Usually spark changes of 1 or 2 degrees will change the knock retard averages. It can be challenging to reduce knock retard, because increasing or decreasing Spark Advance will affect knock retard averages. Sometimes adding Spark Advance will eliminate knock retard averages, which seems counter-intuitive.
dave w
Copy that. When I was trying to see why my knock sensor was not reading, I added a bunch of timing to my tables. I started with 454 tables and having an over square engine, I need more then 454. So I think I am in the ball park.
Pardon for butting in, but I've just started playing with tuning LS1, and spark seems harder to tune than fuel, at the moment.Originally Posted by dave w
Dave W, would it be fair to say that having the tune take timing out due to knock detect costs you more performance than not having quite enough advance and no knock in the first place?
I base that on observation from logging. I have a 'test dyno' bit of road that I log acceleration from 2 to 5 000 rpm in 3rd gear on, and use time scale to measure performance, quicker the better.
Obviously no correction for airtemp, barometric etc, but pretty indicative, and 'real world' conditions.
I am assuming more advance is more power, but upon reflection of that, maybe that's not true.
And am I placing too much faith in my test method, as well?
Again, sorry for jumping in, not wanting to hijack your post rang-a-stang.
Last edited by Mr Weston; 07-03-2020 at 09:59 AM.
You're good Mr Weston (not hi-jacking at all). This is all information I am hoping to discuss.
My understanding from reading other threads on timing is most folks will continue to add timing until they get high knock counts and retard, then back the cell(s) down just a tad. There is a finite point where there is too much timing. We want the max timing possible but as soon as we hit knock, we are losing efficiency (obviously there is knock from other conditions besides timing but for this conversation we are speaking purely of timing).
Thanks for the info, rang-a-stang.
Whilst having a think about it, another thing I remembered was that you're most likely to get knock at max torque, and I've seen that in my logging.
Seems I can add more down low, and no knock,but not so much around 4000 rpm in my LS1.
Knock is an audible event which indicates combustion pressure and temperature have risen too quickly and the energy which would be used to drive the piston is being converted into sound. Knocking can be caused by pre-igniton (before spark plug fires) or detonation (after spark plug fires). Triggers can include hot spots on the cylinder head or piston, lean air/fuel mixture, incorrect spark timing, incorrect octane rating, or a mix. During knocking the engine block, cylinder head, and piston top are heated above average temperature. As chamber temperature rises it becomes easier to light the air/fuel mix and the rate at which combustion occurs decreases. If knocking is allowed to continue there will almost always be damage.MY QUESTION NOW, THOUGH:
I have done a tuning run with my timing tables and have data on my knock retard. Do I just take that Knock retard number (running avergage), subtract it from the value in the timing tables, install new table in my truck, rinse, and repeat? Am I looking for 0's or just less than say .5? Is it really that easy to tune timing? I feel like I am over simplifying it...
Preventing engine damage using knock counts is really a dynamic process. Once knock begins the computer has to reduce spark advance below the value where it first occurred. Timing might be safe to 43 degrees, knock begins at 44 degrees, then the knock sensor activity causes timing to drop to 39 degrees before knock stops. This is because the combustion chamber and piston top require time to cool off. A stock timing table and ESC system are programmed to advance spark to table maximum, and if knock occurs ESC will reduce advance until knock stops. It will then increase advance toward table maximum. If knock occurs during this process ESC will begin the cycle again.
Tuning with knock counts imo requires time travel. Tuning with knock counts is not simply a matter of reducing spark advance at the RPM and MAP where knock is occurring. The goal is really to prevent knock by decreasing timing by the smallest amount necessary. Properly tuning with knock counts IMO requires reading datalogs from a point in time prior to the onset of knock to the point which indicates combustion chamber heating is causing or strongly contributing to knock. Changes such as afr increasing or rapid spark advance can begin a cycle of temp rise that results in abnormal combustion. I have seen cases where increasing fuel delivery at 2000 rpm prevents up to 6 degrees of spark reduction at 3000 rpm. I use the shape of the spark table as a guide. If I see a nice curve that generally seems to work, but requires a substantial change in shape (up to 4-5 degrees of timing reduction in one area) due to knock, I try to find a way to prevent knock in the first place.
In short: Drive the vehicle to obtain logs. Review logs for knock. Adjust timing tables to reduce advance and drive again. If you find you are reducing knock by more than three degrees in any three adjacent cells review the logs for changes that can be made elsewhere which could prevent heat in the combustion chamber from causing knock.
Perfect!!!!! Exactly the info I needed! Thank you!!!
x2
Looking for the 'like' button
Thanking you 1project2many
One of the best " makes sense " explanation I have seen for spark tuning.
I have a thought about this...
Have you verified timing to be correct? That what is being commanded is what is actually happening referencing crank angle, at RPM, I know without a dyno under load will be impossible to verify, but you can at least rev it in park and make sure that the actual timing follows commanded. I have seen a number of times where looking at commanded you'd think that the timing is far too advanced, but actual timing was not correct. There's a bunch of reasons for this, but starting with verifying correct timing is the place to start.
The man who says something is impossible, is usually interrupted by the man doing it.
Posting Permissions
Reply With Quote
Bookmarks