Idle settles then surges
 

Happy Holidays!

I've been having an issue and finally decided to track it down, hopefully you all can help.

When I let off the throttle the idle will settle down to a normal RPM (900-1000) but a moment or two later it will surge to 1800-2000 RPM. Occasionally it will settle back down on it's own but usually it just hangs there indefinitely. If I artificially lower RPM by turning the fan on or holding the power window switches down it will lower 100 or 200 and then drop back down to normal idle. I'm sure I'm overlooking something obvious but I'm all ears for any input.

Attached is a log where you can see TPS at 0 and RPM will jump seemingly on its own. Feel free to critique my stupid rich street tune as well.

Thanks for any help!

You need to post your tune.

I don't get it. I don't see a jump anywhere that doesn't have an associate TPS with it.

My bad. Problem has existed as long as I've had this tune that I can remember. At least the last 6 months. My car has been in various states of disarray for the last year though so I couldn't pinpoint a specific time it started. Definitely prior to the turbo install a month ago.

94 Miata, 96 block, 01 head, 99 intake manifold, DIYPNP (microsquirt) ecu, RX8 injectors, T2 turbo, eBay intercooler, AIT sensor mounted post intercooler outside engine bay, 2.5" exhaust, IWG without boost controller for now, EWG block off plate, VVT and VCIS not hooked up.

My well running but very safe tune is attached.

Mudflap, see 19, 27, and 36 seconds. Clutch in or neutral, TPS is 0 and yet RPMs climb each time

Now I'm intrigued. I can see that your idle duty % goes to 32.5% when your closed loop kicks in. But it clearly has a 930 RPM target and is doing nothing to reduce that idle duty.

I got nothing. Sorry for the wasted words on the internet.

Ended up messing around in my closed loop idle settings and that fixed it right up. I feel like a pretty big tool for not checking something so obvious. Thanks everyone for all the help!

don't leave us hanging, what solved it?

Funny enough I got an automated email from the forum saying there was a new post by you (Mudflap) recommending I change my closed loop idle settings from "use initial value table" to "use last value". I made that change and after one high idle incident it settled and then after a restart it did it once again so I modified the duty cycles of the IAC and now it runs like a clock. Not sure why I'm not seeing that post now but you definitely nailed it on the head. Thanks for catching my goof up!

That is hilarious. I wrote up a post explaining why you should change it to "use last value", but then I looked at my tune (which is pretty decent) and realized that I was in fact using "use initial value table" with your "initial value" table. So I figured it was the wrong approach and I edited my post to "I don't know".

Glad to help in some strange way.

Use last value is not as good as a properly set-up initial values table. The value of the IV table is that the IAC goes to the right value the first time after a start. It also goes to the right value after say, your start and get on the freeway. The next time you go to idle, will be fully warmed up, whereas the "last value" was when the car was cold. Not a big deal, but a properly set up table puts you at the right place every time.

Set up your idle RPM vs CLT curve so that it falls from 1500 @ 0F, to 850 @ 180F. This give the next steps vital information.
Set your Initial values table in "Closed Loop Idle Settings" to respond to "Use MAT".
Also, on the "Closed Loop Idle Settings", make sure the "PID Delay(s)" and "Crank to Run Taper(s) are set to "3". Or leave them higher as you have now.
Then, continue to run what you have been running, but take logs, starting cold. The values used by the MS, when applying "use last value" can then be used to populate the "initial values" table.

Looking at your log, your CLI is not really controlling at all. Possibly because you have the slider so low. It is engaging at the table initial values, and then simply driving to your minimum opening of 15.3. Then the initial values is re-engaging. I'm not sure why that is, instead of starting to control as your RPM gets close to target. Possibly because your Initial values are so way too high?

As a reference on CL PID settings. On my 99, I am running Advanced with slider at 2394, P = 60, I = 55, D = 0.6. In truth that may not be all that far from your low slider, but 100% P & I.

And Mudflap is right. In your log, even the places you mention, you touched the throttle. The surge is occurring WHEN CLI kicks in. It happened because your Initial value is way too high. Then you touch the throttle and the car actually comes out of CLI. Then, when it comes back to the initial value, surge occurs again.

You can fix all this, if you wish to spend the time. It would be interesting to see how your Basic CLI parameters run the idle when they are allowed to work, using your "use last values" to set the starting point of the CLI.

For any future searches, I did resolve this with the initial idles value table. I'm not sure why but the base map I used had very very high numbers in there. Thank you very much DNMakinson!

Certainly. We all once were where you were: stuck on some aspect of tuning MS and not seeing a path forward.

Glad you got it solved.