Bad Idle, But Only Sometimes
 

Hey everyone, I need a hand on getting this weird idle issue figured out. I had my car tuned by a company, but the idle / CL idle seemed a little underwhelming.

I'm very much so still learning but I've changed several settings and have about a halfway consistent 900 rpm idle. The issue is when the car enters closed loop at 1200 rpm, the engine rpm will start to get jerky. It will drop down to around the 900 rpm target but still it will almost seem like the engine 'coughs' and the rpm will rapidly drop. Oscillation isn't the right word, and i'm not sure if misfire is the right one either. Anyways, it's random. I could be coming to a stoplight and it could enter idle great - no pops or jerky engine at all. I could then creep forward or just rev the car a little, and it would do it's thing where it coughs and has a rougher idle. I've tried to turn closed loop idle off, and still experience this. I had been using adaptive idle advance before but it made this much worse so I've turned it off until it's figured out. When it's got the weird idle, when I give it gas to take off from a stop it stumbles momentarily when I give it gas. When it's idling great, I can take off smooth as butter.

The setup:
Poverty DIYPNP with GM AIT sensor, batch fire. The car is a stock engine 96' with a Flyin' Miata based turbo kit. Flow force 640cc injectors. Newer spark plugs. The engine has good compression and leakdown numbers.

The log:
Idling poorly, throttle blip, then the idle settles in perfectly.

Worthy mentions: the car idled beautifully on the stock ECU and when I switched to the DIYPNP it did this weird idle thing on the stock injectors and the flow forces also. The cars alternator squeals a little bit on startup.

Thanks!

"The cars alternator squeals a little bit on startup."

-Get a new belt, tightening it probably won't do much by now. You can try and see if it fixes it.

Sounds like the PID is off a bit and is causing the closed loop to overshoot. I'm not able to look that the tune or log at the moment. Just a guess.

I'm 99.9% sure it's the bearing inside the alt rather then the belt. I did a timing belt service maybe a year and a half back, replaced the belt with a brand new Gates and it's doing the same. For fun I went and had the alt. tested and it came back good from an electrical standpoint. I am not going to discount anything so I'd be happy to try a new belt.

I don't disagree that it could be the PID, but since it's happening with closed loop disabled wouldn't that sort of rule out the CL settings?

"I don't disagree that it could be the PID, but since it's happening with closed loop disabled wouldn't that sort of rule out the CL settings?"

-Yeah idk then since I can't look at anything. I get the idle overshoot occasionally but i haven't felt like fixing it because it's not often for me. I need to do it one of these days.

As far as thinking its the bearing, I would be surprised if it is. Take the belt off and spin it by hand, that will tell you if its the bearing or the belt. How do you know the timing belt is doing it as well? Are there squeals from two different areas on start up? Do you have A/C and PS?

I'm bad at explaining stuff. :loser: I replaced the alt. belt while doing the timing belt service. When I had the alt off to get it tested, I spun the bearing by hand and could hear where it would squeal. I really should just replace both at some point. But yes, I do have both AC and PS.

With any luck we'll both find a solution. Thanks!

It is just me, or does it take a second after you blip the throttle for the revs to jump up? In the logs there's a a half second delay between you pressing the throttle and the RPM even starting to move, and a full second between your throttle depression and the revs jumping to 1800. If that's true you'd definitely feel it.

ASE or VE too rich and choking it out?

Hmm,

You could bump up the 90RPM 10 MAP cell of the ignition table to make it smoother. Could also add a column to the left of your target idle column and bump it up a degree over what you are targeting to help with idle droop. It will simulate what the idle advance does for load, as far as i understand. I don't see anything crazy with the tune or log unfortunately. It looks like its hunting in the beginning part of the log, which is what is bothering you correct?


"It is just me, or does it take a second after you blip the throttle for the revs to jump up? In the logs there's a a half second delay between you pressing the throttle and the RPM even starting to move, and a full second between your throttle depression and the revs jumping to 1800. If that's true you'd definitely feel it."

-Definitely looks like it has a decent idle dip before and isn't triggering AE till the last second when it triggers MAP AE.

Does the car free rev well if in neutral and giving it a little blip? Or does it hesitate OP?

Sorry for the delay, I wanted to put some miles on the car to help answer the questions better. The idle dip that you guys mention when I get on throttle is absolutely noticeable when it's hunting and I rev it. It dips significantly, and the engine sounds like it's breaking up/missing until the rpms rise above abou 1000. If it's idling normally (which it does most of the time) I don't notice really any hesitation or delay. Maybe a SLIGHT delay? But certainly no dip or anything of that nature - it revs smoothly like it should. I will try your suggestions Tuna and report back. I agree that what you're saying about adding a row maybe at 750 with higher timing could help.

Thanks again!!

No problem buddy! Hope it works out for you.

"when it's hunting and I rev it. It dips significantly, and the engine sounds like it's breaking up/missing until the rpms rise above abou 1000."

- Not surprising at all. It's struggling and then a bunch of air is shoved into the motor. Something is making it the mixture lean imo. Once again all guesses. Hope smoothing out that idle area of your ignition table helps out.

That's really strange and hurts my brain though. Symptoms sound similar to my car on a hot restart before I got the issue resolved. Seems like its running lean but the logged AFR looks fine. Is your wideband showing the same as tunerstudio?

Had an idea last night after a couple beers so take it with a grain of salt. Is there anything you can think of that could be correlating with the idle droop? For example, mine happens at low rpm while generally still kind of cold. So I just need to bump up my initial values a bit I think. Any common themes you've noticed it happening with more often? Is it generally only when you are going slow or creeping in traffic?

So, I've got some updates and decent news. I went ahead and messed with the acceleration enrichment and got that dialed in a little bit better. I did notice that on tip in I was going pretty lean and there's still work to do but it's way better. TS and my AFR gauge do match each other great. I also bumped the 900rpm/10kpa row up to 13 degrees. On the initial values table.. I actually lowered the numbers by a few ticks close to the 900 range. I think this is all starting to really help. It's still happening but it's smoother I think.

I thought a lot about what you said and driving around today, I was trying to catch more about what's happening.. to help understand the why. I watched close in TS and the car does enter CL idle every time, but if I'm cruising and then come to a stop to idle (doesn't matter how, after cruising, boost pull, revving, load/no load) when this weird behavior happens it's like the idle drops below the 900 target then comes back up to 900... and then it's just like something fights CL idle from doing its thing. To get it right I have to either just to drive and hope it works next time or rev the car probably 4 times or so for it to 'catch' and drop down to target.

On the flip side, when it works beautifully and I have a rock solid idle, CL catches the rpm's at about 1200 and ever so slowly/gently drops it to 900. I notice in my tune the initial values table goes from 1300 to 900. This is the part of me that doesn't know how to tune all that well, but I may try bumping that up to a higher rpm to give it a broader idea of where it needs to be as the rpm's are dropping. Maybe I am keeping the car in gear for too long as I am coming to a stop, not giving the IAC enough rpm range to react. I dunno. I'll keep this updated.

Don't focus on AFR it's almost worthless for tuning AE. The change happens so quickly that the gauge will be delayed and will pretty much always show a lean spike. You're goal is to get the rpm's to rise as close to the tip in as possible. Sounds like you're about there! Great job!

Was CL active on your initial tune? I thought you said it was off so I didn't even look at the initial table values.

This is your initial values table being correct, the other example of when it's off, this is what you want it to be able to do. It catches a little higher when dropping rapidly from a higher rpm, it tries to catch lower if its falling slower from an rpm closer to your idle target.

You'll get this fully sorted in no time :)

Have you ever adjusted the idle screw and set it just below where you want the idle? Might help as well as a mechanical stop