Please help. First 45-60 seconds of cold startup problems.

[The Alchemist]

Ok, I've read, read and read some more. I've searched, but no luck.

When I do a cold start, it searches bad for 45-60 seconds then settles in. The problem is, that it searches so bad, that it stalls most of the time if it's too cold.

I don't think it's a RAF fueling issue, because at the beggining of the cell, it's adding some fuel, but by then end of the cell it's pulling. Also, it's still in the same cell when it stops fluxuating. The funny thing, is if you look at my idle airflow table, it's at 25 at 20c, and it's pulling 3-4, but just a week prior, it was at 21 and adding 4.

Look at the log from 10-15, in the 20c cell, it was calling for a desired airflow of 26, but on the 11-02-08 cold startup, it had a desired airflow of ~21.

Who knows what's going on. I did notice that in my meddling, I turned on the fuel trims in hopes that it would help, but obviously it didn't.

So any help would be greatly appreciated. This is the only issue with the tune now that I am struggling with. It's weird that it only pulses/fluxuates for 45-60 seconds and then settles in.

Thanks.
Here's tune and data.

[The Alchemist]

Anybody have some suggestions???

[WS6FirebirdTA00]

Looks like your idle airflow is way too high. Target idle speed looks too low. Spark correction tables look too low. Idle timing may be too high.

[loppad]

Curious thing: 45 to 60 seconds is what takes the O2 sensors to warm up and then close the loop; if weather is cooler they took more time. Meantime there’s an open loop routine called “keep alive” to add-pull fuel-time to keep rpm in target based mainly in rpms because it can’t see AFR yet (O2 is warming). Check closed vs open loop status to confirm and also correct idle parameters as WS6FirebirdTA00 says. You shouldn’t have this issue in open loop if they’re right tuned.

[WS6FirebirdTA00]

Yeah, the lean OLFA table could be causing issues in OL on startup.

[The Alchemist]

OLFA table?? Open Loop Fuel Adjustment? Looking at my tune, at startup, it's calling for 1.03-1.04 on startup, so it's not crazy rich.

So I'm going to try this tune. I'm bumping up my startup idle to 1050 and then tapering it back down to the 825 that it normaly idles at without any problem when warm.

So how should I change the idle spark tables? My idle is all over the place for the first minute, but it's weird, the engine temp stays the same when it all of a sudden smooths out, so it's like some timer on startup that needs to be adjusted.

[WS6FirebirdTA00]

You need startup enrichment, if the car is too lean it can cause that hunting. Try your stock OLFA first to see what that does.

[The Alchemist]

But that's the funny thing, it is running rich at startup, it's actually around 12.0-12.5:1 during startup.

[WS6FirebirdTA00]

So, why do you not have the WB data posted too? That is pretty important

You have to give us ALL the info if you want accurate help

[The Alchemist]

So, why do you not have the WB data posted too? That is pretty important

You have to give us ALL the info if you want accurate help

That's my bad. I had forgot to add the wideband data to the log. I used RussK's idle airflow config, and plugged in the wideband, but realized after that it wasn't logging AFR.

I'll try and do that tonight when I get home.

[SSpdDmon]

Try adding airflow to the Startup Friction Airflow Correction modifier.

*removed link*

I know it's in HPT too, just forget the specific name off the top of my head.

[WS6FirebirdTA00]

His start up airflow is reasonable, the number you posted were extremely high. If everything else is correct, there is no need to use extreme values in start up tables. That will not change the searching idle, it should stabilize.

[The Alchemist]

WS6FirebirdTA00 (I don't know your name), did you take a peak at the first minute or so of my startup? It's weird that it just settles in, but it's still in the same temperature cell for most all of the tables. It's like there is a timer of some type that is changing something (do you like how descriptive that is) for the first 30-60 seconds during a cold startup.

I'm ready to just grab a stock tune, swap over VE tables, injector scaling, and see if the car will start better.

Unfortunately I spent last night getting her side of the garage cleaned out since we are getting freezing temps over night and she takes the kids to daycare in the morning.

Hopefully I'll be able to log some more data tonight and we can get this figured out. When I start it, I need to keep my foot on the gas sometimes at 2-3% in order to keep it running, and it sounds and feels horrible, like it's running on 6 cylinders. Again, once it's been running, it's fine.

[WS6FirebirdTA00]

Yes. I did not look through the tune in detail though. What WB do you have? Again, my first guess was the lean fueling. That or you need more aggressive spark correction from your over/underspeed tables (which would now be my best guess without looking back at the tune, with you saying your fueling was right).

[SSpdDmon]

His start up airflow is reasonable, the number you posted were extremely high. If everything else is correct, there is no need to use extreme values in start up tables. That will not change the searching idle, it should stabilize.

Sorry - thought he said he had cold start issues with not enough airflow. Regardless, those numbers aren't necessarily too high...depending on the application.

I'd suggest logging the IAC though. What you can find in a hunting idle is the IAC lags behind causing surging swings. It may be the IAC is too strong in the bigger RPM deltas. One way I fixed this was to reduce the "RPM too high" IAC direct airflow correction factors (not the learned). This helped to prevent the IAC from taking out too much airflow when the idle swung high and let it settle back down to where it was supposed to be a little more smoothly.

[WS6FirebirdTA00]

On a properly tuned car, it will not need 10 g/sec of start up airflow.

99.9% (yes, I made it up to stress my point) of surging is caused by spark (assuming fueling is OK). It is easy to show by fixing the IAC and then fixing the spark (at idle with a cammed car). The motor responds too slowly to IAC but much faster to spark. The only way the IAC is really going to be doing any damage here is from either idle airflow being too high (which I previously stated) or from idle PID controls set way too aggressive causing the IAC to move a LOT (like 30-40 steps).

Try this.

[The Alchemist]

I've got an AEM wideband and have added it to all of my other log setups, I just forgot to add it to the idle config setup.

I'll try your new tune and log the cold startup. Thanks again.