difficulty starting engine

[dan macmillan]

1. I have a 98 legacy brighton with a stock NA 2.2l

   The problem, on first startup of the day it will crank {turn over} endlessly with absolutely no signs of trying to start. On the second attempt ie turn the key off then retry, {key must be turned to off} it fires instantly and usually continues to run. If it is cold {below 40 deg F} it will sputter a bit and I may have to restart again. I have determined that on the first try I do not have to roll the engine over on the starter, just momentarily bump the starter. Cycling the key repeatedly to prime the fuel system does not help. The starter must be momentarily engaged. There are no engine codes and the CEL illuminates when the key is turned to on as a bulb check. I have done the following. New plugs, new ign coil, new plug wires, new fuel filter, new battery. Scan data shows the coolant temp is showing ambient temp. Fuel pressure is well within specs, inj pulse is present as is spark during the first starting attempt. I do not know if cam signal is present during the first attempt. I have searched and found a few people have similar issues but have never found the cause or the fix. Please help as this is not normal and is driving me nuts.

 

[1 Lucky Texan]

I have read of one or 2 people needing to clean debris from, or replace a Crank Position Sensor. als, kInda wonder if the car might start if you held the gas pedal on the floor.

 

But, it does seem very much like a problem in the ignition switch or a wiring harness....

 

 

very odd.

[dan macmillan]

Pedal position has no effect. Like I said the only thing that allows it to start is to momentarily engage the starter then key off and restart. The engine does not have to roll over a full revolution on the first crank attempt. It just needs the starter to engage. I don't drive it in the winter so it sits for a while. This is when I must do it multiple times. But yesterday it was about 32 deg F and my son moved it down the driveway. {never warmed it up} A few hours later I turned the key to crank. No signs of it trying to start. Turned key off then immediately back to start. Fired up perfectly and idled like a new car. Drove back down the driveway without a hickup.

[1 Lucky Texan]

on the times when it cranks on the first key-turn, does it crank slowly?

 

if the headlights were on, would they dim on that first attempt? - stay bright but dim on the second (successful) attempt?

 

suppose you jumped it with another car/battery on that very first attempt? would it start immediately?

Edited April 8, 2015 by 1 Lucky Texan

[ccrinc]

Kinda going out on a limb here, but it sounds like there is a problem with the teeth on the starter and flywheel engaging properly.  Like a small chunk may be chewed off the starter pinion gear or the gear is worn (it is a normal "wear" item).

The problem would have to be on that side or the issue would be intermittent (like my mom's old Ford which had a tooth missing off the flywheel).

 

What makes your situation more curious however, is that you can replicate it daily and that temperature affects it.   It's obviously not electrical/electronic in nature due to lack of CEL and the fact that you've pretty much checked all that and verified it good.  Therefore, it must be mechanical.  Come to think of it, the starter solenoid wouldn't throw a code either, nor would worn contacts.

[dan macmillan]

It has lots of battery power. spins over at a normal rate and will for a long time. There is nothing wrong with the starter teeth. It engages perfectly. I am a licensed automotive/truck and coach/heavy equip mechanic with over 25 years experience and have never seen a car behave the way this one does. It just bobbles my mind. That is why I am asking for help. Maybe someone out there that is more familiar with Subarus can shed some light on this. I don't think it is so much temp related in the way some are thinking. Yes temp is having an effect. On the hottest summer day it will do the same thing but on the coldest winter day it will take multiple attempts before it actually starts. When I warm it up to normal operating temperature it starts usually on the first attempt. However on the next morning [cold engine] it goes back to needing the starter momentarily engaged. Cycling the key on and off to prime the fuel system has no effect on it, I have installed a fuel press gauge and during my first key on attempt the fuel pressure rises to 33 psi and stays there until the engine starts. I have searched many forums and  found people with similar experiences but no posted solutions. 

Thank you in advance for your help and suggestions. Hopefully we can solve this so others can benefit from the results.

[bertmann73]

there are 2 wires on the positive battery terminal. one goes directly to the starter and the other to the electical system....just maybe...the starter may be getting initial full contact and then arching to the smaller wire on the second try from a loose or dirty contact......all i can think of....weird indeed

[1 Lucky Texan]

there are 2 wires on the positive battery terminal. one goes directly to the starter and the other to the electical system....just maybe...the starter may be getting initial full contact and then arching to the smaller wire on the second try from a loose or dirty contact......all i can think of....weird indeed

 

 

yeah, I was trying to dream-up some way the vibration from the first attempt was pulling on a wire or ground connection and making better contact. Maybe before the first bump of the day, he could get under the hood and tap on the starter and push/pull on some cables????

[Silent Bob]

Check the ignition switch.  Make sure the "ignition" circuit is not dropping out and losing power when you turn the key to the "start" position.  Should be a red wire at the ign. switch connector.

[Fairtax4me]

You've checked injector pulse with a noid light? Spark with a spark tester?

 

If the cam or crank sensors were an issue you would be missing injector pulse and/or spark.

 

A common issue on the mid 90's cars is a cracked solder joint in the MAF sensor. That will be affected by temperature and vibration and will make the ECU lean the A/F mixuture. This failure typically displays itself as a random stall while the engine is running or while driving, but it can affect starting as well. Wiggling the wire harness to the MAF sensor can sometimes reproduce the problem.

If the MAF is the issue, unplugging the sensor before starting should allow it to start on the first try.

This problem doesn't always set DTCs, due to its intermittent nature.

[Aabuscap]

I have a 99 Impreza Outback 2.2 with a very similar problem except that on the first start of the day it starts with no problem. Once it's warmed up and shut off and allowed to sit for 15 to 20 minutes or more I then have the "continual crank with no start" problem. It takes 2 or 3 tries before it will start accompanied with some gas pedal pumping. Also did all same remedies ie. new plugs, wires, valve cover and plug gaskets etc.

[1 Lucky Texan]

I have a 99 Impreza Outback 2.2 with a very similar problem except that on the first start of the day it starts with no problem. Once it's warmed up and shut off and allowed to sit for 15 to 20 minutes or more I then have the "continual crank with no start" problem. It takes 2 or 3 tries before it will start accompanied with some gas pedal pumping. Also did all same remedies ie. new plugs, wires, valve cover and plug gaskets etc.

 

next time it happens, see if holding the pedal on the floor (the ECU's 'clear flood' signal) helps it start. That at least would tell you if it's flooded/overfueling.