4 Attachment(s)
3.6L LLT knock retard is probably more normal than is thought
I'm working with an 08 Cadillac CTS 3.6L DI (LLT). The car is stock running on the OE tune. I've been logging data in preparation for future tuning. The car is rated for 87 octane fuel and that's what I have been running in it up until now.
I noticed quite a bit of knock retard scattered over the graph, including areas and during times where conditions suggested there shouldn't have been any at all. Despite seeing as much as 10 deg spikes, I was never concerned there was a problem, considering the number of times it seemed to be applying knock retard as a gift.
I proceeded to test the validity of the knock retard by switching to a tank full of premium 93 octane fuel. Mean while prior to the switch, the car never felt like it was low on power and I never felt the kind of dip in power one would expect from a sudden 10 deg timing retard and I suspected the PCM either registered and ignored it, or gradually adjusted for it with camshaft phasing.
93 octane reduced the knock retard recorded, but nowhere near what I'd expect a 6 number octane increase to do. It should have been gone completely on a stock motor rated for 87 octane, on an 80 ish degree day in FL.
I purchased a can of VP Octanium Unleaded (O2 & cat friendly) that would increase the octane rating of 10 gal by 7 octane numbers. Numbers is important because it doesn't mean the same as points, which would have been .7 octane increase.
I estimated my octane rating to be 99 and parked the car after adding it to the ~10 gal of 93 and shaking the tank for about 2 miles and parking the car. The next day I started the 3rd leg of the test. Initially I felt a sizable drop in power, at about mid full acceleration, that eventually seemed to clear up as the car was driven and as I neared the end of the testing about 20 min later, I noticed the tires started break traction a little on wide open throttle from a stop, where it had not done so previously and less than a gallon of fuel had been burned so the car was not necessarily lighter to account for that.
Again the logs showed an improvement in spark retard, but not all that significant. It was the fourth unscheduled test drive that drove home the importance of giving the PCM the time it needs
to make any possible changes and it was clear beyond a doubt at least to me, that although there is no high octane table in the E69 PCM, it can detect the use of premium fuel and take advantage of it.
We spend so much time watching spark retard, I believe spark advance gets overlooked, so I checked that to and low and behold, the spark advance was steadily increasing with the octane as well as with the next day of driving. It also appears that the PCM has switched into learn mode for the spark advance table, because I began to see negative spark advance on the high end of the table, (that corresponds to the negative advance settings in the stock tables), that was absent while on 87 octane which it had been on for about a yr. Those negative areas also started to increase toward positive with successive test drives, as if the PCM was feeling out how much advance it could apply, or in other words, how close it could get to optimum spark.
In the old OBD-I platforms, the knock sensor was tested by the computer with a deliberate over advance to cause knock, to make sure it was functioning. I suspect the knock so many are complaining about, is deliberate and a means for the PCM to fine tune the ignition advance using a similar method and that's why it is often seen under conditions that make no sense otherwise.
I'm going to keep the car on 93 after the 99 oct mix is used up to see if the changes in ignition timing hold. There is one interesting caveat, although calculated, my HP and TQ peaks were a little higher on 87 octane than on 99 oct by about 5-10 points, despite the significant difference in spark advance in the 90% column. The car and the OE tune are unmodified.
My snapshots are of 87 octane max advance followed by 99 oct and 87 max retard, followed by 99 max retard for the last test drive which was today. I could be wrong, but how else can such a pattern be interpreted.
Also keep in mind, the changes are gradual, there were two days of testing on 99 octane. The first day on 99 oct, the 90% column still showed 8 deg of retard. It wasn't until the second day of testing that it dropped to what is seen in the picture.