Ok, here is what I think is happening and better minds than mine can chime in and perhaps offer a better solution.

The problem: When I cold-start the car, it will crank and start but soon die if I don't touch the throttle. I can do this once or twice and then the car will start ok. BUT if I tap the throttle at just the right time the car will start and run (albeit sloppily) until it warms up sufficiently and then it will run just fine.

What I think is happening (based on looking at multiple logs);

1. The car cranks until it catches and the RPM quickly climbs into CL idle territory, but it's below the target I set for that CLT.
2. CL idle says, "Son, your idle is too low, let me add some advance".
3. But then the engine says, "DUDE! That's too much advance! I can't live with that" and so the car dies.
4. Depending on the ambient temp, the CLT will creep up after one or two attempts to the point where the idle advance is tolerable for the engine and so the startup event proceeds (with a bit of coaxing from some judicious throttle tapping).

So what do I do? Well i could;

• Reduce the amount of advance on the negative side of the RPM delta portion of the idle timing correction curve. But that would upset the beautiful symmetry of that curve (and possibly incur some other drivability problem). Or...
• Go into the ignition table and make some changes in the area (after increasing resolution in the area affected during startup) so that the adjustment incurred by the idle timing correction curve get effectively nullified. And since that area is only entered into during startup, there SHOULD be no drivability impacts.

I monkeyed around with the Crank To Run Taper, hoping that would delay the onset of CL Idle, but that seemed to have no effect. I have no other ides, so if any wiser head has thoughts, I welcome them all.

 

Post your tune. There are a variety of settings that could affect this, from the initial value table to cranking settings etc, that'll help us lead you in the right direction. You shouldn't have to reduce idle advance correction.

 

Tune is here

 

You could even use more idle correction timing. By -100/100 RPM from target, I'm typically at 10/-10 respectively, you're only at 4/-4.

I'd love to see a log of it started and idling at as cold of a temperature as possible until at least thermostat temp without any throttle input (once it's started at least). My initial thought is that your initial values at lower temps are a bit low. It's also weird that you don't have a lower RPM column than 900rpm, typically I see a 750 or 500rpm column. You also target ~975 hot idle, have a 1000rpm initial duty column, interpolate between 900 and 1100rpm ignition/fuel table. The more I look, the more I don't like about this tune. You're sitting at least 115% air temp correction at all times, explains the super low VE numbers in your fuel table. WOT target voltage for alt control is 12.5volts for some reason? Does it actually do this? Does it actually run well and meet AFR targets (which are lean in boost)?

 

Here's a multiple-start log where there are a few fails, and then a start with some coaxing (attached)

If you look at my idle target table it doesn't go lower than 975 so I didn't bother with anything lower than 900 and that was a recent change. before that my hot target was closer to 1050 because I was having problems with an idle droop that would cause a stall when coming to a stop while driving. I got that (more or less) sorted, but didn't adjust the IV table to suit.

"Air temperature correction"? That's something that I may have missed going from the MS2, to the MS3. If you're talking about the MAT/CLT correction here - that was a hold-over from the old tune that I never re-visited. I know that my CLT is too cold and I'm working to address that, so, for now, I just have to deal with that. And those WOT values were defaults from DIYautotune.

The car runs well but it is indeed lean when it starts to get into boost but I've noted that and the VE values above 100kpa reflect the direction I'm going in to address that. I'm using MLV HD Ultra's histogram feature to find and fix things like that (see image). The Z value is the multiplier that needs to be applied to the VE table to get the average of the cell to get to the AFR Target. Those red areas are what my next tune iteration is heading.

 

Your initial value table only ranges from 150 to 190 degrees, You probably need one value at 190, and one at 150, and could easily interpolate between the two, don't need to use all 6 rows for that. It's already a small table, use it for small values.

Like I said, take a log of it ONLY idling from cold to hot. Ideally you take 5 of these logs over the course of 5 days, with your target table locked at each RPM column. You can use these log to set the initial value table, but at the very least your log says you need around 34% at 68 degrees, you're locked in at ~32%. Needs to be a bigger taper at cooler temps.

That being said, you could probably use more fuel while cranking, your settings there are also very low. try taking the entire cranking table and add 50, see if it helps. I'd expect around 150% cranking duty, you're at...102%.