Wayne_Curr's idle woes.
 

(Thread split off from this one: https://www.miataturbo.net/forum/t29965/) -JP



I'm feeling the same way. With closed loop idle, I have to set it to idle at like 1100-1200 when everything electrical (lights, radio, heater) is turned off. When I turn them on, it then pulls the idle down to where it should be (~900). Even with this strategy, i'm getting idle droop when coming to a stop sometimes (yes I have a vtps)

What this means is that during a warm day when I dont need to use my heater or headlights, my idle is high. Its starting to get a little annoying. Something thats also beginning to get old is that I cant get closed loop idle to actually warm up at a higher rpm. Warmup Only works great for warming up, but closed loop only idles at like 1100-1200.

Makes me wish I had just gone parallel and let the stock ecu keep handling it.

Feh?

In closed-loop mode, your idle RPM should be constant whether it's a mild afternoon (no loads) or a searing hot night (A/C & headlights on). You might get a little droop when you first turn on the load, but the steady-state idle RPM should be the same regardless.

What's your Min(dc) value, and what's the typical DC observed in the log during normal (lo load) idle?

To start, i've been playing with the dc values (min and closed) to try and see what a difference they make. Right now they are min:24 closed:22.

I should also note, changing the fast and slow rpm idle numbers doesn't change a thing.

Looking at a log, looks like idledc sits steady at 22.

Then something is wrong. Let's have an MSQ and a log showing some warm, no-load idle, then some warm idle with full load (A/C, cooling fans on, blower at max, headlights, etc).

5 mins, i'll pm it to you so i dont junk up this guy's thread anymore.

Sorry OP, cant seem to attach to a PM. We seem to be having the same issues anyway =P

Joe, my engine has been sitting for 15 mins so its not all the way hot so this isn't as dramatic as it would normally be, but you can clearly see the idle high with no load, and low (800-900ish) with high load. I hit the throttle a couple times but it wouldn't droop for me.

In that log, IdleDC is sitting at 22 the whole time. It's never even getting into closed-loop idle control.

Try reducing the Closed(dc) to 16, and Minimum(dc) to 18. See if that gets any activity out of it. These numbers are probably too low, but it's just a test for now to see if we can make it do something other than sit there with its thumb up its butt.

You can swap one of the gauges in the MegaTune dashboard to IdleDC so you can watch in realtime what the idle controller is doing.

Alright, i'll give that a shot today. As an update, yesterday I played with my idle screw and the idle seems more steady, albeit high. I'm going to re-read the manual on PWM idle as well. I'm not understanding it as well as I need to.

idling in my car right now the gauge just sits at whatever value is in the closed DC box. What kind of behavior should I be observing here?

Ah ha! When you tweaked the air screw on the throttle body, you probably let in too much air. Even with the IAC valve sitting there closed, the RPMs are not dropping low enough for the MS to start trying to control it.

Turn the screw back to where it was, set the numbers in MS like I said, and let's see another log. Hopefully we'll have some activity on the idle controller.

If you find (by watching the IdleDC gauge in MT) that the idle controller is just sitting there at either the Min or Closed value, then you have too much air bypassing the solenoid, and the RPMs are remaining so high without it that it isn't even bothering to activate.

Once you've got it idling in closed loop, you want to observe the smallest number that IdleDC commonly runs at (warm, with no load) and set the Min(dc) number either equal to or one less than that. This will lessen the droop when returning to idle from cruise, as the idle circuit will start out at a reasonable number, rather than starting too low, undershooting, and then having to recover.

Ok, playing with it right now and I dont think PWM is working right at all. If I turn the idle screw down, it just gets lower and lower until it wants to die or just dies.

The idle DC gauge isn't showing anything other than what is in the closed loop box.

Edit: I should throw in some history on my idle issues.

First, it idles fine with the stock ecu but droops BAD when exiting the freeway and pushing in the clutch after a period of sustained RPMs.

Second, when I built my MS, I had left out the insulator on the TIP120. This caused my car to idle at 3k no matter what I did. Needless to say i've put the mica insulator in and now it seems to idle more normal than ever.

So at this point i'm wondering if the damn thing is working at all right now. How can I test this with my multimeter?

if you unplug the idle valve (while idling) is there a change?

So I cant unplug my valve while running because i dont have the tbelt cover on, tried but almost lost a finger.

I did, however, unplug with it off, turn it back on and it was being weird. Seems like the car just wanted to stay where I revved it kinda. In summation, it sees that the idle valve is working atleast somewhat.

Could my TIP120 be bad? It seems the idle valve itself is working. Not sure where to go from here. The idea of having actual working closed loop idle gives me a boner.

Does your multimeter have a frequency setting? If so, try scoping the frequency between the IAC wire on the MS and ground, and see what you find.

I have a cheap MM and unsure if it has a scope setting. If this is something you'd find only on nicer MMs, I dont have it. I'll check though.

I should have just said "measuring" - to do this test, it doesn't need to plot the signal like an oscilloscope, just measure the frequency.

So I was just looking through my settings.ini file to see why my car wont run without my AFM (still working on figuring that out) and I noticed that MT was set for a stepper iac! Wtf?! Changed that to PWM, going to go out and see if that was the plobrem.

Edit: nope, didn't work. And as for checking the frequency, i dont think my MM is cool enough. Either that or I dont know which of the little symbols tests freq. And the car runs fine w/out the AFM, just accidentally unplugged my GND->FP jumper.

I'm actually having similar issues as well with my MS1 build. The IdleDC value never changes so in order for my car to idle less than over 2k RPM, I just leave the IAC valve disconnected and adjusted my idle screw to give me a nice 850 RPM idle speed. As soon as I reconnect the valve, it shoots back up and never changes.

I have check the continuity and everything on my TIP120 and it checks out fine.

Any ideas?

Could you please post your MSQ, Karter74? I can have a look at the settings. Also, how exactly is the IAC wired up in your parallel install?

Attached is both my MSQ and a log file. I should also clarify, this is another install I'm doing for a friend, so it is a complete standalone install, not parallel like my car in my sig. This is from a 95 M-Edition, 450cc DSM injectors (with resistors), and a GT2560.

Ok, looks like this is going to need a bit of tweaking. The settings can be a bit confusing, and the fact that the "dc"s in the settings are really pulse widths doesn't help. The way the control works is that the "Frequency" value is inverted to become the maximum pulse width. Any DC value greater than the Frequency number is actually full open. So on your MSQ you can set the DC values to anywhere between 0 and 26, but 26 or higher correspond to fully open.

Lower your min and closed more. Mine are at like 6 or 7.

I tried this and my car started idling very strange and very high kind of like when I unplugged the idle valve all together.

I think i am just having a really hard time understanding how the idle settings and the idle valve work. Should I just close up the idle screw all together and start playing with the settings in megatune to find what works?

Also, i've noticed that I can change the settings to lower numbers without restarting the squirt, but if I try and increase the values I have to restart to get the changes to take. Is this normal?

I'm not heartbroken that my idle doesn't work 100%, it gets me by just fine the way it is but i'd like to know what i'm not doing right.

After some experimenting, I still have the same issues unfortunately. With the idle valve disconnected, I can watch the IdleDC gauge change values as if Megasquirt thought it was actually controlling the idle.

Appearing to act normal when disconnected from the IAC valve, I plug it in. Immediately, the idle shoots up to about 2400RPM and the IdleDC gauge just sits at the closed value. I attempted to completely close the idle screw, but it changed the idle minimally. I used my Harbor Freight multimeter on the connector and it read what appeared to be battery voltage (13.xx V) and a frequency of .26 KHz, appearing to fluctuate somewhat randomly from .19 - .26 KHz (but I doubt the accuracy of the multimeter).

I also hear a slight whistling sound coming from the IAC valve when disconnected (but engine idling). I doubt any vacuum leaks as nothing has changed since I bought the car and it idled fine with MS2extra.

So unless something has been fried on the idle valve itself, I'm somewhat dumbfounded as to what is going on.

Please help!!

Very interesting that you got the idle dc to change by unplugging the valve itself. I have no idea what that means nor do I know where to go from here. I want to try that myself though.

I actually observed a wackier idle with my idle valve unplugged vs it being plugged in. When I punched in a 6 for closed and min DC it acted exactly as it did when I unplugged the valve itself.

Anyone else have ideas what's going on? Matt, your input would be highly appreciated.

Yes, and Joe's input as well.

I'm about to take a shot in the dark and try replacing my TIP120. I dont know how to test it for functionality and its cheap so i'll give this a shot tomorrow and see what happens.

I think my biggest problem is my great lack of understanding about how this SHOULD work. Without knowing how the whole thing is suppose to work I cannot troubleshoot a goddamn thing.

Played with this more today (haven't replaced that TIP yet like I said I would). Here are my observations from today:

First, i tightened my idle screw all the way so it wasn't playing any role in this anymore.

I noticed on the DC gauge that i'm getting a change from cranking DC (35) to my closed dc when starting my car. This is the only time I ever see this gauge move.

I set my min DC to 7 and played with my closed DC a bit. Got it to idle real nice at 850-900 with no accessories turned on. Turned on the headlights, the idle droops slightly, then turn on the fans and it droops slightly more but really not too much. Revved it a few times and it seemed pretty stable. Something is still not 100% consistant, not sure why though. Sometimes the idle will settle a little too high, sometimes a little too low. If I blip the throttle a bit i can get it to try again and get it to settle properly.

I'm still getting droop if I push in the clutch from high rpms. If I play with the throttle and ease it back down to idle it doesn't seem to be too big of an issue.

Correct me if i'm wrong, but here is what i'm understanding about this PWM valve and MS. When I turn the lights on or some other thing that droops my idle, the DC gauge should increase a little bit to compensate until it settles, correct? In other words, as things change that affect my idle, i should definitely see the idle dc gauge doing something. Is there a difference in MT between the "Idle DC and the "IdleDC" (note the space)? I have both to choose from, both seem to do the same thing.

So overall, with a min of 7dc and a closed of 10 i seem to be seeing the best idle thus far. if I lower the closed to 9, the idle is too high. Higher than 10 it idles too low.

In other news, i finally turned on TPS accel enrichment today. OMG it drives like butter now.

Interesting stuff. I'm still hoping Matt from DIY sees this and can add any input. I'm still suffering from the same issues as I stated earlier.

Yes. You should see the IdleDC gauge moving around when the car is idling, and when you are not idling, the gauge should display the "closed" value.

IdleDC and Idle DC are the same thing. If you open the msns-extra.ini file you see that they both point to the same variable.

Forgive me if this is redundant, but a quick tutorial.

The "closed" value is the duty cycle that the MS feeds to the IAC valve in order to close it fully. When you're not idling (ie: you are driving) the IAC valve should be closed, so the MS sends it the "closed" value. When you drop down into the realm of idle, the MS goes to the "Min" value and then starts seeking the target, and you should be seeing movement on the gauge during this time.

Now, you're seeing the "closed" value all the time, and furthermore, altering the closed value is altering your idle. This means that the MS is never even attempting to go into idle mode- for some reason it doesn't think you're trying to idle.

This is not a hardware problem. At least, not in the idle circuit itself. Although 7 is an astoundingly low idle DC value. Mine runs best in the low to mid 20s.


I would like to try an experiment. Change your idle screen as follows:
http://img18.picoodle.com/img/img18/...em_f2723a2.gif

This allows us to directly input static idleDC values for warm and cold idle, rather than letting the MS seek a target RPM. See if this gets the IdleDC gauge moving to a value other than closed.

Thanks for the review Joe, I think I understand much better now. As you say, it really doesn't seem like the MS wants to get into idle mode.

K, sitting in the car right now.

Warmup only with those settings is different. It started at 26dc, then down to 25, then 24 continually to 22 where it sits now. (the engine was 140* at startup).

Droop is way worse with warmup only, turn the lights and heater on and it is down to ~500rpms.

This is an interesting datapoint.

From everything you've said, it seems that the hardware is working, and the MS is capable of controlling it. However it is refusing to go into closed-loop idle mode when you are in the correct range. Somethin' screwy.

A question- have you tried just re-flashing the MS firmware with a fresh, known-good .S19 file? It's almost like something is just fubar in the operating code.

Yes I have reflashed recently. .s19 is what I create with easytherm, right? I redid that with my last flash as well.

I've had a few beers tonight so my critical thinking skills are not up to par. Perhaps i'll try reburning tomorrow. The idle valve works with the stock ecu just fine so i am reasonably sure that is not the issue. Are you sure its nothing hardware related?

Major Breakthrough...

TPS Threshold ADC was at 10, needed to be higher. This is why the MS was never even trying to control idle.

Playing with it right now but may have some questions shortly.

Question 1:

Idle DC gauge seems high. Was over 200% now at 130, looks to be dropping as I warm up. Is this normal

Now it either just sits at the min or closed value. If I turn my lights/hvac on it still droops, DC changes but it doesn't quite pick up all the way back to normal. Also, it is still idling high in the 1100 rpm range. Changing the values for slow and fast idle has no effect.

Ha!

I was looking at the log and saying "no, this TPS is going to 0 when he's off-throttle" without it clicking that the log only contains scaled TPS, not raw TPS ADC, which is what the idle screen uses. I guess the mechanical calibration of your TPS is somewhat askew. :)

Lets' have a new set of MSQ and Log, so we can see what it's doing now. I'm a bit confused when you say that the idle gauge is showing 130-200%. The gauge is only scaled from 0-100%, and mine never goes higher than the mid 30s. Can we get a copy of your msns-extra.ini file as well? It's in C:\Program Files\MegaSquirt\MegaTune2.25\mtCfg\

Yes, the counting down is mighty strange. I noticed also that after I change any setting to do with idle the gauge goes up to 250% and quickly counts down from there and sits at either the min or closed value.

Here is an .MSQ and my .ini for now. Will grab a log later.

Edit: Cannot attach .ini

Yesterday I tried playing with my idle some more. It never goes below 1100 and usually just sits at 1300 these days. Idle screw is all the way closed.

I pulled the IAC plug while it was idling, no change. Plugged it back in, no change. Unplugged it again and revved the engine a bit, no change. So it looks like my IAC doesnt work anymore. Could I have broken this messing around with MS settings? Seems like it must be stuck open for it to idle with the idle screw all the way closed. I'm reasonably concerned that someday its going to decide to stick closed when I really dont want it to.