95 LT-1 Idle Cell Comparison - Humidity?

Been struggling with this problem for roughly 2 months. After a bunch of experimentation I started looking for compression loss because it was struggling to start and stay running, and most of the time died when it switched to closed loop.

While I'm not leaning toward the ignition controller being the cause, as most of you know I'm running my experimental coil per cylinder controller (link in sig). It seems to run great once up to temperature, but it is certainly possible the ignition controller is a contributing factor.

At this point I can barely get the thing to start - it seems like it's not getting enough air. I've drilled the IAC bypass hole from 11/32 to 1/4 inch with little change. Cold starts barely stay running with the IAC pegged at 160 until the engine warms up. Once warm IAC counts are around 80 at 950 rpm.

So I'm looking for a simple explanation because the problem only appeared as the humidity started to climb right after the 43kpa frame was captured. Note the huge difference in metered airflow. I know the barometric pressure is higher in the 43kpa frame, but so is air temp. Aside from this the biggest difference between the conditions the data was captured in is that the 43kpa frame was from late spring with relatively low humidity (and 11/32 IAC bypass hole). On the other hand the 48kpa frame was captured a week ago in 65-70% relative humidity and 17/32 bypass hole. Can humidity and air density account for this much difference? All the MAF logs I have to compare to are 2+ years old and I couldn't possibly guess what the relative humidity was.

Attachment 14445 Attachment 14446

Edit: I failed to mention that this is the first summer I've run this after adding the stroker crank (384ci - 4.040" x 3.750")

I vote for compression loss due to valves not fully closed.
I guess adding more fuel didn`t cure the problem. Iac counts seems normal.
On the second log engine is not making good vaccuum, possibly the result of lost compression.

Humid air is heavier and harder to move. It takes more effort to move the same amount of air. It doesn`t seem to have that much impact as you expect. If you suspect engine chocking, enlarging the 8 ports on the intake manifold is better than drilling the bypass hole.


I hope your individual cylinder trims are not off chart. They do contribute to engine surging and stalling.

I ruled out compression loss. Engine was torn down to replace head gaskets and installed solid lifters.

More fuel isn't generally the direction I felt it needed to go when it smells so obscenely rich as it is. This is why I was suspecting loss of compression and / or very weak spark. But I just can't imagine the spark is weak enough to cause incomplete combustion.

I've been dabbling with the startup fueling and the injector flow constant since the initial post and have got it starting and idling relatively well from warm (approx 60c coolant temp). I really need to buy some known good / unmodified injectors because I've added an entire lb/hr to the setting I was running in the 43kpa frame and it seems to be getting better. I don't know if it's a meaningful measure of anything, but now when it switches to closed loop you can't even feel a difference in idle. A few days ago it would either die or stumble and then switch back to OL for another 90 seconds depending on how warm it was before starting. I guess I'll have to wait until it cools completely to test if I've made any progress in this regard.

The water vapor in humid air also displaces oxygen and contributes to cooling in the cylinder, which is why I'm trying to get a grip on how much this should be impacting things. Bear in mind I'm speaking in terms of "oppressive humidity". It was 82F/28C in the garage this morning when I began working on it, and my shirt was soaked in sweat in 15 minutes.

Individual trims were set using the wideband well before the problem began to present itself. It didn't take much - cyl 6 is at 0.98 and cyl 5 is at 0.99, all others are at 1.0. I haven't re-checked lately but don't suspect this is a major problem.

Rechecking valve lash with solid lifters might help with cold hard starting conditions. It takes a little for a valve to hang down.
If it is a start up fueling problem, adjust the cranking VE table. It is reponsible for fueling at startup.

At cold only hesitation, fueling will be the main suspect. The hotter the engine, the less fuel and air it needs to keep the idle.
Bump the fuel all over to see if it cures it, and start looking for good known injectors. These knock off, no brand reworked always make more troubles than help.

I haven't checked hot lash yet, but I doubt I'm losing compression here. Primarily because it acted the same way with hyrdraulic lifters, but also because lash has probably not shrunk during break-in, rather increased.

Startup fueling is getting my complete attention now, as is injector flow constant. I'll report back tomorrow once it cools completely.

Trust me - I've read "the gospel" on modified injectors and couldn't agree more, but I ran these injectors enough when the engine was a 5.7L to know they aren't the problem. If anything increasing displacement should have made them less problematic for startup and idle scenarios.

Thanks for your thoughts nonetheless.

just for fun lets force your IAC into operating range. crack the throttle plates open until cold start yeilds iac counts <160 and see if things improve. the air will be less even than a properly sized idle passage but that shouldn't affect startup too much.

Quote:

---

More fuel isn't generally the direction I felt it needed to go when it smells so obscenely rich as it is.

---

it takes a lot for it to 'smell' rich. often a 'fuel' smell is a misfire due... lean will hit a misfire state more easily than rich

I'll try that if things aren't improved today. I actually tapped my throttle body to take a 1/4 pipe plug that I'm drilling to make different sized bypass orifices. 1/4" is the largest orifice I can create with the hex socket drive plugs I currently have. Yesterday evening I had a fairly successful start from 38C. IAC was pegged at 160 for the first few minutes but it was at least able to make the target rpm. I was really hoping to hear from someone who's done a 383 LT-1 in a climate that sees swamp-a$$ type humidity like St Louis. Houston, TX and the Carolinas are the only other areas I've been to with comparable humidty.

I didn't want to hear the word "misfire". That points me back in the direction of the ignition controller.

I would completely drop the humidity idea if I didn't also have a problem with my daily driver - a 1.9L turbo diesel at this same time. The vane actuator pot was worn causing the ecu to lose control at higher rpm and go into limp mode where it would command minimum boost all the time. The difference between a 90hp engine trying to climb a moderate incline with little or no boost in May versus July humidity was immense. In 95F air I could very plainly feel the "boost" from drafting semis.

so the only difference between the two snapshots above was humidity? no other engine modifications?

Quote:

---

Originally Posted by steveo

no other engine modifications?

---

There were - solid lifters and larger IAC bypass holes. But this problem started as soon as it started to get hot & humid, well before any modifications. Nothing really changed with the addition of the lifters.

Cracking the throttle a bit made a huge difference. The MAF was actually registering slightly less airflow at the same engine speed. I'm working on checking hot valve lash so can't post too many details. Will try a cold start in the morning and post an update.

You're at the point where you need to enlarge the idle passages in the intake manifold.

Here's a short log of a drive this afternoon. In general it's not running very impressively, but as I've mentioned the transition to closed loop was unnoticeable if that means anything. I noticed some popping in the exhaust when decelerating from a few PE "events".

I suspect that the humidity is causing the MAF to register inflated airflow #s. I say this because I've always logged less spark knock in humid conditions, and I'm presuming that's caused by the water vapor increasing the cooling effect on hot / heated parts. I haven't tried swapping to SD mode yet but will if I can get it to start and stay running reliably.

Wideband is a LC-2 on D27 using default settings. I'm not sure how meaningful any of the data is except to use for cylinder balance testing. I pay it almost no attention unless I drive the car a few minutes to get the exhaust hot. Even then, Missouri's laws on ethanol are stupid - it's considered required in some concentration but they don't have to tell you how much or that it's even present. I'm guessing the 93 octane I'm running has a stoich about 14.5:1.

By the way, with the help of a definition file I found in kur4o's build folder I found I could display the wideband units in Lambda with this in the definition.csv:

Code:

---

54,WIDEBAND_D27,Wideband O2 @ Pin D27,^,EXTENDED,8,2,0.007874016,,0x01,0xFF,,,,,

---

Quote:

---

Originally Posted by spfautsch

Here's a short log of a drive this afternoon. In general it's not running very impressively, but as I've mentioned the transition to closed loop was unnoticeable if that means anything. I noticed some popping in the exhaust when decelerating from a few PE "events".

I suspect that the humidity is causing the MAF to register inflated airflow #s. I say this because I've always logged less spark knock in humid conditions, and I'm presuming that's caused by the water vapor increasing the cooling effect on hot / heated parts. I haven't tried swapping to SD mode yet but will if I can get it to start and stay running reliably.

Wideband is a LC-2 on D27 using default settings. I'm not sure how meaningful any of the data is except to use for cylinder balance testing. I pay it almost no attention unless I drive the car a few minutes to get the exhaust hot. Even then, Missouri's laws on ethanol are stupid - it's considered required in some concentration but they don't have to tell you how much or that it's even present. I'm guessing the 93 octane I'm running has a stoich about 14.5:1.

By the way, with the help of a definition file I found in kur4o's build folder I found I could display the wideband units in Lambda with this in the definition.csv:

Code:

---

54,WIDEBAND_D27,Wideband O2 @ Pin D27,^,EXTENDED,8,2,0.007874016,,0x01,0xFF,,,,,

---

---

Your phone have a baro sensor built in or have you checked local baro reading from weather bug or another trusted source and see if your within 3 kpa of reported. If it was humidity related, a smaller plug gap would yield a driveability improvement. Wide band would report a richer mix also if the MAF was over reporting due to increased air mass from the humidity. startup ve is usually about 60% of normal VE.

Quote:

---

Originally Posted by spfautsch

Here's a short log of a drive this afternoon. In general it's not running very impressively, but as I've mentioned the transition to closed loop was unnoticeable if that means anything. I noticed some popping in the exhaust when decelerating from a few PE "events".

I suspect that the humidity is causing the MAF to register inflated airflow #s. I say this because I've always logged less spark knock in humid conditions, and I'm presuming that's caused by the water vapor increasing the cooling effect on hot / heated parts. I haven't tried swapping to SD mode yet but will if I can get it to start and stay running reliably.

Wideband is a LC-2 on D27 using default settings. I'm not sure how meaningful any of the data is except to use for cylinder balance testing. I pay it almost no attention unless I drive the car a few minutes to get the exhaust hot. Even then, Missouri's laws on ethanol are stupid - it's considered required in some concentration but they don't have to tell you how much or that it's even present. I'm guessing the 93 octane I'm running has a stoich about 14.5:1.

By the way, with the help of a definition file I found in kur4o's build folder I found I could display the wideband units in Lambda with this in the definition.csv:

Code:

---

54,WIDEBAND_D27,Wideband O2 @ Pin D27,^,EXTENDED,8,2,0.007874016,,0x01,0xFF,,,,,

---

---

Your phone have a baro sensor built in or have you checked local baro reading from weather bug or another trusted source and see if your within 3 kpa of reported. If it was humidity related, a smaller plug gap would yield a driveability improvement. Wide band would report a richer mix also if the MAF was over reporting due to increased air mass from the humidity. startup ve is usually about 60% of normal VE.

That 02s and BLM reading looks weird enough. Did you mess with the closed loop settings. Does the AIR system present and functioning. 02s looks frozen at stratup and really slow moving after that. They should be pig rich at startup and I noticed blms were 120 at open loop meaning they are pulling fuel.

You can set mode4 controls before starting the engine. Like setting target afr and spark for startup.

Humidity revealed some hidden problem you already had. Is it a tune or mechanical.
Humid air takes alot more energy to ignite, You lose about 10-20% power and there is water vapor condensation inside the cylinder untill it gets warmed up, which can foul the spark plugs with water.
But that is likely to happen at much lower temperatures.

The engine gets less efficient and draws more volume for the same amount of oxygen. That can explain the maxed out IAC counts at stratup.

To elimimnate CL tune problem I will start by running Open loop and than CL SD and OL SD.

It could be also 2 different problems mixed up.
1 is hard startup, which can be the result of spark syncronisation, Crank ve table, or slower fuel pressure build up.
2 is crap runninng untill warmed up, the result of weak spark, CL setting off chart, MAf table off chart.

Quote:

---

Originally Posted by Rocko350

You're at the point where you need to enlarge the idle passages in the intake manifold.

---

Anyone have a second on this opinion? I'm thinking the same but damned if I didn't just put the manifold back on 3 weeks ago.

What would baro tell me about? I thought $EE only uses SD if the MAF stops working.