From a tuner's standpoint, tuning with the attempt of getting no knock, is really for their own protection, as someone could use the knock counts to blame a tuner for a bad tune that damaged their engine (which in some cases they would be correct)

Under light loads, and transitional throttle inputs, you'd be surprised how much knock there is on a bone stock evo.. Even under load, with as rich as it runs, also sees knock sums higher than the average tuner targets..

Knock is bad when its due to too much timing where cylinder pressures build to dangerous levels prior to hitting Top Dead Center so your trying to compress expanding gasses, preignition and detonation are something a little different when there the ignition of the mixture is spontanious. A tune and octane of the fuel can have a direct affect on this

You are much better off trying to be as conservative as possible until you understand really deeply all of the cause/effect relationships in the engine. Tuning isn't voo-doo, black magic, or rocket science, and anyone can do it, if they take the time to understand what their working with.

 

Even the slightest amount of real knock can do damage, depending on many factors including your power level, how long the knock is sustained, etc. You have to remember that your engine is made out of metal parts and metals have a yield strength. Real knock dramatically increases cylinder pressure and that pressure exerts a force on the piston, connecting rod, and bearings. If that force is great enough for plastic deformation, you will have immediat damage...if the damage is bad enough you will blow your engine right then and there. If it isn't bad enough, you may eventually spin a bearing due to the bearing losing there shape. And, even if the force isn't past the yield point of the metal, metals also can fail by fatigue, which is the repeated exposure to stresses below it's yield strength over a period of time.

So, even though small amount of knock may not immediately do damage to your engine, in the long run, it definitely can. It's the very high knock sums that can do the immediate damage that you are referrign to, especially at high power levels.

I don't think you really understand that timing was still being pulled in your log. With your 10 counts of knock, you were probably gettign about 3 degrees of timing pulled. You see your timing advancing because your timing maps are advancing in that RPM range. When timing is pulled by the ECU, it doesn't mean that timing cannot advance through the range. The ECU just looks up the number that timing should be and it subtracts the knock correction from it. So you may still have an advancing timing profile, it's just that you will not reach your maximum timing in your map at the point where knock is present.


Eric

 

3gturbo, are you using ECUEdit to view the EVOScan logs? how hard is it to do that and is that program shareware?

 

I'm not going to argue but in my log my timing was never being pulled due to knock sum it dropped at the one point because thats where I floored it (you can't see the tps ) and it advances from there I can send you the complete log if you like along with the rom file.

 

If it's hitting over 5 knock counts and not pulling any targeted timing then your ECU is different than everyone elses ECU.

When you get those 10 knock counts, is the car hitting the exact timing number you put in that specific load cell the car is in?

 

I can 100% guarantee you that it was. You can post a screenshot of your Hi Octane timing map if you want, along with your log. Depening on how many items you were logging, your calculated load may be a little off, so it may be difficult to tell exactly what cell you are hitting, but I can guarantee you that you had timing pulled.

Eric

 

What Eric is saying is correct. I argued with him about this before, but now that I know more, I belive him. Today I did three logs in the pm hours. One log had a KS of 6 @ 3700 rpm and the KS continued to decrement as the rpms increased. So it went from 6-5-4-3-2-1. This continued all the way to redline. Using DataLogLab I ploted the hp/tq curve of this log vs the other log that did not have more than 2-1 KS here and there. As you can see the car lost power in the benchmark (knock prone run) as opposed to the "your power" run.