Hi, I have a 2001 with the MK kit, FF injectors etc. Car had been running fine. It started stalling, only the first couple minutes of running or so, only when the ac was on. Then it started with the ac off. It starts right back up, but may stall 2 or 3 times. Doesn't matter if car is cold or warm, just sitting for a while.

Car used to idle pretty well with ac on.About 1050 rpms. Lately it oscillates between 1300 and about 300 rpms.

Lately I noticed that if I'm at mid to upper rpms (4-6k) and I stop with just putting it in neutral it will stall. It has, within the last year, new cam sensor, crank sensor, fuel filter and recent new plugs gapped to .025. This issue started before I changed plugs. 157k on car.

Seems like a fuel problem, but I have no idea where to start looking or how to test. Any suggestions?

 

Can you post a log taken when these problems are occurring?
What ECU?
Have you seen any leaning out at high loads?
Have you noticed any change in starting characteristics, as in time in cranking mode before it starts?

 

I hope you can extract my log from this zip file. It was a little over the 5mb allowed. If not I'll take another, shorter one. I took a little ride around the block. It stalled only twice. About halfway through I turned on the air. You can see the osculations. It's a MS3 from Trackspeed. I don't seem to have any leaning out under high loads but, I haven't logged to check. The car has been more difficult to start, taking a dozen revolutions or more to "kick". Anything you can see in the log that would help me understand what to do would be most appreciated.

 

Under the Startup/Idle menu you should have "Air Conditioning Idle-Up". There are several options that let you adjust how your tune handles idle with the AC on.

Idle-up duty % will increase the initial % your idle control valve opens, mine is set at 5.1 (meaning 5.1% higher than the tune commands when the ac is off).
Idle-up Target RPM adder increases your idle target by a given number, mine is set at 300 (meaning it idles 300rpms higher than the tune commands when the ac is off).

For oscillating rpms at idle, check your timing map and see if there's a big jump in timing in the area your idling in. If timing is moving around a lot it's hard to get it to idle consistently.

 

1) look through your log. When Status 2 shows "128", you are in CLI. Rarely are you there. Change PID delay and Crank to Run Taper(s) both from "2" to "3"
2) your idle VE is much too rich. The few times you do go into CLI, you are running 12:1.
3) Your Initial values are set to MAT, but the Y axis values (I know, the label is confusing) is maximum of 50, when your MAT is 90. Thus, you are not utilizing this table.
4) Your initial values are too low. They should be set at where you tend to idle, with lights on. However, you must be in CLI to test this.
5) Your dashpot should be a bit higher, so you settle a little higher.
6) Your dashpot adder for A/C (Idle up duty %) is too high (or will be when you increase the base dashpot)
7) I don't think your idle timing correction is aggressive enough. You have settings of -400. You cannot let your idle get that far off target. Thus this is wasted space on the curve.
8) I prefer my base idle timing to be 10, rather than 15. Then I can add more timing correction. I have my non-idle advance table in the same MAP / RPM range as idle at 15. That way, as soon as I give throttle, I immediately jump from 10 to 15, presuming no advance correction was being added in my steady state idle.

I would get my base idle working well without A/C first, then add it.
Your main problem is that you are not in CLI most of the time you think you are, so tuning it is futile. Start very conservatively with the PI delays and initial values so that you solidly get into CLI.

I don't recall what my PID settings are, so I cannot comment on those until I close your tune and open mine. Note: yours are 545 slider, P25, I40, D50.

OK, mine are: 2394 Slider, P60, I55, D0.6

 

I made the changes DNMakinson suggested. I didn't drive car but, it seems to idle in cl ok. I couldn't get my idle timing below 20 though. I don't know what I'm doing I guess. Now when you give it a little gas it hangs up at around 2000 rpms and slowly goes to 850. If I turn the ac on the car just stalls, unless I give it gas and it will idle at 1050. It hangs at 2000 rpms as well. I want to be able to figure out how to get this damn car to idle and run correctly. Am I going in the right direction? Should I post another log, with tune?

 

Joe,

I looked back over the log this evening. Actually, most of the times, you did go into CLI.
In the following picture is where you stalled. In that instance you did not go into CLI. Here the CLT is 142, MAT is 90. So, your target is about 950 RPM, and with the funny MAT to Duty cycle, your Initial is 30.6. Plus 2 for Dashpot, gives the 32.5 that your valve is open to. This is not enough to hold your RPM at 1100 (approx) so you can go into CLI. Exacerbating the problem is that your Spark advance in columns 700 and 1140 go small at higher MAP. Thus, as the engine begins to die, map rises, advance falls. I would make you main Spark Advance tables stay at 14 from 19 kPa up to 97 kPa.

Or, something else besides the act of stalling is causing that rise in MAP, like some kind of leak, but I think not.


Regarding long start times: consider some priming pulse (note, PP is an actual mS of squirts not a percentage like Cranking Pulse). Also, I would not use batch on cranking.

 

Tune + log is much better information than descriptions.

Hang then moving is, yes, much better. That means things are controlled, not they only need to be controlled correctly.

What year car? Did you do an actual set-up on the Idle Valve to determine you min and max values?

I still suggest to get the non-A/C idle working well, then add the A/C on top.

Maybe someone else will see something else.

 

Note to self- DNMakinson is very helpful and knowledgeable.

 

225@90%, about 75@30 these were the numbers when I ran the idle valve test. Don't know if I did it right or not. Made the changes in timing and added prime, which helped start time. I tried to drive it and the idle was oscilating a lot. I took a log when I hit my driveway. I then shut it off and restarted. You can see it starts off idling ok then slowly goes crazy. I don't know where to find a control for the MAT based initial value. You said it was 30 something. How would I raise that? Hate to be such a noob.

 

Honored, thanks.

I don't understand your numbers. Put in idle test. Change % and note RPM with change. ... or does 225 mean 2250 RPM? then things make sense. If at 30% PWM you ran 750 RPM, then set Idle Valve Minimum Duty in CLISettings to "30". Ditto for 90%. On same screen you will see where you have selected
Use MAT in Table Lookup. Then go to Close-Loop Idle Initial Values to see the table I mean. instead of Y axis going from something to 50. Make it -20 to 140. NOTE: I was mistaken, you were probably OK with Dashpot around 2, the issue is more this table. Add the extra there.)

Slowly goes crazy... not having seen log or tune, I would guess move the slider to a lower number.

You tell MS what you want to use for Y-Axis of Initial Values table. You have (correctly) selected MAT.
Now go to:



How to set up the table? WHEN the car is correctly idling, with lights on, look at what the RPM, MAT and PWM are. Then go to that table and fill that cell in. Smooth between them. At first you might think, Oh, I will never be at 1400 RPM target with MAT is 100, but that depends upon what you have your A/C RPM adder set to. (wait... please don't add in A/C yet. Let's get basic idle down first.)

As for taking lots of time to get from 2000 to target, change your PID Ramp To Target Time(s) to 1 or 2 seconds.

Year of car? Please add present tune (that goes with log), if you change the tune, then re-post a new log as well.

 

Since I had screwed up and it was running like crap I went back to the tune where I started. I made a bunch of changes:
Reduced idle ve
Reset the initial tables to the cell numbers you suggested.
Changed the idle timing correction.
Changed dashpot.
Changed timing to 14 up to 97 kpa in 700 rpm column.
Added prime.
Car is a 2001 VVT car. Stock, except for turbo kit.
As far as the idle valve test. If I change the percentage it doesn't do anything. The numbers I got were from when I hit the enable test run position. The car wildly oscillates. When the rpms are high the idle stepper position is 225 and the idle pwm is 90. When the rpms drop it reads stepper position 70 and pwm at 30. I still really don't understand what I'm doing with this tool. I know the results will determain different settings. With the current tune the car idles, with the occasional dip and is slow to come down, but it's driveable.

Funny thing I had to use the car. It was a 85 degree humid night, perfect for top down blast. I was at a light with a long straightaway and decided to run through the gears. Halfway through 3rd gear total ignition (I gues) failure. Coasted to the side of the road, lifted the hood, for it to cool off and 15 minutes later it started. Got me a couple hundred yards from home. A few hours later it started again. Sounds like a typical cam sensor failure that is common on these 2001 cars. It's the heat that gets them. You have no idea how much I'm appreciating your technical help. It's very selfless. With some of your knowledge and luck on my part maybe I can get a smooth idle with the ac on. It was 97 degrees at 5:00 yesterday.

 

To run the Idle Valve Test:
1) Warm engine to normal running temp (CLT)
2) Lock timing to 10 degrees. This is in Ignition Options, Fixed Advance set to 10. (Later you will go back to "Use Table").
3) Under Tests / Canbus Tab, go to Idle Test
4) Set value to 40, Enable Testing, [Idle should rev up to 2K or so]. Then move it down. Idle will fall. When idle RPM STOPS FALLING, the %PMW is what you will enter in the Minimum % on the CLI control screen. I suspect it will be close to your 18 you have now. Ignore the gauges. The #'s you want are PWM% you key in, and the Tachometer RPM.
5) Set value to 50. [Idle should rev up to 2500 or so] Then move it up until idle RPM STOPS RISING, or hits 3000, whichever comes first. That is the %PMW you will enter in the Maximum % on the CLI control screen. I suspect it will be lower than the 90 you have now.
6) Move value back to about 40, and Stop the test.
7) Return spark to "Use Table". (from step 2)
8) Key Min and Max % PWM values in the CLI screen.

Now you will be ready to proceed.
NOTE: These values interact with the slider to regulate the sensitivity of the PID loop.

I will look at your posted files this evening.

EDIT: The picture of Inital values is a stock photo. The minimum temperature should be the minimum for where you live. 20 may be OK, but it may need to be zero or lower. Space rest out accordingly.

 

The above is from your log of today.
I'll walk you through it.
Your idle was quite high at first because you told it to be. Goes like this. Curve of CLT to Idle target at 198F calls for 850RPM Then your initial values table at 850 RPM and 104F MAT asks for something close to 44. Add 2 for dashpot, and you get a total of 46. That is just where it was when you lifted the throttle at 110 seconds (previous to the screenshot).

At 114 Seconds, MS properly puts you into CLI, and the PWM immediately begins falling, trying to bring you to your target. However, at the same time, that action is being fought against by your Idle Correction Advance (ICA). Why, because your curve is wonky. ICA should be zero when the RPM error is zero. It should be slightly minus (subtractive) when the error is positive, and it should be adding advance when the RPM error is negative (RPM is below target). The Advance is faster than the PWM, so your RPM actually rises at 114Seconds, rather than falling.

The idle, nevertheless, gets driven towards target but does not get there before you give gas again. At 126 Seconds, PWM is down to 34.5. However, your ICA is still fighting it with 6.8 additional advance (So total advance is 15 idle advance + 6.8 for a total of 21.8 BTDC).

But, let's say that the idle did get down to your target of 850 (though it looks like it is really 900 that it is trying to hit???) I suspect your fan was on, which would add 50 RPM to your target. At the 1020RPM at 126 seconds, your PWM is 34.5. That is the approximate number that should be in the 850 / 104F cell of you initial value table. That's what I meant about seeing what it ends up being, and putting that in your IV table.

I think what you may want to do is turn off ICA until you tune steady state. Then add it back to help with transients (like A/C, lights on, heater fan on, power steering usage).

Maybe also start with all of your IV table at 34.5, and then tweak as you actually hit various places.

Am I making sense?

Later for ICA, the curve should look similar to this ( It may not be perfect, but it works for me, and I'm starting from 10*):

Note it has Zero Timing Correction at Zero RPM Error.

At 173seconds it looks like you turned on your A/C. I'm sure why, but it took you out of CLI. Maybe because your total timing at that point was 39*. I don't know what happens to the engine with that much advance at 800 RPM, but it doesn't seem good. I don't really want to address that yet, even though you want me to, because I want you to first build a good foundation of a solid idle. Then add A/C.

 

I couldn't find where to turn off ica, so I just mimicked your table. I set initial values all to 34.5, except I see I missed one cell. When trying to find initial values, while looking at the log, you just use the pwm at that rpm? I'm showing 28.4 at 1312 rpm and 35.5 at 1050 rpm. Does this seem right? If so I finally get it.

Car runs now with new cam sensor. Idles pretty good too. I really think it could be close to correct. Then I'll need to figure out how to set ac stuff. It's hot here.

 

Mmm, not exactly. It only counts when the idle is settle at target. Just because it passed through 1312 is not the same as settling at 1312.

But first, did you rerun the idle PWM test and set the Max and Min %? That is fundamental first thing. Not absolutely necessary, but the correct first step. It looks like you may have, as your Max % has changed (or did you cheat?)

Then to the table. Easiest point to fill should be CLT at 190, MAT at 100 (or whatever is happening in steady state, fully warmed). You can log what your target RPM is, or look at your CLT / RPM target is + the adder for you fan, if it is on. Idle long enough to hit your target. Then read the %PWM. That goes in the cell you are on. If you open the table while running, TS will show you where you are settled, and what the % is, and therefore where you should put the value. Maybe it is not exactly on a cell? Then put that value in the nearest cells.

The next time you start the car from cold, you will be at, say MAT = 90, CLT = 90. Again, have TS running and the Initial Values table open. The little circle should settle in the 1300 range, because that is what the target is for CLT 90. There will be an enunciator (near far right, top row) that will show when you are in CLI, so it is easy to tell. When your RPM gets to 1300 and 90F MAT, read your PWM (if you don't have that gauge, create it by right click, designer mode, and figure it out from there). So, read your PWM, and key it into the position at 1300 / 90 cell. Rev the engine, the PWM should go right up to the value you set + your dashpot value., hold there for 3 seconds (because you are going to change the PID Delay(s) from "1" to "3" as explained early on), then move to 1300 over 2 seconds (because you are going to change the PID Ramp from "5" to "2).

Rinse and repeat as CLT warms up, and RPM target moves down. By the time CLT gets to 170, and the Idle RPM target gets to 850, you should have the 90* MAT row completed. You have to work on the other MAT rows when you have other MAT's to work with.

NOW you would be ready to work on A/C BECAUSE, you have a row at 90 * MAT that includes the1300RPM cell, and the 1200 RPM Cell, etc, which, by the way is what MS will look at when you have the A/C turned on. See you have 300 RPM increase in your A/C screen. So, 850 Idle, + 300A/C + 50 for your Fan Control (note that you tell the fan to come on when A/C comes on, and you tell idle to increase by 50). So, as soon as you have A/C engaged, when at CLT > 170 is normally saying idle at 850, MS is wanting to idle at 1200. So it is important that you already had that cell properly populated so you will be able to see the affects of the +3.9% PWM the fan added, and the 7.8% that the A/C added. If you don't do this in an organized manner, then you will chase your tail. When that idle tries to start at 2000 RPM, you won't know if it is the Initial Values table being off, or the Fan % increase being off, or the A/C increase being off.

I'm still suggesting that you set your Idle Advance Settings all to "10". That will make sure that your advance corrections have impact. See, you start with the idle spark retarded, and therefore not at best torque. Then, when a load is added, the spark moves toward best torque and can catch the idle. (Note: the Idle Advance Settings screen [at the top] is where you turn off and on Idle RPM Correction).

Looking at your log. At the beginning, you are idling fine. Then you give it some gas, and it tries to die. Look where you advance went, to 3.8 because you're now out of CLI and onto the normal table, which at 94kPa, is calling for that low of advance. I again suggest that you change your 700 and 1140 column Spark Advance to 14 all the way up to 97kPa, and then put about 8 in the 116 row.

Next, after the throttle stabs, MS never put you back to CLI. BEFORE YOU GO FURTHER: Change PID Delay(s) from "1" to "3"; Crank to Run Taper from "2" to "3"; PID Ramp from "5" to "2. These are fixes for inconsistent going into CLI.

Then you drove or something, and at 86 seconds in your throttle lift had cut all fuel until you hit 1900 RPM. Then you went in CLI and idled well for a while. That went fairly well. PWM % stabilized at about 32. This would be your Initial value. But for that to work right now, you will need to up your dashpot to about 3.5. If you would have let the car idle longer, EGO would have pushed you lean, and that would have affected your idle quality. Thus:
BEFORE YOU GO FURTHER: Change AFR target in your idle range to 14.7. You are too lean there. In fact, you want to make all cells from columns 500-1500, from rows 30 to 80 kPa: 14.7.

Then, what happened at the end of your log? At 170 seconds in, your target RPM suddenly jumped from 850 to 1700, then 10 seconds later to 1470? Then finally, you ramped back to a nice idle.

COMPLETELY SIDE NOTE: Since you have not yet really tuned your VE table, I suggest you step back and go to Basic/Load Settings / General Settings, and change Incorporate AFR to "Include AFRTarget".

Looks like you have made some progress, but it also looks like you have lots of contributing issues, such as bad Spark, Fuel, and AFR maps.

Do the BEFORE YOU GO FURTHER things, then go further.

EDIT: Set your "D" in the PID of CL Idle control to "0". Because, even when you are in CLI, It is taking way too long for the PWM to move to the correct position. Look at the end of your log. after the target goes correctly to 850 RPM, after 9 seconds, your idle is still above target RPM. This is kind of where we started. I know you said you had lots of oscillation with more slider, less D, etc, but I don't ever remember seeing that in a log.

 

I did the idle duty test. I had an issue when I tried to lower values until it stopped dropping. It would drop until it hit 400 rpm or so then stall. I can't remember what that number was. It idled normally at 18, so I left it. The higher value was when it hit 3000 rpm.

I'm pretty sure I understand how to set initial values. I loaded the pwm and mat gauges on tuner studio. The cold start I don't think I got the numbers right. They seem too low. I had problems at 110* because the temps wanted to keep dropping. I shut it off and let it heat soak to raise them some. The pwm gauge seemed to stick on me and I had to let it sit for them to drop and then go to my desired rpm. I recorded what I thought were the most normal looking numbers. After an hour of idling mat never got over 112, I think, so I only changed the 80* & 110* cells.

Changed my idle advance and went to incorporate afr target. I still must have something really wrong because I tried turning on the ac and idle droops then oscillates between 600 and 1400 or so. I don't have ac on log because I know I have to get idle correct first. I'm starting to feel a little foolish not being able to make this happen.

 

Joe, I missed something. In your Idle VE screen, I suggest you change "PID Idle Activates" to "Use PID Idle Activation". That way, the settings you have in the C--L Idle Settings control when you go into CLI, Idle VE, and Idle Advance.

Good call on the 18 Min. You don't want the car to ever be driven into too low RPM.

I would tweak the Initial Values 110* row to be a gentle increase from low to high RPM. Something like 33, 34, 35, 36, 37. You should be getting into the ballpark. I don't think you can cheat MAT, just wait until it happens naturally. Unfortunately, some settings have to be either anticipated, or re-looked at in the right season (weather).

That log looks good. Every time you correctly go into CLI, and then move fairly quickly to target, and EGO moves you to correct AFR without a lot of correction. Does it seem to be working OK to you?

Actually, I'd like to see a log of A/C. First, I'd increase idle -up delay to 350mS. Then get a log with it so we can see what is happening.

One word of warning, with move to Use AFRTarget, you will be quite rich in boost.You will want to run some VEAL and tame your VE table. Note: when running VEAL, make slow, smooth throttle changes.

 

Well I made the changes and it idles real good, even with ac. The log is in the driveway, idling, with the ac on. I guess the spikes in the log are from the ac cycling. I drove it around a bit, in VEAL and it runs pretty good. Still real rich under boost. It goes down to 10.4 on my gauge. I'm going to have to wait so I can get to a road to make a full pull and log it.

Really happy. If you ever get to Sebring, for the races, I owe you a dinner. I'll post a log of my pull when I can.