3.12 load at 750rpm?

Hi Guys,

 

I drive a 2.0 Lib GT, semi-closed deck, VF38.

 

I've just done the spark plugs as well as oil, trans fluid and all the other good stuff so thought I'd do a bit of logging..

 

Noticed a VERY strange knock event and have no clue what could cause it.

 

If anyone has any ideas, I'd love to hear them!

 

I've attached a small log of the event.

 

Also.. I see my total timing slam down to ~2 degs at gear changes..as well as knock events.

 

Does anyone else see this in their logs?

 

Edit:- I see total timing go into negative values sometimes, when I hit +2 load, at lower revs (2500-3500) and full boost (WOT). No knock..

 

Confusing the shit out of me. O.o

 

Car feels great though and Learning view shows no timing being pulled and an IAM of 1.

romraiderlog_worry.csv

What gear was the start of this log done in?

 

You need to add Throttle Plate Opening Angle into your logging parameters, and if you're going to do an extended log like this, you should add in Gear.

 

Regarding the timing, post a log of that phenomenon happening and we can tell you if it's "normal" or not.

you need to smooth the wastegate tables better, duty cycle is being applied rather abruptly. see the 0-0-0-0-0-90-90-90-90-90 and where the load spikes? also look at maf g/s and how it jumps abruptly...

Just realized he has accelerator pedal angle logged. It's not as good as throttle plate opening angle, but it'll do for now.

 

Without looking at his map, you can't say what needs to be done. His WGDC goes up that rapidly because he goes from 0 to 3.53 to 28.63% in the course of 112ms. He's basically going from being completely idle to adding a lot more throttle. Of course the numbers are going to shoot up a bit. But not sure why it'd shoot up THAT much.

 

So, I'll ask again.... "What gear was the start of this log done in?"

you need to smooth the wastegate tables better, duty cycle is being applied rather abruptly. see the 0-0-0-0-0-90-90-90-90-90 and where the load spikes? also look at maf g/s and how it jumps abruptly...

 

Thanks for having a look! Yea, those wastegate duty cycles are the stock values, from the stock rom. I'm running a twinscroll and it whines like a mofo right of the line. Seems I cant attached my ROM.. only images?

 

 

Just realized he has accelerator pedal angle logged. It's not as good as throttle plate opening angle, but it'll do for now.

 

Without looking at his map, you can't say what needs to be done. His WGDC goes up that rapidly because he goes from 0 to 3.53 to 28.63% in the course of 112ms. He's basically going from being completely idle to adding a lot more throttle. Of course the numbers are going to shoot up a bit. But not sure why it'd shoot up THAT much.

 

So, I'll ask again.... "What gear was the start of this log done in?"

 

The very start of the log is starting from a standstill in 1st gear. So that 3+ load is when just getting off the line. I've done a bit more logging and see it alot.. the load and MAF reading goes very high for a split second when taking off. It doesnt feel/sound like that when actually taking off though. Sounds smoooooth. That part of the ROM is all factory too.. I've only tweeked the higher load/rev areas of timing etc.

 

It certainly does appear to happen when that WGDT jumps to 90 right off the bat.. but that's the way it was from factory.. maybe I can smooth that out a little and see what happens?

first of all, is the car completely stock? what mods are done to it? i'm reading that the 03-06 2.0GT liberty came with the vf38 stock so I'm assuming the turbo is stock? if not then clearly you need a tune, I.E. smoothing out the wgdc, lower lowend timing a bit, fueling

 

the twinscroll will certainly cause a quick surge if the wgdc jumps all the way up abruptly. remember how load is calculated- maf g/s x 60 / RPM. that means, if the wastegate is closing, the turbo will spool, therefore more air will be pulled through the intake maf, thus causing a higher g/s reading, us guys with a single scroll aren't used to this instant spool . but it is unusual to shoot across the timing table to the right so quickly(as you can see by the tot timing being so low at that point), especially at only ~33% throttle.

 

what to do next is determined by your answer as to whether the car is completely stock, or has been modified (intake, downpipe). if it's completely stock, i'd first start by cleaning your maf sensor before making any changes.

and FWIW- auto subarus pull around 25* of timing during shifts based on a response from the tcu shift signal, so the fact that you see close to no timing is ok. this causes a small surge of spiked EGT's which can sometimes effect turbo spool during a shift. some say subaru did this to prevent the turbo from loosing spool during the shift, some say it's to reduce the engine output during a shift. you'd have to log manifold pressure during a shift and see if it jumps at all for that, and modify timing according when running a different turbo setup.

first of all, is the car completely stock? what mods are done to it? i'm reading that the 03-06 2.0GT liberty came with the vf38 stock so I'm assuming the turbo is stock? if not then clearly you need a tune, I.E. smoothing out the wgdc, lower lowend timing a bit, fueling

 

the twinscroll will certainly cause a quick surge if the wgdc jumps all the way up abruptly. remember how load is calculated- maf g/s x 60 / RPM. that means, if the wastegate is closing, the turbo will spool, therefore more air will be pulled through the intake maf, thus causing a higher g/s reading, us guys with a single scroll aren't used to this instant spool . but it is unusual to shoot across the timing table to the right so quickly(as you can see by the tot timing being so low at that point), especially at only ~33% throttle.

 

what to do next is determined by your answer as to whether the car is completely stock, or has been modified (intake, downpipe). if it's completely stock, i'd first start by cleaning your maf sensor before making any changes.

 

Thanks for the input Ripemeat.

 

The car is completely stock with just a bit of tuning to the ROM. The tuning has not touched any of the low rev, low load regions of the map however.

 

I'll give the MAF a good clean and see how that goes.

 

Would be happy to upload my ROM but it looks like that's not allowed.

upload the map to google or something

Cool.

 

Well, I cleaned the MAF and it looks like the high load/low RPM issue has gone now!

 

I still see knock very regularly at around 2k RPM's and ~1 load.. like when pumping the throttle in traffic to move foward a few car lengths. That part of the map is all stock so I don't know if I can tune it out or not. I've very recently changed out the spark plugs and I neglected to check the gapping on them before I chucked them in! >.< So I'll check that out asap too.

I'm going try a different fuel vendor for a bit and see how I go for a few weeks and see if thats a factor as well. Have been running 98RON fuel from Caltex for the last few months but have changed over to Shell 98RON this week. Will keep on logging and see whats what!

 

I'm also thinking of changing over to an aftermarket 3 port EBCS/MBCS hybrid setup soon and will have to re do all the WGDC values anyways.. so maybe that will help with the boost control a bit at lower revs, and prevent overboost.

 

Thanks, all of you guys, for the input!

One last thing.. Here's a pic of a recent Leaningview.. I see the fuel trims a bit off in the low range.. could this be a clue?

Nope, looks very healthy according to your LV.

Just realized something...did you recently reset your ecu? Trim D is completely 0, hrm.

 

Also you won't see much advantage over stock even with the EBCS. And it requires redialing in turbo dynamics values also.

Just realized something...did you recently reset your ecu? Trim D is completely 0, hrm.

 

Also you won't see much advantage over stock even with the EBCS. And it requires redialing in turbo dynamics values also.

 

 

I'd reset the ECU prolly about 100kms earlier when I last flashed a map to it. I haven't really had much chance to go to redline much but I'm keeping an eye on it and it hasnt changed in a while.. will post an update soon.. maybe I'll got for a spin later tonight and see whats what. My day to day commute invloves a short 15km trip in heavy traffic so it's hard to get those high rev trims dialed in. Getting into those rev ranges in anything but 1st gear usually results in 100km/h pretty damn quick.. but I'll try my best and report back.

 

I guess I wanted the EBCS/MBC set up as I do see a bit of overboost when going balls to the wall. I have a max target of 17.5psi but it'll often spike a good 1 to 1.5psi over that on a shift. I spose thats why timing drops to nothing as well.. to save the engine from exploding.. lol

I'd reset the ECU prolly about 100kms earlier when I last flashed a map to it. I haven't really had much chance to go to redline much but I'm keeping an eye on it and it hasnt changed in a while.. will post an update soon.. maybe I'll got for a spin later tonight and see whats what. My day to day commute invloves a short 15km trip in heavy traffic so it's hard to get those high rev trims dialed in. Getting into those rev ranges in anything but 1st gear usually results in 100km/h pretty damn quick.. but I'll try my best and report back.

 

I guess I wanted the EBCS/MBC set up as I do see a bit of overboost when going balls to the wall. I have a max target of 17.5psi but it'll often spike a good 1 to 1.5psi over that on a shift. I spose thats why timing drops to nothing as well.. to save the engine from exploding.. lol

 

No, the timing drops in order to spike egt's during a shift, high egt's spool the turbo quicker, that results in faster compressor spool and a spike in boost.

 

When the gear changes to the next, the rpms drop and therefore flow of exhaust is less, meaning potential turbo stall, subaru prevented that from happening by spiking egt's to keep the turbo spooled and a smoother transition between gears without the hesitation or loss of power.

No, the timing drops in order to spike egt's during a shift, high egt's spool the turbo quicker, that results in faster compressor spool and a spike in boost.

 

When the gear changes to the next, the rpms drop and therefore flow of exhaust is less, meaning potential turbo stall, subaru prevented that from happening by spiking egt's to keep the turbo spooled and a smoother transition between gears without the hesitation or loss of power.

 

 

Ah! Thanks for that ripe. Very insightful stuff!

 

Here's another LearningView from today after a good bit of 'spirited' driving.. No 3rd gear pull to redline.. but plenty of 2.5+ load, 6000rpm+ dashes out on the highway.

High rev range fuel trims comming in now.. Not to bad?

keep an eye on those fuel trims, currently the ecu is adding fuel because it's running leaner than expected. also watch for that right hand column knock (which could be related to the fueling matter)

Hi guys,

 

I've been driving and logging for another week and I'm still seeing high load/low rpm evverytime I take off from a stand still.

I've put back on the factory map now and I'm STILL seeing the same issue, so im pretty sure it's not the tune.

 

I've attached another log with three of these events put together.

You can also see in these logs, each time the timing drops down to about 9 as well.. and there's a knock along with it.

 

I'm starting to think these ugly numbers are actually supposed to be like that.

Can anyone else with a 5EAT confirm that taking off from a stand still at low RPM produces the same numbers for that split second? Reduced timing, high load and a knock?

 

Could it just be the transmission gaining traction off the line?

romraiderlog_LOWRPMKNOCK.csv

Hmm..... *thinks*

I've tuned a few autos and have come to the conclusion that due to the transmission and torque converter causing rapid changes in loads, the ecu doesn't like it, simple because it's skipping across the timing table very fast. I personally don't give a rats ass about knock sum ASSUMING the fbkc and flkc ranges(rpm and load) are set wide enough) The general consensus is with higher power setups with a 5eat is to tune it for no FBKC/FLKC on the AT map, this can mean you're going to have to build a hole in the advance table in the 4k-5k range. Some go as far to tune it up until FLKC registers after a shift. :/

 

The stall on the TC is up in the mid 3k's. The solution to these rapid changes in timing during a shift with modified 5eat's have been to use the MT map of that model year, if you have one. This eliminates the ecu pulling timing between shifts and therefore timing references don't skip back and forth harshly, and the ecu is less likely to knock.

Thanks ripemeat. I don't think I'll be switching to a manual ROM but I'll keep everything you've said in mind when watching for problems.

I've recently changed a bunch of fluids and did an Upper Engine clean with the foam from the subie dealer. I am seeing a lot more knock in funny places since. i suspect maybe the Upper Engine clean, cleaned away alot of carbon from the valves and now maybe i need to go get a compression check. :/

Did a compression check and checked intake to manifold for air leaks.. all good. I checked the gaps on my spark plugs too. All good. Infact the engine seems to be running great!

 

I did notice the welds on the DP heat shield have cracked in two places though... That would explain the sudden increase in knock events that got me on this wich hunt in the first place.

I'm thinking this could be the cause of a bit of the low load knock. Maybe even some of the other knock in the power band..?

 

I've also decided that the High Load/Low RPR situation is just normal drivetrain operation.

No one has posted anything saying they see the same on their 5EAT set up but it seems to be pretty regular in my logs now.. maybe I just wasnt looking for it earlier on and now I've noticed it bugs me.

 

Thanks guys, for the help and the pearls of wisdom from ripemeat.. You've all be a great help!