Leather_Belt

Hey guys, this seemed like the place to source help on my issue since I don't believe it to be MS related.

I have a 1994 Miata running a 2003 VVT Miata motor using MS3Pro and an AEM wideband (30-4110, LSU 4.9... however you identify it). My build is pretty much the The Definitive "VVT swap into 90-97 chassis" Megathread to the T. I've deleted the EGR and have a 2.5" catback exhaust. I've left the VTCS butterflies in the intake and am using the NA8 intake runner and MAF, NOT the IAT sensor. The car has a decent street tune and I took it to autocross for the first time a couple weeks ago. Sadly a couple days later when I started to the car to leave work, it would crank and start normally, then it would go full lean and stall out. This would happen repeatedly until maybe a certain temperature I guess? It would finally "idle" about 200rpm higher than normal, while still displaying full lean on my wideband.

While the car is under load, fueling looks like it returns to normal and actually follows the AFR table. But right when it goes into neutral, it will go back to full lean and try to die. There kind of feels like a little less power. Pressing the gas pedal also causes a strange kind of delay in throttle input. In neutral, blipping the throttle causes the revs to go up after a delay and the AFR gauge reads slightly rich rather than full lean. After the revs go up though, they'll hang for a second, then it will go back to full lean idle.

I've pulled the rear shelf carpet back to expose the top of the fuel pump. It audibly primes for two seconds when the key is 'ON' and it can be heard working while the car is running. Could it be 'weak' or about to die?

I've replaced/cleaned a few things, but nothing has made a difference.

• cleaned MAF and IAC valve
• new plugs (properly gapped) and wires (did not replace the two coils on the motor, just the two actual wires)
• checked for any vacuum leaks twice
• replaced fuel filter
• did timing belt, water pump to see if it somehow jumped too many teeth

I think that's about all I've done. My next thoughts are, in no particular order:

• fuel pump
• injectors
• FPR
• coils
• broken MAF
• wonky O2 sensor
• somehow a critical wire has lost its shielding / solder connection and is now grounding out somewhere in the bay

Attached is the tune I'm using as well as multiple data logs to display what's happening. Maybe someone else can see something that I can't, though I can't see how it would be a tuning problem when the car worked fine the day prior. Maybe there's something in the logs that will direct me towards what to start looking at.

Thanks guys

SpartanSV

The MAF does nothing when using speed density and a different IAT sensor.

If it runs fine outside of idle it's unlikely to be a pump/fuel system issue.

The car dying at idle when the O2 sensor shows 18.0 means the wideband is probably fine.

If it's not misfiring it won't be a coil issue.

Have you not tried to adjust the tune yet? It requires minutes of your time and zero doll hairs.

albumleaf

iTrader: (7)

Turn off EGO and see what happens.

Morello

Well, your tune is configured to run speed density, which means your MAF is doing nothing. I'm not sure from your description what you mean when you say you're not using an IAT sensor, but speed density uses air temp to adjust your fueling when making air flow calculations, so that could be part of your problem. If you turn off EGO and jack up the fuel VE in that area will it idle? This might tell you whether it's a tune or hardware issue.

Another issue I had with closed-loop idle control was the setting for "use last value." If the last value was garbage, it will continue to run like garbage. I had much better luck switching to "use initial value table" and setting target values. This won't help with your lean issue.

Leather_Belt

Sorry, by 'not using an IAT sensor', I mean I'm not using the GM IAT sensor and deleting the stock MAF.

Again though, everything was fine and then suddenly was not.

I've looked in TS and tried to change a few things, as well as checked timing and again looked for vacuum leaks. The only thing that yielded any results was to give the EGO correction a little more control (from 4% to 15%) as well as getting the car to warm up.

I've discovered a new potential issue as well, hopefully it's related to my issue. My stock coolant temp gauge won't hit 12:30 like it normally does.

I'm going to check the two coils on the motor to be 100% sure of those.

Here's a datalog and a tune used when I got it to kind of idle after warming up. I drove it around after I got it to idle and it seemed fine? No weird noises or feelings while the car is under load. AFRs appear healthy as well.

Thanks guys I appreciate the help. I'm pretty stumped here.

SpartanSV

You're still lean. I wouldn't mess with anything on the tune before you tune fuel. It's less lean in the most recent log because your coolant temp is high enough for ego to be active and you gave it 16% authority. If you look at the log it's adding a full 16% the entire time.

Add more fuel to your tune.

Leather_Belt

I don't understand how there's a need to do that. Like I said, the car was running great on its street tune, now suddenly it doesn't want to at all.

Would adding more fuel to my table just compensate for a possible mechanical failure?

SpartanSV

Do you have a log of the car idling great previously?

Something like a vacuum leak could absolutely cause this.

I've already gone through and explained why the list of things you think might be causing it are unlikely to be at fault. It also don't think your current coolant temp gauge issue is related. The ecu is getting a reasonable coolant temp. If the bp6d is like the bp05 the gauge is fed by a completely different sensor.

Leather_Belt

Sorry this is the best I can do. This is the most recent data log I have of the car running well. I didn't create any data logs of it running at its peak when the tune felt pretty 'done'. This should show the car idling pretty normally though.

Are there any secret vacuum lines I've missed? Some kind of VTCS failure? I really don't know where air could suddenly be getting in.

Thanks

curly

iTrader: (25)

So above 178*, your air density table starts pulling fuel, 20% by 248* coolant. So if you warm up to 190*, that's about 8% fuel. I'd start by setting that table to 100% above 86%. It ran fine above 178* with this tune before though?

curly

iTrader: (25)

Yeah in the first log you posted, you go from 137% to 109% before you die. You're too lean even at 127%.

With a MAF based system you can have zero leaks post MAF. Get a smoke tester and see where there are any leaks.

SpartanSV

I was interested in seeing the MAP values at idle. If you have a significant vacuum leak I would expect to see a change there. Not really seeing it.

I would add fuel since you know you're lean. If you have no issues after that then drive more and worry less.

SpartanSV

He's only using the MAF for intake temp.

Leather_Belt

I mean, sure... but why would this just randomly happen and then suddenly become the norm? Do tunes just do this? Even my tuner doesn't see how it could be a tuning issue.

Also, shouldn't the ECU be able to keep the motor alive with ASE? It dies pretty much immediately before the ASE can fully complete. Then after that shouldn't the EGO correction do something to keep the motor alive as well rather than just leaning out, sputtering, and dying? Why is the motor able to stay running only after it warms up?

Sorry this just does not seem like an ECU issue to me. Is my FPR failing for some reason maybe? Weak fuel pump?

edit: Idle Control Valve failing?

SpartanSV

ASE adds whatever amount of fuel you tell it to. If that isn't enough the engine shuts off. EGO adds as much fuel as you allow it. I already pointed out that EGO is pegged at the max amount of fuel you have allowed it to add and it's still to lean because your idle VE is fucked.

I already explained why it's unlikely to be a fuel system issue.

Leather_Belt

The question I need answered at this point though is 'why'

The car was fine two weeks ago now suddenly the tune is inadequate.

Why? I changed LITERALLY nothing.

SpartanSV

Votlage issue? Look at your battery voltage in the good log compared to the other logs you posted.

Leather_Belt

Oh yeah interesting. On the dying logs there seems to consistently be lower voltage? Whereas the correctly idling log drops at cranking and returns to normal.

Could this point to a weak alternator or battery or is that not enough evidence to support that?

SpartanSV

Shouldn't be battery. Are you using an NA alternator?

I would measure voltage at the battery and injectors to confirm it matches what the ecu is seeing.

Voltage differences can make a pretty significant change in fueling but AFAIK that shouldn't be an issue if the ecu is seeing the same voltage as the injectors and the injector dead times are set correctly.

SpartanSV

Bingo. Voltage drop caused an issue because someone fucked with the injector dead times. Compare your dead times to the DIY basemap.