tuning cranking pulse on a cold engine

[nismoautoxr]

I don't usually have any problem using pretty close to the default cranking pulse numbers on engines with closely specd injectors but I am having a devil of a time with this miata on cold starts (70*F) . The car is a stock 1.8liter mazda miata with 650cc injectors from DIYat. At 210*F the cranking pulse that works pretty flawlessly is 25% of required fuel. At the temps lately in my part of the world a morning first start temp is around 70*F. Ive tried 80% there and that doesn't seem to be enough . I bumped that to 125% and it started pretty good that day then the next day it acted as if not enough fuel so I bumped it to 140% and it still acts as if that's not quite enough on one day then it acts as if too much another morning . Is there some formula that is a little more hit than miss than this trial and error I'm using at this moment? The wideband seems to be useless as a indicator at that short of a runtime so I cant really get a good read on it . Here is a log and the tune . Somebody take a gander and see what I might be missing. Maybe its not fuel at all. Maybe its RPM signal ...I haven't seen any indicators of that but who knows. Right now I have cranking RPM at 500 but it has been set at several different numbers from 350 to 500 just taking potshots at it hoping to stumble on the silver bullet

[billr]

Have you flowed the injectors for rate and dead-time? If it will start with any cranking pulse under 100%, that is a big hint that the basic injector parameters are off. I think MAT correction is also included for cranking pulse, so having that faked rich may also be a factor, but it isn't so extra-rich that I think it is the main problem.

[nismoautoxr]

I have not flowed them or calculated deadtime. I went off of the deadtime specs furnished by the vendor. I also used the 2 and 4 squirt method to verify it's in the ball park. Do u really think that it could be that far off to cause this much of a.cranking pulse issue Bill?

[billr]

Well, I kind of got fixated on the higher coolant temps, where the starting pulse is in the 45-25% range. Yeah, that is quite low, but I expect you didn't bother fussing with that part of the table yet. Still, at 70F you are running only 130%; about 1/2 of the MS3 "sample project" (251%?). At 70 F mine is over 300%, but that is partly because I use no priming pulse. Point is, when cranking MAP should be at near-atmospheric, so fuel PW should be 100% to get a stoic mixture, ignoring valve over-lap. Most engines will require a rich mixture to fire when they are cold, so 100% or more for cranking pulse is what I expect. Let's wait for some other folks to chime in... Of course, if the req-fuel or dead-time isn't correct, then little of the above "theory" applies!

[nismoautoxr]

I really am in new territory because I typically don't run a huge injector like this but this car is in preperations to be turbocharged so that is why the 650cc injectors have been installed. I've had plenty of other oddities to deal with on this project too. Way more than any other to be honest. It is entirely possible that my RF and/or DT is the problem. We all know what a pita it is to test the injectors especially if it's daily driver and you are a mechanic for a living and have too many irons in the fire most of the time. If this is what I need to do then that's what I gotta do if nothing else but to check that box. I suppose that the bigger the injector then the smaller the margin for error...that seems to be the pattern that is forming.

[subwoofer]

You could try using crank taper to your advantage, this is what I do and it fires as soon as it syncs. First four compressions get a good shot of fuel, then back to sane values. The advantage of doing this is that you won't flood it like you would with a large prime pulse if you key on and off a few times.

Anyway, the attached msq is for a 2.1l watercooled VW flat four, normal size injectors firing line-of-sight against the valves. Feel free to be inspired or disgusted, it works for me in my DD. Don't trust the < 0°C values too much, have not had enough starts on it in the real cold to know for sure it starts quickly in all situations but it wasn't difficult to get it started in < 20°C in the mountains in February.

[nismoautoxr]

Bill called this one. The deadtime was the.problem. I changed the deadtime from 1.2 ms to 1.0 ms and the car no longer would start. I changed the cranking pulse to the default values and it cranks and.runs fine. Still remains to be seen tomorrow on a cold start but point of this post is I am surprised .2 ms of DT made that drastic of a differance in fuel delivery. Must be these big injectors being that sensitive.

[nismoautoxr]

I would call this solved. the car cranks perfect now. Changing the deadtime fixed the cranking issue I was having. I did a quick retune of the ve table and we are back in business. I still cant use the default mat correction table as is because it just plain pulls so much fuel out that in traffic the car runs like poo. Hot restarts are poo too and its all due to the mat corrections. I put the mat corrections at 100 and its all good . the engine runs fantastic and all is well. put the mat correction back to default and its a lean surging pig. I'm not going to invest much into this because the car is about to get a turbo and the mat position is going to change but I doubt the problem with mat will go away. I could easily correct the problem by moving the mat to an area that doesn't see REAL temps but that's not the point. Either I'm missing something very big or there is a missing correction in the equation for me.