I've been playing with the idle a little bit here and there since the tuner set the thing to the stock RPM settings when he put it in the car. At idle before, it wanted to die on me, particularly on cold starts and occasionally at stoplights. It was clear then that the IAC was working overtime via the scanner, so when I had to tweak the tune a few months ago I had him raise it further to 850 at hot idle.

Well it ran better, cranked better, but the idle wasnt quite right. At the same time, the old fuel pump died and I replaced it with a Walbro 255 from Racetronix.

I decided to tinker again with the idle screw to get the IAC counts down. After a couple turns it got the hot idle counts down to about 45. Turning it further raised the idle to over 1000rpm, and I dont want that. I reset the TPS each time I played with the screw, so it does show 0.54V at idle. The TPS sensor is behaving like I think its supposed to, and the voltage increases linearly to WOT from the idle setting.

The engine has now developed a stumble at cold idle, which is very slight at cruise. When it happens, and I watch the IAC counts on the scanner, they blip 10-20 counts higher and come back down. BLMs, which I thought meant nothing at cold start since its open loop, sit in the 140s. Also at cold idle, there is no knock detection, but as I watch the base timing, it starts in the 26-30 range and drops down into the teens as the engine stumbles, then back up to 30.

The engine doesnt tolerate more than 30 degrees, any higher and it will knock. I had the tuner back it out until we saw zero knock.

Once the engine is hot and in closed loop, the BLMs do jump when i feel the stumble, but they remain in the 124-132 range. The IAC counts are at 100 or so and they blip higher as well. However at the stumble, whether the engine is hot or cold, the TPS value remains unchanged to the digits displayed on the scantool.

I thought maybe the plugs had gone, so I pulled 6 of the easier ones, all look good. Checked fuel pressure with the vac line disconnected and its reading 51-52psi. Also, the fuel pressure does remain constant as it should. TPS behaves like I thought it should, and theres no burnt plug wires, they are all less than 4 months old. There are no noticable vac leaks.

I dont know exactly what else to check here. What kind of behavior does a bad IAC (it was cleaned on the rebuild, and there was no cruise stumble before the fuel pump died) show? Usually stumbling is due to a lean condition, but i'm hesitant to raise the pressure further.

 

..... You might try backing out the timing at idle ... try 22-24 and see if that doesn't smooth things out some ... check again for a small vacuum leak too ... possibly in HVAC or cruise ctrl .............

 

Most likely a timing related issue, but mixture is also part of the equation, especially when cold, since open loop mixture is subject to coolant temp, load and calibration errors.

The 140 cold blm suggests that you're using the SAM cell data from a lean trim cell from the previous ignition cycle. This means that either the trim data stored in SAM cell A (idle) or SAM cell B (typically cell 10) was previously running a bit lean. If your 140 was in cell 0, then its coming from SAM cell A, if any other cell then its coming from SAM cell B.

This is not necessarily a bad thing, but it may not be representative of the actual cell's fuel demands.

Address the tuning of fuel trim cells 0 and 10, to be sure that they're not causing unwanted effects for the next cycle. Best to have your BLMs in order in all cells to avoid these kind of effects.

Onto spark:

This is easier said than done, but you want the timing table to act a little like a proportional controller to help maintain the desired idle speed near the target rpm.

If you have negative rpm error, you want to set the spark table in such a way to help promote a torque increase as the rpm is falling in order to help recover the idle since the IAC is very slow to react.

Setting the advance a little lower than the ideal value at the target rpm may make for more stable sweet spot.

You have to play with target rpm and the advance and watch the IAC counts or manifold vacuum to get an idea of how much timing is ideal for a given rpm, then adjust the table to help promote a return to the target rpm.

 

One thing that will help with tuning is the use of a wideband. I have an Innovate LM-1. It lets you know the afr at all times, hot and cold so you will have an idea if your motor is too rich or lean. The new ones are not LM-3s I think. They are not that expensive and will be invaluable for as long as you own the car. I have recorded over a thousand runs with my unit and if the motor misbehaves, I know in a minute whether it is fuel or spark related.

 

First thought it your tuner should be contacted.

That said, I wonder what your FP was before the Walbro install. Did that change your fuel map?

For the IAC, dropping it to "normal" counts (15-30) range shouldn't cause the idle to rise. If FP went up, I wonder if the IAC is trying to feed more air at idle for compensation of extra fuel? I also have to ask (make sure the engine is totally warm when you adjust)?

If the BLMs are high AND rpms try to increase at idle, I have to consider vac leak.

 

If it was mine,I would Test the temp. sensor cold and as it warms up. Also the cold start injector and the TPS with a analog VOM . A digital VOM will not show dropouts in the TPS very well. Good luck!

 

Temp sensor is fine and there is no CSI anymore.

 

Your idle goes up when you move the screw to get those IAC counts down. The idea is to get the valves open further so the IAC doesnt work as hard to maintain the idle. I dont see what youre saying. I readjust the TPS afterwards.

FP wasnt changed when the new pump was put in.

Yes the engine is hot when I play with it.

 

I was trying to think of a reason why the ECM would allow higher-than-commanded idle speed.

Your description of IAC is correct and I agree with it. And, as you go lower (to the 15-30 count range), your idle should not go above the commanded idle speed in your BIN. For it to do so, makes my think you're getting more air than the ECM "sees".

Or, that the ECM has an inappropriately high value for idle speed?

(I wonder if the ECM needs to be disconnected? Does it carry (ala BLMs) some long-term compensating value that commands a higher-than-coded idle speed?

More air is what makes the motor run faster. After all, it only takes the IAC to open and let more air in -- for the engine to pick up speed.

Even if you close your IAC to 0 steps (by opening your TB set screw further), your idle should not increase until you go beyond the point where the IAC hits zero.

That's the way mine works.


Note: The reason temp sensor was suggested is the ECM would tell the motor to idle faster IF the ECM "thought" the engine was colder than it really is. And there are 2 temp sensors. The one displayed on your dash is not the same one monitored by the ECM.