2003 Accord V6 hard start?

[cchase]

I'm trying to diagnose a problem with a 2003 Honda Accord 3.0 V6.

This vehicle has about 54,000 miles on it.

When it is cool (below 15 or 20F) out, the car will just about not start. This is repeatable and predictable. Since new it has had episodes where it will crank and crank and then finally catch.

This winter, it will crank and crank and crank and crank for 5-10 seconds before I decide, that's enough. Wait a few seconds, then try again, same issue, eventually it "may" catch and sputter for a few seconds, then die. If you put your foot on the gas a bit, it will sputter to life, and then you have to keep the RPM's up above about 1500 or it will die again, for the first 20-30 seconds.

After it finally catches, 3 idiot lights are on on the dash (none of which are related to the engine). They are all in a row, one is TCS, one is a brake warning, and I don't remember what the other is (other than it's not an engine related idiot light). These lights will remain on (solid) until you shut the car off, then restart it. They only come on during these episodes where the car doesn't want to start.

The car cranks over eagerly (battery has plenty of life). Air filter is new within the past 5000 miles, car sees mostly short trips but the engine is clean. I tried running injector cleaner through the gas a few months back thinking maybe an injector was just dirty.

What are my next steps? It's kind of embarassing when my truck pops to life within a second under similar circumstances and here is a "new" Honda struggling to life.

 

[Quest]

You need to have this in your mind to begin with:

In order for a gasoline IC engine to work (be it warm/cold), you need (a) spark at the right moment; (b) air; (c) fuel-properly calibrated.

With these 3 points in mind, you can now go right ahead and find out what's wrong with your Honda: e.g. is it getting enough air? are you getting proper, consistent sparks to ignite? are you getting enough fuel (look from injectors backwards: proper fuel pressure in your fuel rail? fuel pump pressurising ok? fuel filter clogged? fuel pressure regulator working as expected?

From this point onwards, you work your way up towards various different things such as EGR issues, air temperature sensor, thermostat, etc.

Good luck!

Q.

BTW: are you having this problem consistently regardless of your engine temperature? or is this just happening when it's only cold?

 

[Jim 5]

This exact thing happened to me with a nearly empty tank after a long trip in cold weather. I parked my wife's Honda Odyssey (basically the same engine) with the rear pointing towards the garage on our sloped driveway.

It has never been parked with the nose down on the driveway.

Van wouldn't start the next morning and couldn't make it start and had it towed to the dealership. The next day it started for the dealer.

Dealer couldn't figure out what was wrong, and the best suggestion I have had is that I had some water in the fuel and that the fuel pick-up must be near the front of the tank. This makes sense as it is cold weather and my wife does many short trips. Also, we bought fuel out of town at a fairly sketchy place, and the tank was empty overnight, and we had a ton of fog (humidity in the air).

In any case, I ran some gas-line antifreeze through a couple of tanks of gas and the problem has never come back. Not sure if you are doing many short trips, but in a cool and humid environment, water can condense out of the air remaining in your tank and over time can settle out in the bottom of the fuel tank after the tank has been still overnight.

This might be an overly simplistic reply, but it's worth a shot to run a bottle or two of gas line antifreeze into a full tank and see if this might solve your problem.

 

[440Magnum]

Wild guess: a temperature sensor somewhere (intake air, maybe engine coolant temp?) is telling the ECM that its warmer than it actually is, so that the ECM isn't enriching the fuel flow as needed for a cold start.

 

[cchase]

Originally Posted By: Quest
BTW: are you having this problem consistently regardless of your engine temperature? or is this just happening when it's only cold?


It's been bad to the point I am mentioning it only when the engine is cold. However, randomly (and it has done this since new) it will be difficult to start regardless of temperature and season.

Originally Posted By: Jim 5
This might be an overly simplistic reply, but it's worth a shot to run a bottle or two of gas line antifreeze into a full tank and see if this might solve your problem.


This seems like an easy enough thing to figure out. However, I would have thought the injector cleaner would have served a similar function while it was in the tank and that didn't solve it. I'll give dedicated anti-freeze a try though.

Originally Posted By: 440Magnum
Wild guess: a temperature sensor somewhere (intake air, maybe engine coolant temp?) is telling the ECM that its warmer than it actually is, so that the ECM isn't enriching the fuel flow as needed for a cold start.


This thought occurred to me. Not sure the best way to find out though. Any ideas?

 

[typ901]

I had a similar problem with my '00 Accord. There was a EGR valve TSB issued. When the temperature would fluctuate, it was very hard to start, similar to what you had and I had the CEL come on too. It was a fuel blend/temperature issue. It pops up in fall/spring when we might have a 20-30 degree temperature swings for a week.

 

[chrome]

I believe there was a TSB on the 03-05 Accord V6 for a fuel pressure regulator that bleeds down far too soon, causing a hard start situation.

 

[cchase]

Originally Posted By: chrome
I believe there was a TSB on the 03-05 Accord V6 for a fuel pressure regulator that bleeds down far too soon, causing a hard start situation.


Any place I could find out if this TSB applies to this vehicle?

 

[Jim 5]

Originally Posted By: cchase
Originally Posted By: Jim 5
This might be an overly simplistic reply, but it's worth a shot to run a bottle or two of gas line antifreeze into a full tank and see if this might solve your problem.


This seems like an easy enough thing to figure out. However, I would have thought the injector cleaner would have served a similar function while it was in the tank and that didn't solve it. I'll give dedicated anti-freeze a try though.

I'm not sure. If fuel injector cleaner contains enough methyl hydrate, then it could facilitate the "absorption" or "soluble-ization" of the water into the fuel as gas-line antifreeze does (forgive the likley improper understanding of the process). I'm just not sure what the carrier is for the detergents in the cleaner.

 

[mechanicx]

I think it sounds like you are not geting good fuel supply like from a bad pressure regulator or fuel pressure bleeding down. You should check fuel pressure when it doesn't start and you can spray a little bit of starting fluid into the throttle body to see if it then starts.

Also this part made me wonder if you might have an electrical problem:
Quote:
After it finally catches, 3 idiot lights are on on the dash (none of which are related to the engine). They are all in a row, one is TCS, one is a brake warning, and I don't remember what the other is (other than it's not an engine related idiot light). These lights will remain on (solid) until you shut the car off, then restart it. They only come on during these episodes where the car doesn't want to start.


So also check that you are getting spark during cranking.

 

[mechtech2]

Spray some starting fluid in the intake on a cold start.
It she pops right away, you know it is a fuel problem.
I suspect fuel.

But it is always good to check all vacuum lines, PCV, and intake for leaks. This is basic.

 

[DC84]

check for TSB's on the fuel pump relay. IIRC there was something about extreme hot/cold starts awhile back. I know that while working at a Honda store we replaced several.

 

[chrome]

Do a google for "Honda Accord TSB 03-079" and you will find a TSB for long cranking and hard starting.

 

[cchase]

I'm wondering about the TSB for hard starting and long cranking.

However, the problem is intermittent and usually is at its worst when cold. I would think that any FPR would bleed back down to nothing in that amount of time, defective or operating correctly. And would not be temperature dependent.

Is this thought process wrong?