Anyone notice any odd fueling behavior during cranking?

md95 · Post by **md95** » Tue Dec 18, 2007 11:16 am

I'm not quite sure, but I may have stumbled upon a very strange issue during cranking, while playing around with my IAC stepper...

The issue: On full warm restarts(sometimes cold starts as well), i have a problem where the IAC stepper goes right to the coolant temp position, during cranking, instead of going to my cranking steps. At the same time of this issue, it seems like hot cranking gets waaaaay to much fuel, per what the cranking percentage is. I had to lower cranking fuel % to 3, and even still have a bunch of black fuel smoke out the exhaust, and has a hard time starting with all the fuel... Now whats interesting, is that when the cranking "works", my IAC goes to my specified cranking steps, and the engine is LACKING fuel, like it's actually going off of my cranking fuel table %(it's too lean because of the 3% cranking perameter, and almost wide open stepper of 16 steps) Whats even more strange, is that Megatune is showing cranking during the issue, but it seems like fueling is going off of something else during cranking, like the VE table..or whatever..not really sure exactly. I believe these 2 problems are related....

I just dont know what think about this, as Megatune is displaying cranking, yet the IAC and cranking fueling dont seem to be referencing the cranking peramaters. When cranking goes properly, IAC steps go to crank position, and fueling references the % table, and i get barely any fuel..motor just cranks and cranks..have to key off then on again, then it fires. It's NOT too much fuel, as pressing the throttle open for more air, doest help...When the issue is present, pressing the throttle for more air to offset all the fuel, it would fire up more easily.

This is almost impossible to replicate on the bench, as displayed values show everything works like it should(except for the IAC steps)...You would almost have to scope the injector outputs to measure the pulse time during the issue, then during the "good" starts. Anyone else notice any odd behavior?

Any way to go through the code to see if any oddities exist or anything that looks suspicious? I just can't think of what would cause it to work fine at one point, then not at another..it just makes no sense.

Post by **muythaibxr** » Tue Dec 18, 2007 3:09 pm

Most likely I think what's happening is that the stepper goes straight to the running value b/c when the wheel decoder first syncs, rpms go a little higher than they should...

Because you're not getting the right amount of air in for the fuel you're injecting at that point, you get black smoke... so the amount of fuel is probably not changing, just the amount of air.

You could power it up and wait a second until the stepper stops moving, then crank it. If that works then it would point towards what I'm saying.

Ken

md95 · Post by **md95** » Tue Dec 18, 2007 3:33 pm

muythaibxr wrote:Most likely I think what's happening is that the stepper goes straight to the running value b/c when the wheel decoder first syncs, rpms go a little higher than they should...

Hmm what do you mean by this? By rpms going higher than they should, does that mean that the rpm read by the MS are not the actual rpm, but it thinks it's higher than actual engine rpm? If the stepper is going straight to running mode, then wouldnt you think the MS is going to running mode with fueling as well? (this is what to me, really seems like what is happening)

muythaibxr wrote:Because you're not getting the right amount of air in for the fuel you're injecting at that point, you get black smoke... so the amount of fuel is probably not changing, just the amount of air.

See this is what i thought too, but it really seems to not be the case. Before i had the stepper working, i just used the base idle speed adjustment on the TB, with a closed throttle plate for set hot idle speed. Even with 3% full hot cranking fuel, it was waayy too much, lots of black smoke, very hard to start, and to get a faster start, i had to press down on the throttle to offset all the fuel(you can see the cloud of fuel vapor and smell it in the car!) Then when the IAC does what it's supposed to, it lets in the extra air, but now there is no where near enough fuel to start the engine(this is where I believe that it's NOW referencing the 3% cranking fuel value). There is no difference from me stepping on the gas to get that fast start(before i hooked up IAC) or weather the IAC is doing the extra air.. I cant see how going from way too much fuel to too little, with just some IAC bypass air could have THAT much effect for a no start, from me using the gas pedal(BTW, i'm not pumping the gas to trigger AE, just holding the gas down, then crank over)

muythaibxr wrote:You could power it up and wait a second until the stepper stops moving, then crank it. If that works then it would point towards what I'm saying.

Ken

I always wait for a few seconds at power up, just to know fuel pressure is primed, and IAC is fully retracted. It makes no difference. From my testing, the IAC ALWAYS retracts, just sometimes at crank it ignores cranking rpm, and goes straight to running steps. When this happens, there is little extra air, and something is injecting more fuel along with it because it starts up pretty good when hot(because it seems like it's not referencing the 3% cranking value anymore). When the IAC goes to my 16 steps at crank like it should, it just cranks and cranks and cranks..not even a blip from the engine.

I have tried a cranking rpm of 475-500, and didnt help. The strangest thing is, is that i CANNOT replicate the IAC going to full run steps on the bench with my test CAS. It goes right to cranking steps, just like it should. Plug the MS back in the car, crank light and rpm show cranking, but IAC steps go right to full hot steps! I dont notice any odd rpm display behavior either, like spikes or anything like that, but i'm assuming the data rate of the display gauges are a bit slower than what the processer is reading..so even if there was a very, very fast rpm spike, it may not show up, but would put the MS into running mode. Only thing about that though, is you said there is a 5 sec wait time for cranking mode, so any quick rpm spike would just be that, a spike, and it would go back to crank mode(UNLESS this feature isn't working like it should??)

One more thing of interest- When the IAC goes to full hot steps at cranking, when the engine does initially run, IAC steps back off, then go down to full hot steps! When IAC works at crank like it should, engine starts, then steps go down incrementally like normal, no backing off of steps..

I'm dumbfounded at this point, and have racked my brain trying to figure it out...i'll do more bench testing tonight, but at this point, i dont know what else to try..

UnaClocker · Post by **UnaClocker** » Tue Dec 18, 2007 5:39 pm

What is your cranking RPM set to, what do you see your engine cranking at in the logs?
Hate to revisit the basics.. But how are your grounds? Got enough of them going to the engine, got one coming straight from the battery to the engine and straight from the battery to the body? Got a good thick main ground for the MS going to a good ground point? Your engine draws alot of current during cranking, and if it's not getting enough ground, bad things happen.. I just say this, since you can't get this problem to show up on the stim..

md95 · Post by **md95** » Tue Dec 18, 2007 6:08 pm

UnaClocker wrote:What is your cranking RPM set to, what do you see your engine cranking at in the logs?
Hate to revisit the basics.. But how are your grounds? Got enough of them going to the engine, got one coming straight from the battery to the engine and straight from the battery to the body? Got a good thick main ground for the MS going to a good ground point? Your engine draws alot of current during cranking, and if it's not getting enough ground, bad things happen.. I just say this, since you can't get this problem to show up on the stim..

Right now i have 400 rpm set for cranking rpm..I've tried higher and lower, no differnce. Rpm gauge shows roughly 250 rpm while cranking in the car, which is what i spun my test CAS at with a hand drill on the bench.

For power, i have a dedicated 8 gauge power cable right from the pos battery terminal, switched by a relay from my ignition switch ignition wire. Right before the DB37, i have a 20 amp audio power noise filter on the swtiched 12 volt wire going to the MS... I have upgraded all grounds to the chassis, the engine, and ran a ground from the battery to a metal part of the center console, which is where all of my grounds are at for the MS, wideband, other gauges. The OEM ECU grounds are at that same location as well. Multimeter tests show .1 ohms resistance from the batt terminal to my ground point. It doesn't get any better than that.

You are suggesting basically the only thing I can think of at this point when comparing the perfect bench tests to in car tests, which is the sudden voltage drop from power on(lets say 12.3-4 volts) to when the starter kicks in(my volt gauge shows about 10-11 volts). Maybe the sudden current demand and voltage drop just freaks out the MS? I dont know, I just cant reason anything else..nothing makes sense..

I was going to do some bench testing tonight, but it seems my power supply took a crap..go figure.

EDIT: Speaking of grounds, it got me thinking and i doubled checked the MS ground. I have it grounded with the OEM ECU grounds, which are bolted to this metal frame piece at the center console. I have my wideband o2 ground there as well. I thought i had brought in a dedicated ground cable to that metal piece, but i dont... So...without having any other ideas at this point, i'll run a ground from the battery, then ground the MS and everything else with it. Also, the CAS get's it's ground right there as well(acording to the OEM schematics), and i want to run a separate 12v switched power wire to the CAS's power wire in the engine bay. This way i'll know the cas is sharing MS power and ground, to rule out any noise/oddities with the CAS's signals. I'm desperate to try anything at this point!

md95 · Post by **md95** » Wed Dec 19, 2007 6:14 pm

Ok, did some re-wiring today... I ran 2- 14 gauge grounds, twisted, from the battery, to the MS, and the CAS sensor ground. I ran a switched 12v wire from the MS switched 12v wire to the CAS power wire, which is right after my stereo noise filter.. So now, the CAS, and the MS have perfect battery grounds, and battery voltage.

Still no change! I then put the MS on the bench to test again with my test CAS, and of course, it goes right to 16 steps at cranking rpm. Every time. I must have tried this like 30 times, power off, on, crank.

Put it back in the car, steps go right to running. What the ****. In the car, i tried unplugging the injectors, the mitsu power transistor ignitor, the MAT sensor...nothing, no change. PW shows LESS on the bench at crank than does in the car at cranking rpm. Bench was 1.1xx ms, while in the car(with the "issue), it was 1.5ms. Not sure what to think of that, as MAT was not identical from car to bench(bench was higher), so that may be why PW was showing lower on the bench(coolant temps were very close)

For what it's worth, lowering the cranking %, has no major effect in delivered fuel when the issue persists. Lots of black smoke..3% isn't what i would call too much fuel! I dont know what to think anymore..i've tried everything i could think of.

One last test i thought of: Take off the CAS on the engine, turn the ignition on, and spin it over with the drill(i'll unplug the injectors!). This test would rule out any current draw/voltage drop from the starter.

UnaClocker · Post by **UnaClocker** » Wed Dec 19, 2007 6:58 pm

md95 wrote:
One last test i thought of: Take off the CAS on the engine, turn the ignition on, and spin it over with the drill(i'll unplug the injectors!). This test would rule out any current draw/voltage drop from the starter.

Exactly what I was going to suggest.

Post by **muythaibxr** » Thu Dec 20, 2007 8:55 am

Battery voltage does have an effect on pulse-width. Current draw from the starter would explain a higher pulse-width.

Ken

6040solder · Post by **6040solder** » Thu Dec 20, 2007 12:30 pm

Ken, shouldn't voltage only affect opening time like it does with dwell? fuel pressure is regulated so a pulsewidth is a pulsewidth to the engine.

or does that pulsewidth include the opening time?

if it doesn't, can you explain why more fuel should be injected depending on battery voltage? i cant see a reason.

md95 · Post by **md95** » Thu Dec 20, 2007 1:14 pm

6040solder wrote:Ken, shouldn't voltage only affect opening time like it does with dwell? fuel pressure is regulated so a pulsewidth is a pulsewidth to the engine.

or does that pulsewidth include the opening time?

if it doesn't, can you explain why more fuel should be injected depending on battery voltage? i cant see a reason.

Voltage effects injector open time, and injector open time effects final PW. So changes in voltage effect final PW. Lower voltages require more injector open time to inject the same amount of fuel, compared to when voltage is higher. So a sudden voltage drop from cranking the starter, would effectively increase final PW.

6040solder · Post by **6040solder** » Thu Dec 20, 2007 2:17 pm

perhaps your injector opening time or voltage corrections are incorrect then.

it would be very nice to be able to see these two seperate variables independently rather than as a clumped together single parameter.

what you care about at the end of the day if your settings are correct is how much fuel, not how much time it took to start injecting and how much fuel.

it would be nice to know that time aswell though to diagnose such issues.

is it an ini change away to datalog them? or do they not exist as available variables to suck out?

Post by **jsmcortina** » Thu Dec 20, 2007 3:50 pm

6040solder wrote:perhaps your injector opening time or voltage corrections are incorrect then.

it would be very nice to be able to see these two seperate variables independently rather than as a clumped together single parameter.

They are two parameters in all code versions I've seen?

James

md95 · Post by **md95** » Thu Dec 20, 2007 4:22 pm

James/Ken- Is there a gauge, ini edit, ect, than I can bring up to show exactly what "mode" the MS is in? For example, cranking, running, ect? Something that tells a bit more than the "cranking" light? This way i would know for sure if it really is in cranking mode or running mode, while i crank the engine over.

It just makes me wonder, if the MS is telling the IAC part of the code to go to full hot running position at cranking, then why would fueling be in cranking mode, and not in running mode along with it? I know if the fueling part really is an issue, it's obviously related to the IAC going to running steps as well.

I know this isn't a code issue, just wanted to make sure that was clear for everyone. It works as intended on the bench. If it *is* voltage drop from the starter, how would you go about fixing that? I mean i have direct power and grounds right from the battery to the MS.. A new, good battery, brand new starter.

Could the 12 volt lead going to the stepper chip be backfeeding into the MS chip? Like when the starter kicks in, the sudden voltage drop gets to the stepper chip, and then backfeeds into the MS chip somehow and freaks out the code? I'm grasping for straws here!

md95 · Post by **md95** » Fri Dec 21, 2007 4:31 pm

Well, the plot thickens! I did the latest test today. I unplugged the ignitor, and the injectors, unbolted the CAS from the engine, turned the ignition on, then spun the CAS with the drill. IAC went right to running steps! Soooo it's NOT current/voltage/starter related.

At this point, it's got to be something thats different from the bench to the car, and the only things left that I have not yet isolated are:

CAS
Fuel pump relay
Cooling fan relay
Coolant temp sensor
TPS

My first suspects are the relays. Possible backfeeding into the MS? It is rather interesting that my IDL out is ported, controlling my cooling fan relay..it does work fine though as programmed..more work to do. I'm going to figure this out.

UnaClocker · Post by **UnaClocker** » Fri Dec 21, 2007 5:26 pm

You used the CAS on the car. Maybe something about it or the wiring to/from it is funky? Try the CAS from your bench on your car?

md95 · Post by **md95** » Sat Dec 22, 2007 9:29 am

UnaClocker wrote:You used the CAS on the car. Maybe something about it or the wiring to/from it is funky? Try the CAS from your bench on your car?

The car runs great, and shows perfect RPM, which is why I dont suspect the CAS..i mean if it were bad or a wiring issue, it either wouldnt run, show correct rpm, wouldn't sync, ect.. I kind of beat up my test CAS on the outer nub that's driven from the cam, but i can at least wire it up and use the drill...My next test is disconnecting the cooling fan relay control wire, then the fuel pump relay control wire, if thats not it, then i'll try the CAS. Not much time to play around with it though until after x-mass..

EDIT: You know, i just thought about this...The 2 stock CAS signals are being shared from the OEM ECU..basically i just tapped into the crank and cam sensor wires at the ECU..The stock ECU is pulling up the signals to 5 volts, just like how the MS input circuits are wired up as well. I wonder if they are conflicting with each other..I can just have the CAS run the MS, but i'm not sure what would happen if the stock ECU doesn't see an rpm signal..At this point, the stock ECU isnt controlling much of anything, except the MPI relay, which gives power to all the engine sensors, injectors, ect..The only thing that may not work would be the cruise control..I know it uses the VSS, but since it's a manual trans car, it may need rpm signals as well... It'd only take a few minutes to cut and isolate the 2 sensor wires at the ECU from the CAS, so it only gets to the MS.. More testing! This is really a very strange issue indeed.

md95 · Post by **md95** » Sat Dec 22, 2007 8:35 pm

Wired up the "test" CAS on the engine, bolted it up, and still the same results! Goes right to running steps on the IAC.. I cut and isolated the CAS wires from the stock ECU, no change either. So the only things left would be the 2 relays i have wired in. I guess the DB-37 as well..may as well check my pin connections.. When i was searching threads about the IAC on the msefi forums, i had read where someone else experienced the same issue..nothing was said about it, or was referred to..so i guess i'm not the only one at least.