How's my LV?

[iNVAR]

FOOOOOOOOOOOOOORK. It's been about 2.5 weeks since I last checked my LV. And here's what it looks like today! WTF....

 

That's not good. Temperature's been rather cool here lately too.

 

Can running really, really low on oil do this? Not to the point where the oil light comes on or anything but nothing showed on the dipstick yesterday morning, oops. Oil pressure had been fine though according to my gauge (sensor at sandwich adapter)

 

I'm going to try switching to a different gas station too.

[Spec B]

Sucks...do a few logs. That looks like bad gas to me...knock everwhere.

[iNVAR]

Yeah, wish I had been bringing my laptop around with me and logging every day, argh.

[bmx045]

looks like your car dropped a deuce

[NSFW]

I doubt that low oil could do that, but I have noticed that the ECU is much more likely to use IAM shortly after a reflash or reset. So if you did a reset before that LV, then that could be it.

 

Also, that an IAM of 0.875 could just mean that the ECU pulled timing via IAM once, down to 0.75 (the default IAM step value is 0.25), then added timing to 0.875 (when the IAM changes direction, the step value is cut in half, so 0.125). I've seen that happen in a single pull shortly after a reset.

 

So it's worth investigating, but it's not necessarily terrible news.

[iNVAR]

I hadn't done a reset for about 2 weeks. I'm concerned about that knock still.. IAM's sitting at 0.875 and there's still red in that cell I've been working on. I need to pull some logs and see what the deal is.

 

Perhaps the cooler temperatures are permitting higher loads now and it's knocking in a cell I didn't pull enough timing from hrm...

[Ridgeracer]

I would say look at the IAT timing comp table.

 

[Table2D]

-40 -22 -4 14 32 50 68 86 104 122 140 158 176 194 212 230

2.11 2.11 2.11 1.05 1.05 1.05 0.00 -1.05 -2.11 -3.16 -3.87 -4.92 -5.98 -5.98 -5.98 -5.98

 

Mind you I run 93 oct with a FMIC and HTA68 but I heat soak a lot with the AVO intake.

[iNVAR]

Mine's not too far off from yours.

 

[Table2D]
-40     -22     -4      14      32      50      68      86      104      122      140      158      176      194      212      230
3.16    3.16    3.16    3.16    2.11    1.05    0.00    0.00    -2.11    -3.16    -3.87    -4.92    -5.98    -5.98    -5.98    -5.98

So it looks like my 86 degree cell is the one that's off right now. The cold side is also more aggressively advancing the timing.

[black318i]

I had a ton of knock with my tune as well. I compared it to the stock map and discovered it was running like 8 degrees more than stock IIRC in the 1.5 4k area. I had to remove 6 degrees to make my engine happy. Also maybe I missed something, but your timing tables do not look very smooth.

[iNVAR]

Okay, went out for a log, what in the fork happened here? I have other issues too, like knock EVERYWHERE, but this one is confusing the hell out of me.

 

Total timing goes from 20 degrees to -14 to -8 then 8 and back up, but no indication of knock... I definitely FELT the car lurch when it did this...

 

My current timing and advance tables.

http://legacygt.com/forums/attachment.php?attachmentid=104431&d=1314306035

http://legacygt.com/forums/attachment.php?attachmentid=104211&d=1313783748

 

reattached LOG file with headings this time, oops.

wtf.csv

[Spec B]

Wtf indeed - I have never seen that before. I don't get it...are you sure you are running the right map? I don't understand if why your timing would do that without any learned or feedback knock correction.

[iNVAR]

Yeah... I thought maybe the cable was reading wrongly or something, but I definitely FELT the lurch in my car, which is why I went back to look at the log for the exact moment!

 

I am scratching my head like crazy here. The LV was clear, and like you said, don't see any FLKC or FBKC and according to the 2 tables, that should not have happened.

 

I'm certain I'm looking at the right map.

 

:confused::confused:

 

Are there any other tables I should be looking at that would affect timing that significantly? I did make a change to one cell in the IAT compensation to subtract -1.05 only from 86 temperature but that's it. Initial WG Duty has also been reduced to 0 for everything below 36% throttle.

 

edit: Actually, now that I look at the log and RPMs again, maybe that's not when the car lurched. Hmm... So maybe there was just a blip in things? Just an anomaly?

[NSFW]

Random numbers do appear in logs from time to time. You got two weird numbers in a row, which makes me think it may not be noise, but then again I don't have any good theories on what would cause timing to change 30 degrees so suddenly. I dunno.

 

But you should look at your total timing using this tool:

http://www.romraider.com/forum/viewtopic.php?f=32&t=4439

 

I have a hunch that you'll find that it's pretty lumpy, and smoothing it out might make the car smoother and/or stronger in general.

 

The 1.65 - 2.25 columns caught my eye in particular. When you pass 4000 RPM, the advance table drops 5 degrees, but the base table is smooth, so the net effect is that you drop 5 degrees between 3600 and 4000 RPM.

[iNVAR]

Yeah, I did notice that... I didn't touch the advance table, that was from Infamous. All I really changed was the base table. Not sure what to do about this, hmm. What reason would someone have to limit the advance table to anything other than 8? I mentioned doing it earlier, just putting 8s across the board, and working entirely on the base table only. Any disadvantage for that?

 

Thanks for the link, lemme look at that too.

 

edit:

You can chnage everything to 8 and adjust accordingly on the base map side. That is how I have my map setup. But i would not do it in the cells below 2000 RPM - because if IAM ever drops to 0.5, your timing will be negative near idle. Not exactly sure what would happen in that case.

Other than that I mean. I'll adjust accordingly. Any other reason NOT to put in 8s?

 

And also, another question. When in the world would my car (or anyone's car) ever see a load of 2.0+ with the engine RPM under 1000? When would that part of the table even come in? Because I see lots of negative base timing in that corner.

[NSFW]

It's my understanding that the advance table exists to allow the ECU pull the right amount of timing across-the-board to run on lower octane fuel. So basically you should be OK to run 87 octane on the base timing (advance table set to zero, or IAM set to zero - same thing really) and 91-93 (whatever you have locally) on base+advance. It's easier to induce knock at midrange RPM and high load, so the factor advance table has its highest numbers there, and lower numbers at lower loads, and lower numbers at high RPM. I haven't actually tested the 87 octane theory, but I think that's what Subaru's tuners had in mind.

 

Lots of tuners just fill it with one value, so that you can see knock just by logging the dynamic advance parameter. If it's less than 8 (or whatever the table was filled with) then you're looking at FLKC or FBKC pulling timing. That approach is less common for cars that support Knock Sum, and I expect to see it get even less common now that RomRaider's "fast polling" mode lets you log IAM+FLKC+FBKC and still get 3 times as many rows of data than we used to get.

 

My guess is that he was pulling timing from the advance table and leaving the base table as-in. That would explain the shape of it. I prefer to set the advance table just once (shaped a lot like the stock advance table) and do everything in the base table. As long as IAM stays around 1.0, it doesn't matter.

 

I don't think the high-load / load-RPM areas of the table is reachable. But if you ever did manage to get there (I have no idea how), you'd want low timing, since high load and low RPM both tend to be more prone to knock.

[iNVAR]

Thanks NSFW. BTW, I grabbed your Stage 2 map off RR, as well as MickeyD's. I need to read up some more and learn everything, but I have an exam coming up that I need to study for first, so I have to put things on hold. This is really time consuming!

 

Also, more importantly, tonight the outside temperature was about 55-60F and I had a chance to go out somewhere late tonight and do my 3rd gear pulls, repeatedly. I probably made somewhere on the order of 10 pulls, stopping each time to make adjustments.

 

This... can't be right, perhaps it's just some crap fuel, and I shouldn't have been messing with the timing? I had to keep pulling timing from the high load/high RPM area! Look where I'm at compared to where I used to me when I got infamoustuned.

 

This is a comparison of my current base timing map compared to the original stage 2 that I got from Infamous approximately 1 year ago. Temperatures outside were the same, car mods were the same.

 

What factors, other than gas, would I be looking at that might cause all this knock? I did change my air filter to an AVO washable one, but from what I understand that shouldn't affect the scaling really.

 

Spark plugs? I changed them about... 27K miles ago, the stock iridium ones.

 

Anything else ? I mean... this is pretty disturbing to me. Should I be worried? Also, haven't had a change to smooth the entire thing out using the timing editor. Thanks all for the help/input.

 

edit: Did some reading on the RR forums, so based on what you told me NSFW and some others on the RR forums, I think I understand the point of the Advance Table now. The 8s (or 6s) are in those specific cells because those are the "important" (and prone to knocking) cells to pull timing from when the ECU decides to drop IAM. So while I could just pop 8s everywhere, that would mean that if IAM dropped, I'd be pulling a lot of timing from other areas unnecessarily. On the other hand, I don't really care if I pull more timing than necessary. If IAM drops, it means I better fix my map. So I think I'll plug the 8s in a larger area and make things easier to adjust.

[iNVAR]

YUCK, here's my new advance table and base timing. The base timing looks horrible, not smooth at all!

 

Thanks again for the link to that timing editor NSFW. It made re-doing the base/advance tables much easier.

 

Now, any input? Anyone? Bueller? Bueller? If I want to smooth this out properly, that would mean I have to yank a lot of timing from the cells directly above and to the right of the lowest cell areas, no? Sigh.

[Ridgeracer]

Lets try something different here. Your LV from page 1 shows that D range is pulling close to 3% fuel off the 40+ g/s range...I have a hunch this is all related to a MAF scale that is really off up top. Unfortunately it's a shot in the dark since you don't have a WBO2 sensor. Can you replace the AVO panel filter with the stock one and just drive around in CL and lets see how much the trims change..?

 

I only mention this approach because I always remember Mickeyd being very adamant about aftermarket panel filters or removing the intake plenum changing the MAF scaling.

[iNVAR]

Okay, I'll have to order a new stock OEM filter, I ran out of them and never got around to having a few spares around. I should probably clean the sensor too huh?

[Ridgeracer]

Couldn't hurt but yes get an OEM filter and see what happens. This is all assuming you where originally tuned with the OEM filter and resonator in place.

[iNVAR]

Hello hello all. Thought I'd post an update. I've been logging every single day and have learned quite a bit and also noticed quite a bit.

 

1st, with regards to my knocking issue, I'm still using the AVO panel filter but have 2 stock filters on hand. I tweaked one cell of the IAT ignition comp table to pull 1 degree from 86F. It seems to have helped. I never smoothed out the total timing because it looks like in order to smooth things out, I'd have to pull a LOT of timing from all over. I'm gonna leave it as is for now... probably slightly suboptimal.

 

Second, we had a cold week last week, around 50-60F and I threw a P0244 code while pulling hard in 4th gear. Luckily, I was logging at the time and found the overboost. It's spiking hard to 19.5-20 (target is 18!) in 4th gear at 3200-3600 when I mash the pedal. I unfortunately didn't log TD values. Intake temp was 70F.

 

Here's a sample if interested:

www.crimetank.com/misc/overboost-cutoff.csv

 

Ridgeracer suggested that I change the WG IAT comp table, which I have done by subtracting 2.3% WGDC at 68F. Strangely enough, everything above that is completely zeroed out. I don't like that, but I can't really tune that area right now being that it's cool. Maybe next year.

 

Then yesterday, I pulled another log:

www.crimetank.com/misc/spike-and-maxed-out-td.csv

I did log TD this time and you can see where the spike is. It seems that TD integral is not winding down quickly enough once target boost is achieved, hence the slight overboost, which is not bad. I guess there are THREE ways to handle this:

1) adjust TD Integral so it winds down faster OR

2) adjust TD Proportional so that it instantly lops off some WG based on the overboost

3) or I could just leave it alone. I know 0.5PSI isn't a bad spike but I think I'd like to fix that anyway, or try to minimize it.

 

What do you guys suggest doing? I think I'm leaning towards adjusting TD Integral but not sure if that's the appropriate way to do it.

 

Also, after 3600 RPM, WGDC is capped out around 80-81% because TD Integral is unable to go above 15% (I fixed that. It's 20% now, hopefully that helps) but again, is that the right way to do it, or should I adjust TD Proportional to add/subtract WG more aggressively? It seems like having TD Proportional too aggressive would cause oscillations whereas TD Integral seems to be a "smoother" way of controlling things albeit "slower."

 

This is quite a learning experience... and burning up a lot of gas too. Any pointers?

 

Also, I think maybe the last LV was actually bad gas, or maybe RidgeRacer's suggestion of -1 degree for 86F really helped a lot, because 3 straight days of non-stop logging and about 75 miles mixed city/highway in various temperatures and my LV is still completely clean!

[black318i]

I had a same issues and I worked everythig out with TD Int. cumulative, Int. WGDC, WGD Comp. IAT, and removing a lot of from the Base timing table.

[iNVAR]

Figured out why I got the P0244 error code. It was the cold weather coupled with the fact that my Init WGDC tables were mismatched to my Boost Targets. TD Integral and Proportional were both off the charts trying to compensate, capped out at 30% (Init WGDC was 0%, so 30% was the ceiling for compensation) trying to produce 10 lbs of boost. NOT happening. So when I mashed the gas pedal, boost built up quickly and Init WGDC went up to 65%, plus another 30% from TD, which capped me out at 90% based on the Max WGDC hard ceiling. When I hit target boost finally, TD integral was unable to wind down quickly enough so I was still at very high WGDC. TD proportional tried to remove some, but failed miserably. And BAM, P0244 code!

 

Have since fixed my Init WGDC tables and Boost Targets.

 

Can someone answer this for me though: How do tuners determine what boost targets to run at? My peak boost is at 18PSI. But I have seen both less and more. How is it decided what to run? Like if you had no baseline to start from (i.e. we know Stage 2 LGT's go between 16.5 to 19.5 PSI peak, depending on the tuner), how would you figure it out blindly?

[black318i]

Diminishing return as boost is dialed up. Also you shouldn't have any significant boost issues caused by temperature if everything is dialed in.

[iNVAR]

Yeah, I realized that after figuring out and understanding how all the tables work. If TD Integral and Proportional are properly calibrated, then the temperature shouldn't affect too much... Even the Init WG IAT comp table is just a baseline/estimate to open up the WG... all the work appears to be done by TD Integral and Proportional.