on this forum takes this as me dogging Mike in some way, he's a great tuner, and knows his stuff.

cruising.csv

idle.csv

Interesting. I'm seeing quite the opposite from idle to driving. I'll change my monitors a bit and get a full A/F look. When on the pipe it's 0. Learning A is the high one at +10 ish, but as soon as I apply throttle it drops to .78 or 0.

My learning at idle is within the 5%. So I'm hoping it's a MAF scaling issue. Here are some idle and cuising logs to add context. I really appreciate you taking a look at this.

 

Mike is a busy guy, and I don't really like bothering him if it's just something hardware related that I'm overlooking. I hope no one on this forum takes this as me dogging Mike in some way, he's a great tuner, and knows his stuff.

 

Cruising: https://log.malonetuning.com/chart/253076#H8KLCADDscKuwrdcAgNNwoxBDsOCIBREw6%2FDssOXQ1IsPQ1hwoHDtgdJKjTDsGvCqsKnFzAxLmfDnsK8wrHClm7DuXHCjcKJVSjDucOYK2Fyw6jDpcKWS3xzwovCoMOAwonCixfCpi9JTy7CoiTCq1PDuTN2A3rDrsOIwq%2FCq3rDvToqMcOcZTjClcOlD8OYwrbDh8KlOcOQBjMMFmjDrUB6bMO9IcK5w65bwpTDsWvCnMO7AGFRw4lDwqMAAAA%3D

 

Idle: https://log.malonetuning.com/chart/253077#H8KLCADCssKvwrdcAgN9w5BNDsKCMBAFw6DCu3Q9TRjDvj0LYVFlwoQmSEkZDHp6WzTDhsKFw4PCssOvw6vCtMKZw5c0w6rDom5nO8KRw67CvVvDp0VBw5JCDEfDp8OtwpPDghFUTxN5w4PCpMOeMsOdw4nCs2bCpzdtNhsnAMKzSMKmw6vDtMOjwpspb8O7woHDt8KZwoXDuAfCmnAfw5IwA8KYQw4FwpRQQQ0YwqIgw5jCgsOCfcOKwqzDrMKWecK0wrzDv8KQw4dowrAdw6nDjsKwCS%2FDhjQBw7Yrw70TFCUVJRMlF8KlEMKlEsKlFMKlFsOlJG96UMKCw5wCw4orwqHDnAMewpV3NcOjIjQuEn7CqH0Bw4ddRSHCiQIAAA%3D%3D

 

From your cruise log, you can see that the ECU in general has learned to add too much fuel in the C (about 8.6% A/F learning) and D (about 7% A/F learning) ranges. The learning is positive, but the correction is negative and relatively large at times. In your steady-state C range it's oscillating between around -5 and -10%. Ideally it should oscillate around 0% correction (the oscillation is due to the rear O2 sensor attempting to keep the catalytic converter operating efficiently; this can be disabled in the tune). It's possible your knock is also caused by rich knock, as well. If your MAF scaling, sensor, and O2 sensor is all good (so you're hitting your fueling targets), then it's possible that there's just too much fueling in certain areas.

 

Again, I don't see too much issue with your logs. If it keeps knocking in the area your first log shows some knock, then maybe I'd look into tweaking the tune there, but otherwise it looks pretty good. As usual, make sure your tuner thinks the same

 

FYI, you can check your fuel trims when you're parked and the engine off. Just turn the key to on, and then display A/F learning A, B, C, and D on your AP. That will show you your 4 different fuel trims. Part of the "learning view" that people talk about in the opensource world.

Yeah the issue is drivability, tip in feels jerky and so does wide open throttle. It’s not a smooth acceleration, and can feel pretty disruptive.

 

Mike has already gotten back with me. We’re kind of wondering if the 13 year old MAF might be on its way out. Either that or there’s still a leak in my intake track that I haven’t figured out.

Old MAF could definitely do it. If you haven't already cleaned it and checked the connector, I'd try that.

Cleaned the MAF, no discernable change.

 

Going to do another intake leak test this weekend. If nothing is found then I'll pick up a new MAF, or maybe swap the MAF from my wife's Tribeca if they're compatible. Shrug.

 

I'm kind of wondering if there's a sneaky torn inlet leak that only manifests under significant amounts of boost?

I'm kind of wondering if there's a sneaky torn inlet leak that only manifests under significant amounts of boost?

 

 

The inlet shouldn't see boost pressure, that only appears post-compressor (so intercooler or after). If you didn't catch anything with a smoke test or pressurized intake test, then I wouldn't suspect any vacuum leaks. I'm assuming your boost leak tests would have covered this.

That bit me a while back on my first intake. Just a wee leak at the turbo side after 120K or so.

May just be the tune cause I can't figure why coasting or at idle learning A will be at 8 to 10%, however learning and correction report a solid 0 to -.78 or .78% while driving.

Off the pipe and coasting correction shoots back up to 10+%.

The inlet shouldn't see boost pressure, that only appears post-compressor (so intercooler or after). If you didn't catch anything with a smoke test or pressurized intake test, then I wouldn't suspect any vacuum leaks. I'm assuming your boost leak tests would have covered this.

 

Smoke test didn't catch anything other than a small leak at the ebcs. I have since replaced all of my ebcs lines with new ones, and new clamps, so I assume that one has resolved, but perhaps it hasn't. I haven't smoke tested since the changes. My trusted mechanic has a long wait list, and I don't really want to go back just for a smoke test.

 

I'll do the poor man's boost leak test (plug intake with paint can, and blow on the recirc hose) this weekend and see if it holds pressure. If it does then I suppose it's time for a new MAF unless you think there's any other diagnostic option?

I was looking for a test of the MAF sensor and gave up. Just bought a new once since I had +200K on mine and several cleanings since 2010.

Funny I never paid attention to the car other than OCI and gas for the first 4 to 5 years owning the car.

Now wonder if I'm too hyper sensitive looking for issues. But getting ripped off by the stealership will do that.

I have a MAF sensor sitting in my intake that's not being used as I'm now running speed density. I'll send it to someone for the price of a MAF block off plate.

I'll buy it if my boost leak test this weekend reveals no boost leaks. How much is a block off plate?

I'll buy it if my boost leak test this weekend reveals no boost leaks. How much is a block off plate?

 

More expensive than I remember. I was figuring $20 but they are more like $50.

 

I'll take $20.

I'll buy it if my boost leak test this weekend reveals no boost leaks. How much is a block off plate?

 

Denso 197-6040 Mass Air Flow Sensor will work. At least it did on my Spec B. Surprisingly it has gone up in price a little over $20 since I bought it. (March 29th)

 

https://www.amazon.com/gp/product/B007WMG8A0/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1

doing more research on this, my spark plugs are pretty old, do you think these logs are showing symptoms of spark plugs on their way out? I think they have about 60k miles on them. I'm going to change them either way, just not sure if the jerky acceleration could possibly be a result of this.

doing more research on this, my spark plugs are pretty old, do you think these logs are showing symptoms of spark plugs on their way out? I think they have about 60k miles on them. I'm going to change them either way, just not sure if the jerky acceleration could possibly be a result of this.

 

 

Log "Roughness Count" on each cylinder. If it's not misfiring, I doubt it's spark plugs.