Mech5700

Good deal.

I'm off today too.

Mech5700

A couple rough starts from this morning:

On the log for this one, I was seeing the spark advance start spiking when it would hit the low spots on the RPM graph. I raised the crank rpm to 415 so it wasn't trying to advance timing too early. Engine temp 88*


Here is another. Engine temp at 95*, second start.


There were a few more, even uglier starts, after that. But the last fully warmed up start attempt was better, CLT 200*:


I still want to know why the relays are clicking of/on when cranking. What it looks like in the logs, is it thinks the engine has stopped running, "RUN:Y" switches to "RUN:N".

The first 3 logs are the 3 videos I uploaded.

Braineack

iTrader: (62)

RUN:N will always be on if the RPM is lower than the cranking rpm is set, in this case 375RPM.

after work ill work on a new map for you. this is sad cause it seems to run very smooth once it catches.

Mech5700

RPM @ cursor is 202, crank and run both indicate "Y"

Mech5700

Dude. I know. It's a curse, bittersweet. I dread starting this thing but it runs and drives like never before. A ~25% increase in fuel economy, smoother across the board, it's just a better driving car now and I love it.

I've spent hours and hours reading, trying different things, changing settings, an assload of wasted fuel for start/idle tuning, and I'm still where I started with this issue... I have learned A LOT in the last week or so about this, but still feel retarded that I can't figure this out.

I just got off the phone with FAB9, inquired about an issue I saw that came up before about MS3 and FAB9 COP igniter incompatibility. He suggested bumping the cranking dwell up to 8, instead of 6. Also suggested cleaning the plugs and gapping them to .35-.38, so I did. I gave it a shot and it worked better, but I then reloaded the base crank PW% values:

10*---326%
30*---301%
50*---276%
70*---251%
90*---226%
110*---201%
130*---176%
150*---150%
170*---126%
180*---101%

and this was the result (CLT 160*):

Braineack

iTrader: (62)

that's much better. I was wondering about the COPs and dwell settings as I have no experience with the FAB9 ones... glad you inquired. I'll still get something new for you to try today.

Mech5700

Ok. I'm looking forward to trying it out.

At least I can be a test dummy to better your future builds, maybe even be added to the Gravy Train guide...

-8ms cranking dwell w/ FAB9 COPs
-VSS input to Tableswitch instead of N2O in
-eventually a better basemap for the NB guys
-make sure the harmonic balancer hasn't walked to set base timing
-change the LC2 output 2, to the proper voltage settings.

I'm sure there are some other things we can take from this experience lol

Braineack

iTrader: (62)

dont hate.

Mech5700

lol not hatin, just sayin

Braineack

iTrader: (62)

try this one.

Mech5700

No good, wouldn't catch at all :( I switched back to this tune (attached) and it started. ECT was about 170*. After it was running on my tune, I was comparing the 2 and saw yours had 10* of cranking advance. Reloaded yours, set crank advance to 0* and then it started on your tune, slightly warmer than before.

I'm gonna try my tune on a cold start tomorrow morning. Then yours cold start the following morning and see the difference. I see you have the crank pw% cut way down. It may have been too lean when I tried yours first just now.

On mine, I can start it nicely at 160* and crank PW% at ~140%. So I kinda tapered off the curve once it hit 155% and will make adjustments when I start it at different temps. I don't want to experience what I did yesterday morning again...

bmxfuel007

iTrader: (21)

Yes, I know my previous posts were a bit off since I was basing my settings off NA settings, rather than proper NB settings :P oops

But have you tried also adjusting your idle valve during cranking, like the increasing the Idle Cranking Duty/Steps table?

Also in the closed-loop idle settings, right under the dashpot adder, you have "Use Last Value" instead of "Use Initial Value Table" selected. If you use initial value table, it gives you an initial values table that you can adjust for a better idle

It's just more stuff that could affect your idle, after I looked through some hard start threads. Might be of interest if adjusting only fuel settings haven't helped enough

stuff like in this thread, despite the different injectors, etc.
https://www.miataturbo.net/megasquir...5%2A%2A-74991/

Braineack

iTrader: (62)

okay, that's actually good to know. Makes me think the biggest factor was the dwell settings for your coils.

I copied y8s's last startup tables where his cranking pw% dips well below 100%, so if it didn't start it really just didnt have enough fuel.

I would suggest you still use the fuel table I smoothed out for you from that. and I might have something better tp try for you tonight based on the settings that worked for you and other little tweaks.

Mech5700

Had a little bit of struggle this morning. Ect 85*, crank pw% @ 155%. It even did the "lockup" a few times... how can it start beautifully at 160* and pw% @ 140, and then when it's 70* cooler and only 15% more pw, it acts like it's flooded? Is there another factor I am missing? Does ASE or WUE effect cranking as well?

Braineack

iTrader: (62)

crank with the throttle at WOT (flood clear) to determine if the fueling amount is causing that; it shouldn't be.

The only time I've experienced the cranking "lockup" like that is when I'd have the spark outputs incorrect (using inverted instead of non-inverted, or wrong firing order). I've completely flooded the motor during cranking (no map sensor plugged in) and it still cranks with ease, just will never start, then if it does ignite scares the ever living **** out of you when it backfires.

so one thing I'd also try is using stock coils...

Mech5700

Firing order is set to 1342 and ignition input capture is "rising edge", and output is "going high."

Bryan from FAB is sending me the updated coil driver to try, there have been some issues with the older coil driver module playing nicely with the ms3.

And i still have the old coil pack I can try as well.

Mech5700

Check this out. I did some looking and found this Bosch 0 227 100 211 Ignition Control with MegaSquirt-II and a few other threads about rising vs falling edge. Seems like there are varying opinions about this.

If you look on the link, it suggests using falling edge for ign input capture for a similar igniter to the one I am using.

Look at the frist log "falling edge". First start is falling edge, no sync errors. Started very well, but had slight hesitation when attempted to rev. Second and third starts are rising edge, and both have 4 & 3 sync errors, and no start at all. Last start is falling edge and it once again, it started great but also had a slight dip in rpm before reving up.

CLT started at 106* and was 117* by the time I was done with this log, and crank pw% is a flat 140% so I know this can't be because of crank pw%...

I tried a few more logs with falling v rising, and slightly different trigger angles to see what happens.

Also attached this mornings struggle.

Braineack

iTrader: (62)

dont change settings you dont understand. im surprised it will even start when you flipped it. falling/rising is how the MS decodes the crank/cam sensor pulses.

also don't read microsquirt guides to help with your ms3x.

there's actualy a warning for that page:

Mech5700

I understand that warning, and I wasn't even following it as a guide. I just came across it when I was searching for more information on rising v falling edge and saw that it was relevant to my igniter, just so I would understand these settings.

That being said, I verified base timing, set it to falling edge, this is what I got:


The best start this car has had since I've installed this ECU.

Log and tune attached.

hector

Might be kinda off topic but how does Fab9 deal with reducing dwell to these coils so they dont burn out with dwell of more than 3ms?

When I swapped to these type of coils there were warnings everywhere regarding max dwell when researching what else to change so to not burn these up. It seems to me that the igniter they use must be limiting the dwell to the coils for use with factory dwell settings of 4.5ms that a stock ecu sends.

So how do they have a different igniter for stock ecu vs aftermarket ecu that cuts down the dwell?

I know you said you had to revert to the stock ecu the other day in order to get the car started and it started well. Assuming the base maps from diyautotune are using factory type settings for dwell control, their recommendation is 7.5ms cranking dwell for stock coils.