BBro

I'm working on tuning out some weird startup behavior and have found differing answers for what the function of some settings are.

The settings in question and definition from manual (https://www.ampefi.com/wp-content/up...evo_manual.pdf) :

Crank to run taper: - How long after starting the MS3Pro will wait to engage PID control.
Pid Delay: This is how long in seconds all other conditions for entering PID control must be met for before the MS3Pro will engage PID control. Set this as low as possible without being below what is normal RPMdot jitter with the engine RPM not changing. Typical values will be between 50 and 75 RPM/sec. (In tuner-studio this is a setting in seconds, not rpmdot, and its my understanding its a lock out of how long cl conditions need to be met before going into cl idle)
PID ramp to target time - Once PID engages, the MS3Pro will gradually move the target value from the current RPM to the target speed.This can be used to help larger P-values be used, making it easier to tune PID to catch sudden drops in idle speed.

So my understanding is after startup the IAC will stay at cranking duty until the crank to run taper time is up. Next if CL idle pid conditions are met it waits your pid delay in seconds before engaging cl idle. Next your IAC finally tapers from cranking position to the position it requires to hit CL idle target utilizing initial value table and this will happen in the time allotted by the PID ramp to target time.

Is my logic here right because according to my logs I am wrong but I cant find a better clarification.

BBro

Bump.

Included is a startup log and the tune used during it.

This had a crank to run taper of 5s but about 2s after cranking rpm was surpased the IAC duty dropped straight from cranking to the initial value table with no taper. What I'm trying to get rid of is the massive rpm dip after startup and this is playing a large part.

18psi

iTrader: (76)

try disabling the iacv voltage comp as that may be causing the dip.

DNMakinson

iTrader: (3)

Remember this? On your log, you are hitting -1600 RPMDot at 5.47 second mark. It goes even lower even before the PWM DC falls. I'm not sure why that first fall is occurring so pronounced, but I think the cause / effect of your issue is the same as what I saw.

The PWM DC is actually tapering correctly from 3.8 to 5.4 seconds. I would try an even more negative slow deceleration and see if that helps maintain the taper. Then figure out why your flare is dying so quickly, and solve that.

18psi

iTrader: (76)

I see this a lot in 1.5.1 fw 01+ cars. silly engine states. I wish they would be done away with as they were never needed in the first place and cause all sorts of issues.

BBro

Thanks guys good tips all around. Tomorrow I will start with a much more negative slow deceleration threshold and see if I get the expected taper behavior then start chasing the issue of why revs drop so aggressively in the first place.

FWIW I noticed the engine state would enter decelerating after startup but I assumed the startup code ignored engine states until the car entered normal running code

BBro

Logged a first startup of the day using -2000 as my slow deceleration threshold and got the expected startup behavior with no dips. On this startup there was no large negative rpmdot to begin with so I'm not convinced its fixed so I'm gonna get a few more logs with various slow deceleration thresholds to see if this was truly the fix. Have to wait for the car to get cold for these logs because it only seems to happen on the first start of the day.

18psi

iTrader: (76)

the log looks great. but with slow decel at -2000 I don't think you'll be engaging overrun fuel pretty much ever. but testing is key

BBro

I agree with you about -2000 messing other things up so after that log I switched back for my normal driving. I mainly just wanted a proof of concept if that would fix it.

Today I tried -1000 and the poor behavior on startup still happens because I hit around -2000 rpmdot after startup and the engine went into decel for a moment.

I'm gonna keep incrementally bumping this rpmdot to a more negative value to see where this issue stops between 1000-2000 rpmdot.

What I'm not understanding is the initial dip that actually kicks the engine out of idle state happens in this log and my first logs but not in the 2000 log.

BBro

Been doing some more experimenting and the big dip after startup doesn't seem to be due to the car leaving idle engine state or the iac duty. You can see in this log that the car never leaves idle engine state (-2500 rpm dot slow accel threshold) and the iac also behaves as it should but the dip is still present. It seems the dip follows the injector duty cycle pretty linearly so I am going to try to add more ase fuel and see what happens.

18psi

iTrader: (76)

more? its already at 10.40

DNMakinson

iTrader: (3)

That is what we are trying to do. We put the big -RPMdot so that you stay in CLI, and the taper works. Then you could concentrate on the big dip. After that gets solved (sorry, I don't have a clue) then put the -RPMdot back to a more reasonable number.

Have you seen an issue with overrun fuel cut not coming on with trialing throttle while you have the big negative? 18 mentioned that it might be an issue. Asking only for educational purposes.

DNM

18psi

iTrader: (76)

I would also play around with cranking timing. sometimes if you run too much it will spike up and then dip when it transition to commanded idle timing. but too low and it will take longer to fire. there's a very delicate balance

BBro

I didn't know how accurate afr was immediately after startup so thought it was worth a try. Just interesting that even when its reporting so rich the rpm still rises when the PW rises.

Yep it was an easy way to confirm that those weren't the issues. I would turn slow decel to a more reasonable number after the startup so never got to see if overrun would work with the larger value I will leave it large next time I go out for science.

Thats a good idea, I will play around with this on the next few startups and report findings.

BBro

Took some of Vlads advice and messed around with cranking timing and smoothing the transition as it enters the normal timing map and cl idle timing and I think that is the ticket.

Here is a log of me increasing my cl idle advance and it really improved the dip. I'm gonna work a little more on smoothing this transition and I think it should be good.

@DNMakinson do you have a log of your startup I could look at to compare?

DNMakinson

iTrader: (3)

I will take some fresh ones tomorrow. Not the same at all CLT. Likely to be able to make them more equal, but no motivation. I’m enjoying not having the computer constantly in the car.

18psi

iTrader: (76)

Look at you learning to tune thats awesome man.

DNMakinson

iTrader: (3)

The msl has a 75* first start, and then, at the end, a 185* restart. I paused key at ON to allow re-connection with computer. Interestingly, the Priming Pulse did show. Sometimes it does, sometimes not. I think it depends on the hold-up caps on the MS3. Anyway, the huge pulse just as the car starts at 185* is likely the effects of EAE. FW guys say it is active always, but I think the Wall Fuel build is active during cranking, but the control functions are not, thus ATW and SFW equations actually affecting PW come on suddenly at the transition from Crank to Run. Kind of makes sense, as you want to control PP and CP directly. I digress.

As I mentioned before, I get a flair at 75 that I do not get at operating temperature, but both start fine for me, and I've not tried to sort that out. Also, I have gotten that shelving at around 800-1000 for a few years now. Reverant suggested a 36-2 timing wheel to fix it, but that made no difference. Still, he and 18 might be onto something; that it is related to spark advance in some way.

Off Topic: My cycling partner's son, who just started ME at AL: Rollins Baird.
Off Topic 2: My Corbeau LE-Pro's came in. Will have to make brackets and weld them to sliders.

DNM

BBro

\



Thanks for the logs! I understand the feeling of not dragging the laptop to the car. Lately I only ever log the first start then the laptop goes away and its been nice. I think I'm also gonna lower my IAC cranking duty because the car revs so high after start and enters a ign zone with more advanced timing.

I will ask my roommate who is a senior ME if he has heard of him! If he ever is struggling with class let me know, I can set him up with the right people to get help. I'm interested to see how you mount the Corbeaus, have you looked at the flip side brackets?

DNMakinson

iTrader: (3)

I will mount them like Revlimiter did. See revlimiter.net website and search Elise.

Not looking to buy ready-made. Although that would get me there faster.