High idle on cold start: vacuum leak or IAC setup?

[jangell]

Time for a new 3.0L/MegaSquirt/EDIS thread. :) First, I’m running the coil pack now instead of coil-on-plug. Redoing the wiring for the CoP didn’t fix anything, so I’m guessing a coil is bad or something. Since the coil pack works, I’m tuning with that for now. I’m going to go to a junkyard later and try to source a more easily mountable pack and some longer wires. I might switch back to CoP after I’ve run the coil pack for a while.

Anyway, the car now starts, but it shoots up to 2000 RPM and stays there. This is after I turned the throttle adjustment screw all the way to closed, too. However, when I used IAC Test Mode in TunerStudio to completely close the IAC, the car starts and idles at around 800 RPM when cold. It starts the lose RPM as it warms up, so I adjusted the throttle screw a bit until it stayed running, and adjusted the fuel table to keep it in the 14.7 AFR range for good measure. It seems to idle fairly well once warmed up.

My question is about this high idle on cold start with the IAC. It’s about 45 degrees ambient today, in case that’s relevant.

I’m thinking I just have the IAC set up wrong. But I want to make sure there isn’t a vacuum leak before I delve into that. Since closing the IAC seems to let the car idle properly, that implies there’s no leak, right? Otherwise the car would still be idling way too high, wouldn’t it? But it seems odd that I have had to open the throttle a little to keep it running as it got warmer, too — I thought I’d need to close it more, if anything.

I also have an analog vacuum gauge hooked up. With the engine up to temp it shows >30 inches Hg at idle (it maxes at 30), well outside the marked green zone of 17-22, but I really have no idea what that means. Some Google searches suggest that it should be idling in that green zone, so that implies a leak.

But TunerStudio shows a MAP of 40 kPa, which I think is OK?

I previously set the ignition timing as best I could, with a 1 mm gap between the VR sensor and missing tooth wheel, centered as accurately as possible six teeth before TDC. So I think the timing is right, anyway. At least the exhausts aren’t glowing.

I’ll attach a log and tube when I get home. It’ll be of a warm engine; I didn’t take one of the engine cold.

It’s looking like I’ll be able to drive this thing soon. :)

Thanks!

— Joe

[jangell]

Log and tune attached.

In a video I took a couple weeks ago of the engine idling at 2000 RPM, I can see the vacuum gauge showing in the 18-22 in Hg range for vacuum. With the IAC closed and throttle almost completely closed I'm maxing out the gauge at 30 in Hg at 800 RPM. I had assumed a vacuum leak (or at least a misconfigured IAC) since closing the IAC got the idle down, but 30 in Hg seems like a lot of vacuum at idle. I'll grab a video of the gauge while the engine is running and revving later today.

I adjusted the first four columns of the fuel table together to keep the AFR around 14.7, so there's a jump between the fourth and fifth column. Nothing is really tuned yet.

58873
58874

Thanks again

-- Joe

[FABombjoy]

Your combo of spark advance and idle steps is causing the runaway idle.

There are several means to solve this. I would first set up closed-loop idle control. "Warmup mode" is less sophisticated than the original idle control system. Then you can also involve spark control in the idle system for ultra-stable idle.

Setting up idle advance will help stabilize warm idle.

Add an idle zone to your spark table - all the same value so little shifts in rpm don't cause shifts in spark timing.

[jangell]

You have confirmed that it was my ignorance. :)

I took your advice and set up closed loop idle control, which was much simpler than I thought t would be. I closed the throttle all the way and am now relying on the closed loop to keep the idle at 790 RPM (+/-30 RPM). It’s working quite we so far.

For the advance table, I just set it so that all 9 cells (3x3 grid) around the cell it was idling at (14 degrees at 40 kPa at 790 RPM), but I can’t test that today because the starter doesn’t want to crank anymore. (Applying 12v to the solenoid wire doesn’t seem to do anything, and the battery isn’t discharged, so something’s up). I’ll have to check further once the engine cools down and I have more time, probably on Friday.

Also, I’m thinking my vacuum gauge is junk. I finally thought to look at it without the car on (or even plugged into the vacuum system) and it shows 12 in Hg vacuum instead of atmospheric pressure. So that was just a big red herring.

Thanks again!

— Joe

[DMC5180]

Sounds like you fried the contacts in the starter solenoid. Does the solenoid pull, click with 12v applied? If it does but the motor doesn’t spin that’s likely the problem.


Sent from my iPhone using Tapatalk

[jangell]

It had been intermittent the last few days of testing, but it seems to have finally just stopped working. I have a push button hooked up in the engine bay so I don't have to go into the cabin to crank the engine, but it's basically the same wires as used when turning the key. Fiddling with the wires seemed to help, but it was probably just coincidence. I thought it was the button, but replacing the button didn't help.

I finally tested the wiring with a PowerProbe; I'm seeing ground on the wire going to the solenoid, and applying 12v doesn't cause anything to happen to the starter.

Similarly, I get nothing when I push the button -- no sounds at all from the starter motor or the solenoid. It's as though I hadn't pushed the button.

Luckily, compatible starters are readily available. I'm off work on Thursday and Friday, so I'll go down on one of those days and see if I can figure out if it's a wiring problem or the solenoid itself is broken. At worst I'll pick up another one at NAPA or Autozone.

Thanks

-- Joe

[Josh]

unfortunate to hear that you went away from the CoP. I was fairly proud of that. 99% likely it was a bad coil.

I may have a bracket for the coil pack that i ran before. Mounted to the passenger side valve cover. Its in my junk pile somewhere.

[jangell]

I got three more CoP coils while I was at the junkyard, but at the moment I just want to get it running. Don't worry -- I plan on trying CoP again in the future. I like the idea of CoP more than coil packs. I do wish EDIS would fail in a more obvious way when a coil is bad.

I already have the coil pack mounting figured out, so no problems there. Thanks!

-- Joe

[opethmike]

You should hug the car.

[jangell]

Apparently hugging worked, and I was able to actually drive the car today. I mean, I just drove it around the block, but still, it drove.

I never figured out what was wrong with the starter. After removing the header and pulling the starter, I just tightened the wires and it ran without a problem. I think one of those wires was just loose.

The brakes aren't doing so good -- the car stops eventually, but I wouldn't want to trust them for real driving. My mechanic friend is going to help me bleed them this weekend, but I plan on putting delorean.co.uk's performance brakes on sooner than later (I've been wanting to get those for years).

I also noticed that the car wants to stall if I quickly take my foot off the accelerator. I tried adding more fuel to the leftmost column of the VE table, but it didn't help as much as I thought it might. I'm not sure if there's something specific to handle this case. Decel Fuel is at 100 still, so I'm not quite sure what's up there, but I'll keep tinkering.

Beyond that, I have an exhaust leak to fix (clamp is loose), and the TPS keeps seizing on me (I'll try adding a gasket to space it a bit), but otherwise things are looking good.

Thanks again to everyone for their assistance on this conversion over these past nearly five and a half years!

-- Joe

[FABombjoy]

I also noticed that the car wants to stall if I quickly take my foot off the accelerator. I tried adding more fuel to the leftmost column of the VE table, but it didn't help as much as I thought it might.

58921

[jangell]

:) I took a log, but forgot to bring home my laptop yesterday. I’ll be sure to bring it back today and get the log off it.

— Joe

[opethmike]

Did you know that if you make out with the car, that it will LS swap it self?

[jangell]

Here's the log and tune. This includes a couple of stalls and revving at idle before driving around the block on auto-tune. I was tinkering a bit with the leftmost VE table cells while idling, and auto-tune messed with the rest of the table while driving, and thus the tune is from after all that tinkering/auto-tuning.

58922
58923

There's a lip at the edge of my garage. I have to give it more power to get over the lip, then release the gas as soon as I'm actually over it so I don't rush into traffic. That's how I was able to trigger the deceleration stalls so easily.

My uninformed guess about what to do next would be to slightly increase the timing on the leftmost cells? Since I've already upped the fuel in the leftmost VE cells, that seems like the next thing to try.

Thanks!

-- Joe

[FABombjoy]

Based on my experiences, building a wall against stalling should be done using spark advance. The VE table will probably only have a minimal effect and will probably cause warmup fuel to be excessive.

A few immediate thoughts:
-Where is TPS? I know that TPS is critical in MS3 for accuracy of idle control, but it doesn't look to be quite as used in MS2. I didn't see a concept of "engine states" in your tune, so TPS may only be useful for accel enrich and such.
-Revert the VE and bring idle AFRs back into the 14.x range
-Open up your base idle screws a bit. When you have a stable warm idle, MS opens the IAC valve to 6-7 steps. See if you can get that to almost 0.

Your spark table has 5 duplicated columns: 3100 through 6000 are almost the same. There are some shenanigans going on in 3100 but we'll pretend that's a keyboarding error. This means you could rescale the table and add some low-RPM resolution.

Your low RPM breakpoints are 500, 950, and 1200 but the target RPM is 790. This means you're always going be at an interpolated spark advance value, making it challenging to fine tune those columns. The high advance values in 1200 are causing frequent, consistent oscillation when you transition from on- to off-throttle.

I would shift the advance table around to free up some low-end advance columns. Then set your 500 RPM column to +1 or +2 degrees as an anti-stall measure. Put your VE back to where it was so that you don't have a big rich idle and you should have much better off-throttle driving response.

[jangell]

TPS: I do have a TPS, which is located on the Mustang throttle body opposite the throttle cables (the body is designed for one). I’m having a problem where whenever I snug the TPS bolts down, the it no longer turns with the throttle. I’ve gone through three of these over the years of testing as this binding evenrually ruins them. I’m worried that if I make it too loose that I’ll have a vacuum leak. I’m going to try making a thicker gasket to space it a bit more and see if that fixes the binding.

From what I can tell MS 2 mostly uses the TPS for more responsive acceleration enrichment, so it’s optional in my setup, but it would be nice if it worked. I don’t recall seeing anything about “engine states” mentioned anywhere before, but I know MS 3 has some more advanced stuff, and I mostly just didn’t read anything MS 3 specific so I wouldn’t be tempted to upgrade. :)

VE: will do. I had a feeling that the leftmost VE column wasn’t doing much to prevent decel stalling.

Base Idle Screw: ok. I had been planning to let the IAC completely control idle, and just close the throttle all the way, but it seems that I need to crack the throttle to get the idle up above 650. My theory was that this was the first tim is out the air filter on, and now it can’t suck enough air through just the IAC, but That doesn’t seem right. Without the air filter, it seemed to be able to idle around 790 just fine with the throttle completely closed, and with the IAC open all the way would rev to 2000, so I didn’t think I was losing any useful headroom by letting the IAC manage with a closed throttle. Of course, I haven’t actually tuned the warmup yet (I was going to let warm up auto-tune handle that, which is more likely the reason I needed to crack the throttle when the air filter was on (I literally just realized this; it makes a lot more sense than my other theory).

Anyway, I’m fine with the IAC being zero at idle to. I was watching the RPMs more than the IAC steps, so I’ll monitor that on next start.

Spark table: that’s almost completely auto-generated. I have no idea how to adjust it at all. The only cells I changed are a 2x3 block around my idle, which I set to 14.

I set the low columns to 500 and 900 but put 14 in adjacent idle cells in the theory that it would just interpolate to 14 at 790 RPM. As opposed to setting the second column to 790, as that would cause interpolation in a 3x3 grid. Or at least that’s how I understood it.

I’m good with adding more low RPM columns. I’m just not really sure how important the high ROM columns are or what values I need in those columns when I reassign. Since you noted that I have duplicate columns, can I safely just reassign the the lower columns fornidle ranges, and just interpolate the higher columns?

Thanks!

— Joe

[FABombjoy]

Base Idle Screw: ok. I had been planning to let the IAC completely control idle, and just close the throttle all the way, but it seems that I need to crack the throttle to get the idle up above 650.
You can also set the IAC minimum steps, but then you might cheat yourself out of the max ability of the IAC to control RPM under all conditions.


Spark table: that’s almost completely auto-generated. I have no idea how to adjust it at all. The only cells I changed are a 2x3 block around my idle, which I set to 14.
https://www.youtube.com/watch?v=CiqrSJIITGY


I’m good with adding more low RPM columns. I’m just not really sure how important the high ROM columns are or what values I need in those columns when I reassign. Since you noted that I have duplicate columns, can I safely just reassign the the lower columns fornidle ranges, and just interpolate the higher columns?
Exactly! The above video shows how to rescale with interpolate, which will help you shift everything over and make room for some fine-tuning in the idle ranges.

[jangell]

Thanks for the video -- I had seen that button in the some post months ago, and completely forgotten about it. It's one of those "complete visible but well hidden" bits of UI. I had no problem using that to add more columns.

I set up the new ignition table and VE table, adding leftmost columns for better idle control. It seems to idle fine (+/- 30 RPM), but I'm still having problems with stalling when I quickly take my foot off the throttle (ie: foot down and up fast, causing it to rev then go back to idle, but it drops down low and stalls). I boosted the low column of the ignition table to 17+ degrees, and tried as high as 25 degrees at one point, but it didn't seem to help. Not sure what I'm missing here. Not sure how much it matters, but the throttle screw is only slightly cracked. Idle motor steps are close to 0 (seems to be around 6 sometimes, but usually 0).

The TPS is still no good. It seems to work sometimes, and not at other times. I finally looked at it more closely, and the part number seems to be a Dodge/Chrysler unit that is MOSTLY the same as the Ford TPS that the Mustang throttle body expects, but not exactly. For example, the "ears" that grab the throttle bar thing are smaller in the Dodge, and the diameter of the barrel is a little different, which may be what's causing all the binding. Of course, a Mustang TPS has an upward-pointing connector, which interferes with the plumbing for the air duct, and there isn't enough clearance to point it downwards instead. Still, I might hit a junkyard tomorrow morning and source one just to try it out, and see if I can come up with a way to route around the air intake.

Other than that, I flushed and filled the brake system, so if I can get this stalling stuff figured out I should be able to take it for another drive. It starts pretty readily, so if I did stall at a stop sign or something I could restart it easily enough, but since I'll need to get the stalling resolved at some point I might as well do that now.

A couple of logs and the tune are attached. The longest log includes some auto-tuning, which I reverted while I tried to get the revving figured out. The shorter log contains contains a clean VE table and the new ignition table.

Thanks!

-- Joe

58950
58951
58952

[Bitsyncmaster]

I set that curb idle screw by turning it in on a hot idling engine until the closed loop idle can not comenate any more (idle goes higher RPM). Then I back off that screw about 1/8 or 1/4 turn. That should help your stalling problem when you take your foot off the gas quickly.

That's with the stock K-jet but it may help you the same.

[jangell]

Thanks -- I'll check that tomorrow. I'm trying to get it so that the IAC steps are close to 0 when warmed up, which would mean that the IAC is completely closed and that the lowest RPM at idle is held by the throttle screw alone. But it seemed like it only needed a very slight turn to do that -- I think a half turn, but I don't recall exactly.

I did notice that once the engine was in gear, it dropped to the 650-700 range -- I only just remembered that because I spent most of my time tinkering with the car in Park. Maybe I should target closer to 850-900 for an idle, and keep the idle screw a bit more open, then see if it drops back to 790 or so when in gear. I'll have to check again tomorrow and see what it's doing, as it's likely I'm remembering this all wrong right now.

-- Joe

[opethmike]

That sounds about right. Putting an auto in park will drop the idle usually around 200 RPM or so. That could also be why you are stalling on lift; throttle plate closing too far. Not saying it is for sure, but it ain't helping.

[jangell]

I do have MS set to closed loop. You'd think that it would open the IAC to get the idle up to 790. I probably have something set up wrong if I was seeing ~650 when in gear. I also don't know how fast the IAC can react -- it might be too slow to account for the rapid change in MAP from tapping and releasing the throttle, perhaps?

I'll open the screw a bit, raise the idle, and see if it still stalls out on me. And go over the TunerStudio settings again. Thanks!

-- Joe

[FABombjoy]

The spastic TPS is a problem - per your tune settings, closed-loop idle won't activate until TPS is less than 1%. RPM drops but the stepper valve never opens. You may just need to further open the base idle to compensate a little until you have TPS working.

Or set the threshold for idle activation higher than your stuck TPS value and tweak the RPMdot and load values so it activates only when your likely to be idling.

[jangell]

I keep forgetting that TPS is used in places like that. That would explain a lot.

I went to the junkyard and got a TPS off of an actual Mustang, and of course it doesn’t fit (slightly different mating size for the barrel, slightly different screw hole spacing). I should have trusted that’s Josh had the right TPS in the first place. I put the old TPS back in with a much thicker O-ring, and it seems to be working OK for the moment. We’ll see if that lasts.

I was still getting the idle dropping to the 650 range until I changed Use Last Value it Table to Table and configured a basic table. Now it soft lands around 790 RPM like I want it to.

I think I have the decel and idle sorted well enough for now. Next is to figure out why it wants to stall when I open up the throttle even a little quickly. Probably time to play with accel enrichment and the VE table, or something.

Thanks!

— Joe

[opethmike]

I keep forgetting that TPS is used in places like that. That would explain a lot.

I went to the junkyard and got a TPS off of an actual Mustang, and of course it doesn’t fit (slightly different mating size for the barrel, slightly different screw hole spacing). I should have trusted that’s Josh had the right TPS in the first place. I put the old TPS back in with a much thicker O-ring, and it seems to be working OK for the moment. We’ll see if that lasts.

I was still getting the idle dropping to the 650 range until I changed Use Last Value it Table to Table and configured a basic table. Now it soft lands around 790 RPM like I want it to.

I think I have the decel and idle sorted well enough for now. Next is to figure out why it wants to stall when I open up the throttle even a little quickly. Probably time to play with accel enrichment and the VE table, or something.

Thanks!

— Joe


Definitely AE, not VE. For this particular issue, leave the VE alone.

[jangell]

Will do — thanks. I’m using a clean, untuned VE, so I need to get to auto-tuning as well.

I drove around the block. Got a few stalls on acceleration, as expected. The idle also dropped to >650, because apparently the TPS got stuck again (this time at 5.5%). <sigh> Guess I’ll tinker with that for a bit, but I’m getting closer to just unplugging it and switching everything to MAP.

— Joe

[jangell]

I’m giving up on the TPS and have disabled it (unplugged the connector and jumped ground and sense), then set up MAP AE, and it revs fine now without stalling out. It’s getting dark soon, so I’ll do another drive tomorrow. Also, I need to tighten the belt, and it’ll be nicer to do that without risking burns from the exhaust.

Thanks again!

— Joe