Another closed loop idle question. Engaging early?
 

So I'm trying to "tune" the lower map regions before I turn on overrun fuel cutoff, and found an interesting anomaly and was wondering what I need to do to make it stop happening...or if it's normal.

When I'm decelerating, the VE table sweeps across the bottom row, but then the PWM idle duty triggers and everything goes lean (derrr extra air), but I'm just deceling. I haven't done AE yet, so that's currently off while I tune these areas (ignoring those regions of TPSdot spikes). It "should" be off right during deceleration? I don't know if I should be obsessed tuning this region of the VE table (<26kPa) - but I'd like to see it "be" in the vicinity of the target AFRs before I turn on overrun.

Also, car idles around 40kPa, and seems to decel below 20 kPa.
MSPNP2, 1999, EGO off, AE off, MAT correction = flat 100%.

https://cimg9.ibsrv.net/gimg/www.mia...5aaa7bd34f.png
https://cimg9.ibsrv.net/gimg/www.mia...d7c91cb9ac.png

If the car requires 40kpa to idle, something is wrong. I would start with cam timing. Assuming you idle spark advance is in the 10-14 deg range.

Yeah, idle is 40kPa. I've had a small vacuum leak since the beginning. Traced it all around - replaced hoses...the only thing that hasn't been changed is the intake manifold gasket. Timing in idle region is 10-14.

Check your brake booster. It is often overlooked as a possible vacuum leak. I see many Acadia and Enclave GM vehicles where the master cylinder leaks into the booster and causes the diaphragm to go bad. Doesnt mean that can't happen to a Miata. I would also block off your evap system and see if that changes.

Could you upload a log and your tune? Some print-screens are much harder to analyze.

Also, what RPM are you idling at? If you are idling at a normal RPM but at 40 kpa, then a vacuum leak is not your problem - same as what Ted75zcar is saying.

For the situation in the OP, I attached the tune and datalog...

I've also added the most recent tune and datalog of my idle - (only changes are lowered AE threshold, and turned on EGO back on).

You can see large changes in VE vs RPM at decel loads. The peaks and valleys can be really narrow, like less than 1000 RPM wide. Autotune for them. I see a range of 30 - 70% in my decel rows from 1000-7500. I use 30 RPM bins (16 x 2 with tableswitch, 2 overlapping bins).

In order to tune these regions well, your car and tune have to be super dialed. It is not worth the effort in most applications IMO.

Fix your idle vac. Tuning to this level of detail on an improper setup is a total waste of time.

Thank you for that.

I think I'm missing something (on free version of MLV so I can only see the first ~26 seconds worth of data, and TS so no autotune). On the datalog, on decel loads (7.1-10.9sec, 15.6-19.5sec) I only see VE change from 48-49? I see the AFR spike seemingly timed with the PWM idle duty and was wondering why it's even triggering at all.

And as far as "not worth the effort in most applications" - do you mean just turn on overrun and call it a day?

How would I go about fixing my idle vac (aside from searching for a potential source of a vacuum leak)?

Edit: ah, unless you meant the initial, verify the cam timing on my car (take off shields, look at hash marks on crank and cam gears? and verify after rotations, etc.?)

Sorry bud, I am going to bow out here.

support EFI and get the full versions of these tools. You need them.

There are some MS 101 type threads and youtube videos, I encourage you to have a look. Read the Megasquirt manuals.

Have a nice day

Ah bummer. Indeed, I've looked through many of the tutorials, videos, manuals, and old/new MT threads - these seemed to be specific questions that weren't addressed hence coming here (primarily why my IAC duty was coming on - and if that was normal). And I also just realized how you were saying the previous comment on VE values (aka, not my datalog, but rather VE values in general in those regions).

And I do want to get the full versions of the software - even necro'd an old thread on which to get (MLV or TS full - if I didn't have the means to do both just yet), still waiting.

Looking at 2019-01-21_08.13.51.msl, your closed loop idle settings are pretty bad.

-edit- my bad, didn't take a very close look. Apparently .4% pwm difference for your idle valve makes your rpms oscillate over 100. Seems fishy.

My suggestion would be to fix fuel and closed loop idle while stationary, then take a look at whether your idle is at the right MAP - if it's not, resolve that first. After that works, you can go for a full VE tune.

Thanks. Yeah, my RPMs seem to follow my spark advance. It's bouncing between 10.5-11.6. I should probably get another column in my ignition table. Because right now the cells that it's sitting in, are 10 @ 750rpm, and 15@ 1200 rpm.

I saw one of Brain's old posts on using adaptive idle advance...where being off -200, adds 20 degrees, and being +200 removes 20 degrees (obviously 0 is 0). It sort of worked for me, but I had a noobish question on whether or not having that idle advance running all the time was a good thing or would it wear something out prematurely vs having the timing dialed in without that additional table. My assumption would be that adding in a 1000rpm column with timing closer to 10 would dial that idle better, etc.

But all this still comes to the bit where I have a question on why my PWM idle duty is triggering at all during deceleration. I mean, I know it doesn't effect anything ultimately since I'll have that covered under "overrun fuel cut" - and it was mainly an exercise of seeing ~13-15 AFR while engine braking.

Edit: Looks like I'll be going back to open loop to get the controls out of the equation, and just work with CLT, IAC position, and VE/spark tables. THEN I'll go to closed loop..still not sure if there's another issue with my 40kPa idle...that would possibly mean I would need to close off the valve more...to increase vacuum...and then figure out fuel and spark to still make it run stable, and hopefully in the end, that ends up closer to 30kPa...

My bet is that you still have a vacuum leak. 40kpa is a classic small enough leak. Should be at 29-32.
Could be your vacuum line between the ECU and the intake.

It's been something I've been trying to find since I've bought the car so many years ago. The only thing that hasn't been changed is the IM gasket...and life gets in the way at the moment. Main question would be...does that effect everything else - PWM idle duty settings, etc...?

It is part of the reason why you're going circles with your idle, vacuum leaks affects afr.

I see. Let's say I'm "content" with how my car idles at the moment (fluctuating, high map and all), would you happen to know why my PWM idle duty is triggering on decel? How would I make it not trigger so I can tune the <25map region to be at ~13-15AFR. I want to see if I can run it without having to do overrun fuel cut that'll throw stoich out the window anyways (no fuel).

In the interest of preventing the propagation of misinformation I will chime in once more.

40kpa idle vac is not a leak thing or anything that one should just live with if the engine is idling at anything that would be considered normal, say 800-1000 rpm. If the engine REQUIRES 40kpa to sustain rpms in that normal range, something is wrong. A normal, healthy, properly setup miata motor that isn't cammed will be at well over 2000 rpm, probably over 3000 rpm with 40kpa manifold pressure and no drivetrain load. My car, with a supercharger and a turbocharger, is at 4000+rpm with 40kpa manifold pressure and no drivetrain load, cold. As I suggested in my first post, incorrect cam timing can and will cause this, especially on the exhaust side of things. Ultra rich and ultra lean idle afr, misfire, improper spark timing ... all can cause this same thing.

Do a compression or leak down test. You might have an intake valve not closing completely or something. Maybe bad rings on a cyl.

Looks like I'll be spending a little more time with the engine. Leaving this morning, I saw a puddle of coolant under the car. :violin: Don't see anything obvious for location - so high chance it'll be the water pump. great. It'll give me some time to verify the timing of the crank v cams though!

I've got another thread where I'm losing sync at ~6500 rpms. Maybe this all is coming to be a perfect cam timing storm. Hoping for something minor.

I may just go and do this...as I've asked this particular idle question (back when my car was stock) in many forums to no avail...but this thread seems quite relevant.

https://forum.miata.net/vb/showthread.php?t=640939

I'll double check to see if I had done the timing belt job correctly. Still surprised it "ran this well" - but then again, what I thought to be "just how it was" might have been that the previous owner made the same mistakes that I did, and thus I didn't really notice. Symptoms (of prior to MS) included, randomly shakey idle... intake burble/pops on decel, etc.

Otherwise the car runs "well" at the moment, other than the weird hiccups here and there.