Cold start missfire 2000rpm with video

[CyFi6]

My car has had this misfire after cold start for as long as I can remember. I have rebuilt this engine twice, have had stock and different piggback computers on it, and it has always done it. When I get in the car in the morning and drive off, the car feels jerky in first and second gear right around 2000 rpm and it is quite annoying. After 3-4 minutes of driving the misfire is completely gone and the car drives normally, but every time I drive off in the morning the car is jerky for the first few minutes, almost feels like when you let go/reapply the gas on a manual trans car and the car bucks back and fourth.

The colder it is outside the worse it gets, in fact, in the summer when ambient temps don't drop below 100, its barely noticeable, but when it gets cooler out it is very noticeable.

Things I have checked or fixed include:
Fixed any vac leaks, vf voltages are normal once fully warmed up
Replaced injectors with RC 550's
Replaced CPS with NEW
Fixed all broken connectors on the harness
Grounded the igniter
Adjusted TPS
Set timing to 10deg
Checked plug wire resistance/inspected plugs/checked ignition coil resistance-swapped coils for another used set, made no change

My current setup is 550cc injectors, AFPR, Walbro FP, stock AFM with SAFC, recirculated bypass valve, rebuilt engine with all valve adjustments checked. Like I said earlier the problem was always the same, even when 100% bone stock, and even when I used a MAFTPRO with speed density.

Not sure what else to check. I have basically put up with this problem for as long as I can remember but I would really like to fix it.

Here's the vid, you can see it misses right between 2000 and 2500 RPM, if you go above or below that it generally will not misfire. At ~2200 rpm, it feels like I have to really press the gas hard for the rpm to increase just a little bit, almost like the engine doesn't want to accelerate. You can see my vacuum gauge and A/F gauge. AFR sticks around 14.7 as expected.

http://www.youtube.com/watch?v=zqSSXIycL_M

Forgot to mention that there are no stored codes.

Thanks let me know if you have any ideas.

 

[CyFi6]

Bump any ideas? I should add that when the mixture is rich, the car drives smooth. When its very cold outside, sometimes it takes up to a minute or so for the air/fuel ratio to go to 14.7, and the engine runs smooth. As soon as the mixture reaches 14.7 it starts to buck and miss only at 2000 to 3000 RPM until the engine is fully warmed up, then it is smooth through the whole rpm range.

 

[BryanDyer]

I know you said you've rebuilt it twice, but have you done a compression check? Last time I had similar symptoms and went through and checked everything like you have and found out finally it was a bad valve. Just a thought.

 

[IndigoMKII]

what does the vf do when it misfires? how much correction are you adding with the 550 on stock afm?

what is the gap of your plugs? what is your base fp? it almost sounds like the ecu enrichment is causing it. would it be possible to replicate it by unplugging the ecu water temp sensor on the neck?

 

[CyFi6]

Yeah, I have actually checked the compression, they were all pretty even and right around 175PSI in all cylinders. I also had a valve job done on the most recent build just in case. Valves were all adjusted when rebuilt as well.

I am not sure about the vf when it misfires, I can take a look, but it was my understanding that the vf voltage should only be read after fully warmed up? Plugs are NGK coppers (heat range 7) gapped at the stock .031". I've run stock platinums with the same gap and same results.

I can try unplugging the sensor and see how it changes things. It seems like as the ECU stays in open loop and doesn't aim for 14.7 the engine runs smooth, problem is that it generally gets into closed loop almost instantly and causes the engine to hesitate until the engine is almost fully warmed up. As far as I know there isn't anything that would cause the ECU to go into closed loop too early though?

Here's a pic of my SAFC settings. The way it is right now is pretty much dead on for vf's (2-3v at idle and while driving) and the drivability is pretty damn good. Fuel pressure is 35 PSI with vacuum.
These are the low throttle settings for regular driving, I haven't finished tuning WOT.

-10% the rest of the way

 

[supraguy@aol]

One of the temp sensors at the thermostat neck is responsible for telling the ecu that it's at the proper temp for closed loop.
Maybe that's bad? Or your cold start injection circuit? Just thinking out loud.

 

[Poodles]

IIRC ignition timing is retarded during cold start and it's pretty harsh to the ignition system. Have you looked into the coils/wires/plugs?

 

[Nick M]

When I get in the car in the morning and drive off, the car feels jerky in first and second gear right around 2000 rpm and it is quite annoying. After 3-4 minutes of driving the misfire is completely gone and the car drives

If you don't have a correlation problem to actual ambient temps and interpreted temps with IAT and ECT, then something is leaking in the chamber over night. Like say, oil or fuel.

 

[jetjock]

... When its very cold outside, sometimes it takes up to a minute or so for the air/fuel ratio to go to 14.7...

Need to fix that...

 

[CyFi6]

IIRC ignition timing is retarded during cold start and it's pretty harsh to the ignition system. Have you looked into the coils/wires/plugs?

I checked the coils a while back and all the primary resistances were in spec, couldn't find any cracks on any of them. I had a spare set so I changed them over anyways, these also in spec, and made no difference. I just changed the wires and changed the plugs to iridium's, no difference in the problem. Also, base ignition timing and valve timing are dead on.

I have checked the IAT and ECT before, but its been a long time so I will check them again. I'll check the wiring between the ecu and the sensors as well.

 

[3p141592654]

The decision to go closed loop after cold start hinges mainly upon coolant temp > 45C and Ox sensor transition to rich. When those two conditions are met it will start cross counting. Prior to that the ECU tries to get the fuel close using wall wetting algorithms and such like any other professional fuel management system.

I would agree that you should check the ECU coolant sensor. They don't usually go bad though but a code will only be set if it is shorted or open.

 

[Nick M]

Pull a couple of plugs when cold and look. It doesn't cost anything but time. And if they are clean, then you don't have to worry about fixing anything in that area.

 

[Nick M]

Resolution?

 

[CyFi6]

Iat/ect and their wiring both check out. There's evidence of some slight oil burning but Im not convinced that is causing this.

 

[Nick M]

If you have oil on the plugs when cold, it will misfire. But you could clear it out in 20 seconds if it was just oil by increasing the throttle angle.

 

[3p141592654]

You may have already done this, but a secondary ignition analysis would be a useful diagnostic here.

 

[CyFi6]

The plugs weren't wet and neither were the pistons, there's just some light white crust on the ground electrode which tells me there's some minor oil burning happening. Doesn't smoke on startup or anything.

Unfortunately I don't have the means to do any secondary analysis. I could potentially take it somewhere but it would be sort of difficult, they would have to let the car sit out overnight and monitor it for the first minute or two after startup, if they miss it they would need to let it sit overnight again, not to mention I would have no idea who to take it to. I have a DSO but I have no way of monitoring the secondary voltages.

 

[3p141592654]

Then look at the primary with your DSO.

 

[CyFi6]

Yeah I might be able to do that, I need to learn how to properly use it first though. I also remember that even the primary voltage spike was exceeding the safe max voltage for my DSO, so I would need to somehow attenuate it. Biggest issue is I need to learn how to use my particular DSO which I haven't really been able to figure out. I guess that is my next step.

Is there anything else to look into/test minus the ignition analysis?

 

[jetjock]

I'd get the mixture right first...