Tag Archives: knock

Subaru Knock Monitoring – Cobb Tuning

I read Cobb Tuning’s Facebook postings and saw the following the other day:  http://www.cobbtuning.com/subaru-knock-monitoring-and-dit-knock-logic-changes/

Most people that have turbocharged Subarus and either have a Cobb Accessport or want one always ask how does one monitor for knock.  I’ve elaborated on this in the past (here) but the information in the link comes straight from the source!

Excerpt:

Detonation events are inevitable and will occur from time to time on any modern vehicle running on pump gas.  Your car is built to recognize these events and take the appropriate action to defend against them causing any damage.  The Subaru knock detection system tends to err on the side of caution.  This makes it common to see “false knock.”  False knock occurs when the ECU corrects for a knock event but the noise registered is due to other noises that aren’t necessarily detonation or harmful to the engine.  This often happens accelerating from a stop while letting the clutch out, during gear shifts (more so if grunting while shifting), accelerating at low RPM in a high gear, under cruise on the freeway, and during abrupt throttle changes.

Read the rest at the above hyperlink.  Enjoy!

Cobb AccessPort – How do you use yours?

I’m a member of several large WRX/STI sites/forums and while I lurk on every one but IWSTI, I read a lot between them all (and sometimes subscribe to those that attract my attention).

I’ve noticed a trend with Cobb AP usage.  With version 3 out now, people are using the AP as a gauge (or a set of gauges).  That’s OK, but I don’t think the unit was designed to replace an actual boost gauge and I’m pretty sure the mechanical gauges will always be more accurate.  I guess it’s OK to use the AP in that capacity, but it can be dangerous if you’re depending on it in that capacity.  It’s one thing to use it while knowing that it may not be 100% accurate…it’s another to think or assume it’s the better tool, because it isn’t.

As well, I’m also noticing a trend where people will use the AP as gauges and monitor real-time in lieu of logging events via the datalog feature.  This thread in particular highlights this problem.  That is definitely dangerous and here’s why:  Monitoring gauges real-time isn’t really real-time if you’re not watching them (you’re not going to be able to watch them and drive at the same time…you might think you can, but you can’t).  Now, under each virtual gauge, there are areas that monitor the lowest and highest measurement of that particular gauge.  Let’s say you’re trying to watch the Fine Knock Learn (FKL) gauge while driving.  You eventually look at the AP and see that while you weren’t looking at the gauges, the AP logged an FKL event.  The counter under the gauge shows the value of, for example, 1.35.  The thing is, you don’t know if it counted that once, or many times while you weren’t looking at the gauge.  Also, this also tends to create a false sense of security because you’re thinking, “oh, there was only one event” when there could’ve been 20 while you weren’t looking.  One measurement showing within the peak value field does not mean there’s nothing to worry about.  That one measurement could be happening more than once.  Very rarely do I actually see an FKL event showing within the AP gauge as it occurs (that’s only happened twice).  Now, if you’re datalogging as well as live-monitoring, that’s a different story.  The analogy that I used in a post at IWSTI is that live-monitoring is akin to reading Cliff Notes, while reading datalogs is akin to reading the actual book and not a summarization of the book.  Based on what I’ve been reading in the forums, people are using the AP’s live monitoring as the only method of monitoring.  I saw at least one guy today state that he has 70psi in cylinder #4 but when someone asked him to post logs awhile back, he didn’t have any, yet he stated that he saw knock events…it is assumed that he saw those knock events on his AP.  He never did post the logs but then posted he had a bad cylinder/piston/ringland.

Subaru owners will need to be extra diligent in following their cars’ maintenance schedules, datalogging (especially after filling up the tank with new gas), not beating up on the car when the engine is cold, ensuring the oil has the proper level, ensuring the oil is clean, ensuring the engine isn’t being lugged in a high gear (no flooring the engine in high gear or on inclines)…following those instructions can lessen engine damage, but logging is key, in my opinion.

Lots of Knock!

A few days ago, it got warm enough to where I wanted to go to a car meet.  I took the car (of course) and ran logs on the way out.  I also ran logs today when driving to a doctor’s appointment.  Well…I saw lots of knock…WTF.  The only thing I can think of that’s causing it is bad gas (I filled up on Friday…those logs were partially clear).  Thinking on it now, it could also be the fact that the weather is warmer…warm enough to have heat-soak (which could contribute to knock on the OTS Stage 1 tune).

I got scared enough to where I unmarried the AP…it may stay unmarried until I’m ready for a protune.   While the DAM was still 1.0, the knock was very prevalent (finke knock learn, actually, which pulled timing quite a bit in some instances) .  I’ll see if I can post the logs.

EDIT – The logs are here.

UPDATE – I think this might have been a combination of bad gas and the fact that the engine was loaded and in top gear. No, I wasn’t lugging the engine, nor was I boosting significantly in 6th, but I was in 6th with light load and running the AC, which the 93-octane Stage 1 OTS tune doesn’t like. When I went to the Stage 1 91-octane OTS tune, it stopped. I’m still running the 91-octane Stage 1 tune.