Boost exceeding target?

[serx7]

Curious what was tweaked. You mentioned in an earlier post that you were seeing FBK at low rpm, part throttle, no boost, 'high calculated load'. From what I see in datalogs 61, 62, 63, your problem area is 0.8x-0.9x g/rev load and 2100-2300 rpm under somewhat light throttle. I personally don't consider 0.8x-0.9x g/rev to be 'high load', esp given that you're seeing nearly 2.6 g/rev load @ WOT in 3rd above 4500 rpm.

 

I'm not sure if this is possibly shift knock, or false knock. Have you checked/banged around under the hood/under the car for a potentially loose/rattling heat shield or something else that might be making a metallic rattling sound?

[ERLoft]

Nope, haven't done that. I'll try to check that as soon as possible. Logan actually made ignition timing adjustments to the OTS map and sent me a copy, so that's what I've started running this morning. Still saw some knock, but I probably need to give the ECU a bit of time to figure things out before logging. Even so, I'm monitoring it closely and backing off when knock appears on the AccessPort.

[McDowell Performance Tunin]

Yeah, timing was backed off a bit in the area's of knock. Nothing else changed on the OTS map

 

Sent from my SAMSUNG-SGH-I747 using Tapatalk

[serx7]

ERLoft: As iNVAR mentioned earlier, shifting adds another layer of potential complexity. Your datalog60 included WOT but you didn't go WOT until around 4k RPM in 2nd, and the upshift to 3rd kept RPMs above 4300 RPM.

 

I would highly recommend doing a 3rd or 4th gear (SINGLE gear) pull from ~2k to 6k+ (if possible) so that you can test the tune at high load & across the entire RPM band. Obviously, if you see FBK show up on the AP, abort the run immediately, you'll probably need to reduce/smooth timing anywhere you see FBK reproducibly show up @ WOT.

 

When I do this type of testing, I usually do it w/ a full tank of gas and on a straight stretch of road that's up a -slight- incline. This loads the car up about as high as it would ever see in real world conditions (without roping in a passenger, my wife is only sometimes amenable to assisting in this regard .

[iNVAR]

When I do this type of testing, I usually do it w/ a full tank of gas and on a straight stretch of road that's up a -slight- incline. This loads the car up about as high as it would ever see in real world conditions (without roping in a passenger, my wife is only sometimes amenable to assisting in this regard .

 

Get fatter and you won't have this problem.

[ERLoft]

Four new logs.

 

Log 65 has a second gear pull. Log 66 has a third gear pull. Log 67 has a fourth gear pull. No knock on any of the pulls.

 

Log 68 is regular driving and shows a couple of Feedback Knock hits. Still not sure what the heck is going on here? I'm having a compression check on Friday - maybe it's time to use the rubber mallet to bang around the headers and such looking for a vibration? Bad knock sensor? Not really sure.

 

I'm hoping the compression check comes up clean. If it does, I'm actually thinking about going Stage 3. BNR HTA68 with upgrades to the fuel system, intake and EBCS...

datalog65.csv

datalog66.csv

datalog67.csv

datalog68.csv

[iNVAR]

Wow, that's quite huge knock at that one spot in log 68... -8.4 degrees.

 

Inspect your knock sensor or anything else that may be loose in the engine bay. Also, have you cleaned your MAF? Your AF Learning 1 shows you adding a non-trivial amount of fuel in. You may be leaning out and the ECU might be trying to compensate. Before the knock event, your AF sensor shows 15.16 and 14.7.... Hmm

[ERLoft]

Yeah, I've seen even more knock. Nothing on the WOT on the acceleration runs though. It also seems that when the Feedback Knock shows up, I can back off throttle, then go back on and it's not likely to recur, even though the conditions are very similar.

 

I'll clean the MAF on Friday when I'm getting the compression check done. As noted, I'm going to bang around exhaust system too looking for rattles and loose items.

 

Any way to determine if the knock sensor is bad?

[iNVAR]

I don't think there's a way to tell other than swapping it out.

[serx7]

Comparing datalog68 w/ the big -8 deg FBK hit to your previous logs (particularly datalog63), it doesn't look like you saw FBK at the same RPM and similar (or slightly higher) load (0.8x g/rev and ~2900 rpm). Your AF Learning #1 was in the 3.12% range in datalog63, whereas it's 4.69 in datalog68. This would sort of point to false knock (rattling, faulty knock sensor) although it could be you're borderline-enough lean as iNVAR noted.

 

As far as your plans to go Stage 3 w/ a bigger turbo and supporting fuel and boost control hardware, I would recommend -against- going there until you get your current knock situation squared away. Going to Stage 3 or beyond while seeing un-corrected/un-explained knock in your much milder current state of tune, is probably asking for more headaches than it's worth.

 

I don't recall what Stage 3 is exactly, but I'm running a VF52 @ 20 psi peak tapering to 14.5 psi by 6k rpm, on a built motor w/ upgraded TMIC, DW65 pump, DW740 injectors, TGV deletes, catless UP, PnP/coated headers, Grimspeed EBCS, Cobb catted DP w/ wrap, Nameless Performance midpipe/Y, stock mufflers (303 whp / 339 wtq on a Mustang AWD 500 SE dyno in 83F temps), and trust me you are going to want your knock situation well under control. In 30-40F temps I routinely see peak loads of 3.4x g/rev. You do not want to see knock there or anywhere near there

[ERLoft]

@ Infamous now - compression check came back clean 150-155-157-150. Mike found a lot of blow by - intercooler gaskets were bad among other issues. Still not sure if this was phantom knock - car is going up on the rack to check for loose items that could cause phantom knock... At least I was able to finally breathe once that compression check came back.

 

Sent from my SCH-I535 using Tapatalk

[caramall2]

Cool! Did he suggest leak down or think it was unnecessary w good compression numbers? Have had good compression but so so leak down before.

[ERLoft]

He did a smoke test first, but we haven't done a leak down test at this point.

 

Sent from my SCH-I535 using Tapatalk

[ERLoft]

So, no leak down test. Todd did the compression check and said because there was no fluttering on the compression check, it shouldn't be needed. Going to go back next week or so to have the PCV valve. Also, when we put the car up on a lift to check for possible rattles and such, found a pinhole in the driver's side front outboard CV boot. Going to get that fixed too.

 

On the original knock issue, it seemed a bit better on the drive home, but not gone by any means. We'll have to see after the ECU gets a few days of learning what happens, but we may need to dig deeper to find the source of the issue.

 

Even though I've been told that knock sensors have issues only very rarely, I'm definitely wondering if that is a potential source. It'd be nice to have a spare to swap out and see if the same symptoms persist. As mentioned above by serx7, it would be a bad idea to go bigger until the issue is sorted out.

[ERLoft]

Well.... Apparently the work thus far hasn't done much for the knock issue. (Or phantom knock issue as the case may be.)

 

Made a log today from some 'normal' driving. Getting onto the freeway, short cruise down 2 exits, then off the freeway. Two different knock incidents - nothing on WOT runs getting on the freeway, but later while cruising in 5th, on lines 472-475 we see some. Then on lines 531-552 there are 3 separate incidents, all in 6th gear, all nominal load, all in the 2200 rpm range. Feedback knock goes as high as -11.2, which seems crazy high to me.

 

Not on the log, later in the day, I saw a -7 number on a throttle blip when heel-toe downshifting from 6th to 5th.

 

Can anyone help make heads or tails out of this?

 

On an unrelated note, with the boost leaks fixed around the intercooler, I ended up throwing a P0244 code. I'm guessing that with no more boost leaks, that for the time being, I should adjust the wastegate actuator arm length to lower the boost levels a bit. Code was thrown in 5th gear under hard acceleration.

datalog77.csv

[serx7]

curious that your DAM starts out at 0.5 in datalog77. not sure if that's being caused by the knock you're seeing at lower loads around 2200 rpm but i suspect that's the cause.

 

regarding the P0244: as seen in your datalog77, you're boost spiking over target by 2.08 psi in 3rd and 2.9 psi in 4th, presumably higher than that in 5th. see lines 259 and 272 in that latest datalog.

 

before you go adjusting the WG actuator, you might want to have the WGDC and/or WGDC IAT Comp and/or Turbo Dynamics tables looked at. to get a better picture of your boost control situation, you'll probably want to add TD Boost Error (you can add this and remove Target Boost to avoid affecting logging sampling rate), as well as TD Integral and TD Proportional. You can probably remove Knock Sum and MAF (g/s) to preserve logging sampling rate too.

[serx7]

as for that crazy FBK at middling load and ~2200 rpm, i'd probably avoid tooling around in 6th at 2200 rpm in general, esp if you need to apply throttle. you're lugging the motor by doing that. that isn't going to fix the issue, but it should reduce the occurances of it.

 

in 5th and esp 6th, I try to keep cruising RPM closer to the 2500 rpm mark. i get occasional FBK 6th if I go too heavy on the throttle below 2750 rpm. i can avoid that knock by pulling a little timing at that load/rpm range, but doing so has a somewhat noticeable effect under load in the lower (and more frequently used, esp for acceleration) gears, so I just adjust my driving habit instead.

[ERLoft]

I also noticed today that I had a small amount of knock on initial throttle tip in when leaving a light in 1st gear. Again, not the kind of situation you'd expect to see knock in.

 

As for not using 6th gear below 2500 rpm, that's pretty tough. That means keeping my speed in 6th above 75 or so, which isn't always feasible in normal traffic and running 5th gear all the time kind of defeats the point of having a 6 speed, doesn't it?

 

Either way, I have to figure out what the issue is. I'm hopeful that I can get my hands on another knock sensor so at least I can eliminate that from the list of possible culprits.

[McDowell Performance Tunin]

Knock sensor is a good place to start with the kind of FBK you're getting, at those engine loads and rpm levels.

[serx7]

I also noticed today that I had a small amount of knock on initial throttle tip in when leaving a light in 1st gear. Again, not the kind of situation you'd expect to see knock in.

 

As for not using 6th gear below 2500 rpm, that's pretty tough. That means keeping my speed in 6th above 75 or so, which isn't always feasible in normal traffic and running 5th gear all the time kind of defeats the point of having a 6 speed, doesn't it?

 

Either way, I have to figure out what the issue is. I'm hopeful that I can get my hands on another knock sensor so at least I can eliminate that from the list of possible culprits.

 

i wasn't suggesting that you avoid 6th below 2500 rpm. i was suggesting to avoid it at/below 2200 b/c of my commute (mainly highway, mainly an hour+ after morning/evening rush), i can pretty routinely cruise in 6th above 2500, usually above 2700 rpm.

 

def let us know if swapping out the knock sensor helps (make sure it gets torqued correctly or you'll potentially get other undesirable behavior). after working w/ my tuner for a period of time last summer, we ended up disabling FBKC at low loads and low-ish RPM. all testing that was done around that pointed to increased noise from the built motor causing false knock. most of it actually goes away once the motor is fully up to temp, but unless i let it idle way longer than i'd like to wait for the internals to warm up (esp in winter), i'd see FBKC b/n 1600 and 2000 rpm getting up to speed on the road leading out of my neighorhood in any gear at low load, and DAM would eventually drop quite a bit within a day.

[ERLoft]

New knock sensor should come in this week from AllSubaru.com Talked with a Subaru master tech @ Frank Subaru in National City last week - one who also happens to be an enthusiast with a fully built Forester XT and Outback XT. He concurred that I should check the knock sensor and the parts manager there said they replaced those quite a bit.

 

Over the weekend I also did an oil change - old oil was very clean, no sign of metal in it. Had the intercooler off and checked all the vacuum lines, those looked good. I did notice that the connector for the knock sensor was hanging, not clipped onto anything, although the connector looks like it's supposed to mount somewhere...

 

After the oil change, I let it warm up and was checking the feedback knock on zero load revving of the motor - turns out that I can see as much as -4.2 degrees of timing pulled just on revving the motor in neutral under 4k RPM. No way that there should be detonation in that scenario.

 

Presuming no shipping delays, I'll install the new sensor this weekend and we'll see what happens...