Stalling at start-up

[Slick1100]

My car is acting strange lately. If the car is already warm and I start the car without the A/C on, it immediately stalls unless I have my foot on the gas and rev it for a second. If I start the car with the A/C on or it is totally cold, it starts and idles fine. Once the engine RPM has stabilized, it idles fine until I go to restart. No engine codes are stored.

I checked the TPS, and it is within limits, though it acts in reverse of what my book says (Haynes POS). The IAC solenoid ohms out at 10 ohms, though the book says is should be 9. I don't want to change this though unless it is definitely the problem. What else can I check? Call me silly, but I don't like starting the car with the A/C on all the time.

[Legacy777]

probably the coolant temp sensor.

the reason the car starts fine with the A/C on is that the computer gives the car extra fuel because it's under more load.

Replace the coolant temp sensor, it's like 20 bucks at liberty subaru www.newsubaru.com

[Slick1100]

I take it the coolant temp sensor can indicate fine, but be bad? Or does the indication come from a different sensor?

[vrg3]

Yeah, the gauge and ECU use different sender units.

[ciper]

ECU sensor is two pin, and I think its 17mm while the dash is single pin and 10mm

[rallysam]

Really? A faulty coolant temp sensor can screw up the mixture enough that it won't start?

[ciper]

If the car is cold already its running static fuel maps right?

[Legacy777]

If the car is cold already its running static fuel maps right?

yup I believe so.

[vrg3]

When it's cold it has a "water temperature increment coefficient" that increases fueling until about 140 degrees Farenheit. Since that means it will run rich, it can't run closed loop.

[Slick1100]

Just a little update... I replaced the sensor, and the car starts better, but not like it should. I've noticed that it's idling very low when the A/C is not on (say 500 RPM or less on the tach). Any thoughts? The alternator recently died and was replaced, but that didn't make any difference in idle, though I've noticed my lights don't dim as much now when idling.

[scottzg]

according to the tach, i idle at ~400 rpms. I heard somewhere that 500 on the tach is really 750. How is that wrong? is it shaky?

[vrg3]

Yeah, the upper mark between 0 and 1 is 700 rpm. I believe idle speed is specified as 700rpm +/- 100rpm. You can see Laurel's car idling correctly here:

http://www.imagestation.com/picture/sra ... 3d4437.jpg

If the idle is unsteady, then that's a problem. Does it ever feel like it might stall out?

[georryan]

I know my car lopes at idle a bit. I think new plugs and wires and correct gappings will fix it, but I haven't done it yet. It has been this way since I've bought it and I"ve put on close to 20k miles on it. So It may be due.

-Ryan

[ciper]

The lowest line on the tach is 500. The line that it usually points is half way between the first line and the 1000 rpm mark, or in other words 750 rpm.

[Legacy777]

Our tachs are not very acurate.

Watching the idle on the select monitor, it's right around 700-675, and the needle shows 500.

The loping idle can be caused by a couple things. However my idle has gotten a lot better since putting the alternator in.

I think injectors have something to do with it too. I will hopefully take care of those too....

[ciper]

Thats not 500, its 750.

[Slick1100]

Ok, I splurged a little today and got a tach and ignition tester (helps that I could put this on my account at work, as it's a bit pricey!). Idle is 700 RPM. The kV is the key to where the problem lies. Here's what I found:

#1 ignition wire: 5 kV
#2: 25 kV
#3: 9 kV
#4: 30 kV

That's right, the left cylinders have low volts and the right have good volts. Sounds like the coil is bad to me.

[vrg3]

I'm not sure you're measuring/interpreting right.

We have a waste spark system. So, for example, when either rear cylinder needs ignition just before its power stroke, the rear coil fires. The spark travels out one coil tower, through one cylinder's spark plug, into ground, out the other cylinder's plug, and into the other coil tower. So, there are two sparks, one in each cylinder. The spark actually burning air/fuel is fairly strong and the other "wasted" spark is fairly weak. Around 25-30 kV for the combustion spark and 5-9 kV for the waste spark actually sounds right.

So I suspect your tester is measuring the waste spark voltage for cylinders 1 and 3. It may be triggering wrong or it may just not be meant for use with coil pack ignition systems.

In any case, it's not really possible for a bad coil to cause the left cylinders to spark correctly while the right cylinders don't (or vice versa), since the spark has to go through both cylinders' coils. If one coil goes bad you'd get a good spark on the front cylinders but not the rears or vice versa.

[Slick1100]

I'm not sure if you're familiar with it, but I'm using a SmartTach + to measure the ignition voltage. The instructions say that it measures "peak" voltage. If I'm understanding you correctly, then the voltage is high (20-30kV) during the compression stroke, and low during the exhaust stroke. Still, if the voltage read is supposed to be "peak" then the left cylinders are both low. I suppose it is possible that the reverse polarity of the spark plug on the exhaust stroke could throw off the readings, but it seems unlikely to me. I guess the next step is to ohm out the coils and wires and see if any problems are indicated there.

As a side note, oddly I got almost 30mpg on my last tank of gas! Then again, I have been stuck in traffic (LOW speeds) alot more lately.

[vrg3]

I'm not sure if you're familiar with it, but I'm using a SmartTach + to measure the ignition voltage.

I'm not familiar with it at all... I'm just telling you what I know about our cars' ignition system.

The instructions say that it measures "peak" voltage. If I'm understanding you correctly, then the voltage is high (20-30kV) during the compression stroke, and low during the exhaust stroke. Still, if the voltage read is supposed to be "peak" then the left cylinders are both low.

You understand me correctly.

How does the SmartTach+ connect to the car?

If it uses some kind of trigger from the primary side of the coil, then it's possible that it measures peak voltage during a certain interval or something. I dunno.

I suppose it is possible that the reverse polarity of the spark plug on the exhaust stroke could throw off the readings, but it seems unlikely to me.

I don't know if polarity really matters so much... It could be... the polarity doesn't actually change between the strokes; each cylinder always has its own polarity.

I guess the next step is to ohm out the coils and wires and see if any problems are indicated there.

Like I said, both left and right coils and wires are used for every spark, so I don't think you'll find a problem there... If anything, you may have bad or incorrectly gapped plugs in your #1 and #3 cylinders (a narrow gap ionizes at a much lower voltage than a wider gap).

[Slick1100]

The SmartTach + is inductive, and has no physical connection while reading RPM or ignition voltage. For RPM, you simply tell it the number of cylinders and 2 or 4 stroke, and hold it near the engine. For reading ignition voltage, there's a mast with a "U" shaped piece at the end that gets looped around a sparkplug wire, or ignition lead. Very easy to use, and so far seems to be accurate.

I suppose my next step is to ohm out the coil and sparkplug wires and check the gap on the plugs. It's possible that the gaps are off on the #1 and #3 cylinders, as I haven't checked the plugs since they were installed when the engine was changed. I don't exactly trust the mechanic that did the change, but I didn't have the time to change it myself. Hopefully I can find out something soon