One topic that has always stumped me was how to go about producing a timing map after increasing requested load beyond stock values.  For example, if you've increased KFMIRL, LDRXN, KFMIOP, etc, you likely have KFZW(2) rl axis values that fall short of your highest LDRXN value.  So you can increase your load axis on KFZW, but the actual timing angle is pretty looney for that specific load now.  What have people done in this situation?  Retard the timing in the higher scaled load columns?  By how much? 

I know the true answer is to log, log, log, but realistically, how do you know the optimal firing angle for every load and every rpm by just logging without pushing the motor until it knocks?

No, this is bad tuning.

Yeah, I've modified the line that WOT follows... basically a warped bowl shape within the map.  I worry about all the other values at partial loads though.  I guess worst case if it was too advanced I'd hear/log knock.

The reason I ask is: I recently snapped a rod at low rpm, just cruising around town, at around 3000 rpm and very little load request (maybe 5-7 PSI tops).  I haven't yet determined what caused this (maybe a K04 torque spike).  Part of me wants to blame my tuning, but part of me realizes that shit just happens.  Basically, I was running KFZW(2)(OP) from the TT225 BAM. I am actually running an AWP 1.8T.  I realize this engine has a slightly different crank, but I don't think it'd really affect timing that much that I'd be snapping rods before knock recognition kicked in.  The only CFs I ever logged in this configuration were at WOT, and they were -3 at most.   I'm probably worrying too much... and my new fully built bottom end will be fine down the road.

Give advice then, to be honest I notice a pattern with your posts... Ever heard of Eric fletcher?  We don't want to just be told we are wrong like you like to do, if you going to take 30 seconds to say that take an additional 30 to drop some knowledge since you continually imply you have a lot of it.

There is no right or wrong way to tune, just a method which is safe, works well and which makes the driver grin ear to ear.

I always scale last load/rpm site then log during steady state testing.

Yes, please, for the OP! LOL

LOL nye

I agree, but all of that can be done by re-interpolating intervening points.

The only data that is missing is whatever load columns you add... and those columns generally are only used during WOT, since they are the maximal load cells...