Cranks but no start

[Zeke]

So, my wife pulls into the gas station and fills up. Gets back in the car and it cranks but won't fire. I go there and I check some fuses and find all is well. I call the flatbed and send the car to Wayne's favorite place, Callas Rennsport.

It runs right off the truck. But, after fiddling with it for a day, they decide that it runs alternately between way rich and way lean. I get a call asking if the car was super low on fuel or overfilled.

Overfilled? I don't fill past the first click and I don't think my wife does either. However, not all nozzles work perfectly. But the car dies because of being out of gas or overfilled?

CS says this is not common but does happen. The charcoal canister gets wet and game over. They're still working to find the problem. No codes at all.

I'll keep you updated.

[stew]

Might be fuel pump . Had A similar problem

[BYprodriver]

Fuel pump is cooled by fuel in the tank. If the tank is real low on fuel the pump has to work harder to suck in enough fuel to supply engine which makes it hotter & less fuel to cool it.

[Wayne 962]

I agree. The closed-vent system on the Boxster can cause the fuel pump to maybe stop working when the conditions are where there is not too much fuel (pulling a lot of vacuum), or too much sometimes. It does indeed sound like a fuel pump problem - it will probably happen again sometime in the future. What happens is that if the tank gets down really low, the stuff floating in the tank can get sucked in and clog the pump. When you let the car sit and then had it towed, it might have settled, and then was able to start up again. The way to test for this is to check fuel pressure in the engine compartment.

-Wayne

[Zeke]

Apparently Callas was not able to work on the car much on Thurs. So, when Tony Callas returned from training at Bosch, he took a look. He called me late Fri to say that the car is "learning" to run again. No codes.

He suspects a sensor or something in the intake that was telling the computer that the car was flooded. The charcoal canister has been mentioned as a possibility if it somehow sipped some raw fuel.

Some abberation triggered the problem and it's not all that objective. The car stayed the weekend for more evaluation come Monday.

[Wayne 962]

I'll put my money on the fuel pump. Just about anything else causes rough running or poor performance, but a fuel pump failure or intermittent failure causes a breakdown at the side of the road. The coincidence of filling up the tank at the same time is too big of a clue.

-Wayne

[Wayne 962]

Small world, same problem? obd2 dtc code

-Wayne

[Zeke]

Now I'm told there ARE codes, 4 of them. One is one I reset thinking that if it's chronic, it will come back. Never did (O2 pre heater). Another is the CEL light doesn't work. A third points to a problem with the charcoal canister and that vapor system. And the last is idle mixture.

Now we realize that the O2 sensor was and probably is bad all along. No light, no concern, although the car ran fine.

Sorry, I didn't write them down when the shop called. Weired since at first they told me that there were no codes. Or I misunderstood what I heard. But I remember the statement that it's hard to chase down a problem when there is no information pointing to it.

I'd agree with that.

At least there is information now.

[Wayne 962]

Right. I would find it difficult to believe that the car would not start and there would be no trouble codes registered. That's almost automatic that a code would be generated.

-Wayne

[Zeke]

Well, what we have now is a broken oil fill tube. This was found using smoke introduced to the motor's internals.

[BYprodriver]

Originally Posted by milt

Well, what we have now is a broken oil fill tube. This was found using smoke introduced to the motor's internals.

I think that is standard on the 986.

[Wayne 962]

Originally Posted by milt

Well, what we have now is a broken oil fill tube. This was found using smoke introduced to the motor's internals.

Yup, I just got off the phone with Tony (called him about something else), and I asked him about Milt's car. The "smoke test" that they talk about is documented here in my tech article. I actually borrowed Tony's smoke machine for the article (then I bought my own - they are too cool):

http://www.pelicanparts.com/techarticles/Boxster_Tech/21-FUEL-Vacuum_Leaks/21-FUEL-Vacuum_Leaks.htm

FYI, the first thing Tony said he did was to check the fuel pressure from the pump - looked okay to him.

-Wayne

[Zeke]

Here is a full update as copied from an email from Tony Callas to document the various issues:

"When your car arrived it started immediately once it was unloaded from the tow-truck. We ran a complete control unit interrogation and located various DME OBD-II fault codes including:

P1128 & P1130 Oxygen sensing adaptation RKAT (Both banks) below limit.

P1115 Oxygen sensor heater pre-catalyst bank #1.

P0650 MIL (CEL).

P0446 Shut off valve, activated charcoal filter (function).

--------------------

Expanded descriptions from Tony:

P1128 and P1130: This is due to a fuel trim variance with the RKAT range. The standard tests here are to check the crankcase vacuum with a Manometer, yours was low hence an obvious air leak, then to check the fuel pressures, yours were great and stable. Then to smoke test the intake and engine crankcase systems and look for a vacuum/false air leak. We located a torn engine oil filler accordion tube, replaced the faulty tube but the system was still stuck at the RKAT rich limit of 4.50. There is still a large air leak apparently involving the ORVR and EVAP systems. I now know this because when I activate the ORVR (On Board Vapor Recovery) system, the car goes straight rich, I mean hunting, smelly rich, hence the back and forth rich to lean readings we got when your car first arrived. Now, the no start after refueling condition is starting to make some sense. Tomorrow we will check the EVAP and ORVR systems operation including integrity and let you know how the Shut off- AKA the close valve is. We are also hearing a hum from the fuel lid vent area which usually is the result of some sort of blockage in relation to the EVAP system. It’s always interesting here at CR. More information to come.

P1115: This is usually a faulty Oxygen sensor heater which is internal to the oxygen sensor but it can also be a faulty connection. The HO2S is controlled and grounded by the DME. I am not focusing on this at this time. At least not the first item of concern but not to be ignored either because this can affect the fuel trim values but only one bank.

P0650: This is because your MIL (Malfunction Indicator Light AKA CEL or Check Engine Light) is inoperative and is most likely a faulty bulb. You can do this yourself if you wish. I can tell by our scan tool values that your car has a problem because the Mode $06 values (Thresholds) are being exceeded. This is where the MIL light illuminates.

P0446: This valve works in conjunction with the EVAP Purge valve to help check the integrity of the EVAP system and is located in the RF wheel house area. This is an indicator that the EVAP shut off valve had intermittently stuck in its operation at some point. This is not terribly uncommon and is usually an intermittent issue. Normally, this is more of a problem IF it is accompanied by an EVAP leakage fault but you have no EVAP system leak faults. This seems to be playing a key part in why your car wouldn’t start immediately after a petrol refill but there is a missing part we are looking for which will put the puzzle together. More information to come."


You bet this was not something I could have tackled at home. So, the question is to you folks following the saga, do you think I can change out the CEL bulb myself? I don't yet have Wayne's book on the Boxster but I do have a Bentley that has never been opened to a page. So I don't know how the instrument cluster comes out.

If it involves the air bag, I'll probably pass even though I have been around cars and worked on them for many, many years.

[Wayne 962]

Saw your car over there today. Yes, changing out the check engine lamp is a relatively easy task - see the article here: Pelican Technical Article: Boxster 996 Gauge Cluster Upgrade - 986 Boxster (1997-04) -

-Wayne

[Zeke]

My head hurts.

Callas is now changing the O2 sensor that threw a code last year. How do you change an O2 sesnor? Well, let me tell you how it's done at Callas.

Change out the old for the new and wire up. Remove the battery cables from both terminals for a battery reset. Touch the cables together using a resistor in between so that you don't get an instant drain of residual energy stored in the DME, etc. Wait. Reinstall the cables in a way that voltage is not spiked (I got an explanation up to but not including this step). Reset the radio code and test all systems. Test drive car and recheck parameters.

Yes, this is not cheap. But if there is one thing I have learned during the time that my car has been in the shop (since late last week), it is that any work I might have done before or attempted since the car died may very well have ended up costing me more in the long run.

My advice is do not read forums and then think you can DIY these cars unless you have more than some background with DME 7.* systems. The Bentley I have says to remove the O2 sensor with an O2 socket and replace. Not a thing more.

So where does that leave you? By the side of the road again?

One more piece of advice: don't neglect these cars. I have no idea when my CEL light became inoperative. It's not something I look for every time I insert the ignition key. But it appears that the car has been driven for some miles with problems even though it seemed to run great.

Finally at a gas station fill up everything caught up and stopped the engine from running. If I find out anything more about why, I'll post.

All in all, given some of the strange things these cars can do, I think this incident has to be one of the more complicated ones yet talked about here.

I'd much rather deal with a convertible top cable. And I may have a Durametric used one time for sale.

[Zeke]

Well, it's done. Took a week. The final analysis was that gas got into the EVAP system because a hose was off in the front wheel house. All of a sudden I realized that the car had gotten its first fill up since it came from the body shop where they painted the front bumper off the car.

You can guess who knocked the hose off and cost me a few hundred bucks.

There were other issues that combined to shut the car down. Cracked oil fill tube let air into the crankcase altering the vacuum. The hose off the front let air into the intake causing the DME to read the air/fuel ratio wrong.

It all came together in a single moment at the gas station.

[Wayne 962]

I don't think that all that is absolutely necessary simply to change out an O2 sensor. I think the Callas guys took extra precaution to reset the system on your car because it was dishing out so many codes.

It is my opinion that these car are far easier to work on than say the 1984-89 911s, which don't output any fuel injection codes, and simply leave you guessing as to what the problem is. At least with the modern cars, you have computers that will give you clues as to what to replace.

-Wayne

[Zeke]

Originally Posted by Wayne at Pelican Parts

I don't think that all that is absolutely necessary simply to change out an O2 sensor. I think the Callas guys took extra precaution to reset the system on your car because it was dishing out so many codes.

It is my opinion that these car are far easier to work on than say the 1984-89 911s, which don't output any fuel injection codes, and simply leave you guessing as to what the problem is. At least with the modern cars, you have computers that will give you clues as to what to replace.

-Wayne

And as I was leaving and paying my bill, Tony told me that when he changes an MAF he does the same reset. In fact, the O2 reset was the 2nd reset in the same week of working on the car. He showed me the print out of the fuel map and what the 1st reset did after the purge valve was changed. What that did for him was to prove that the DME responded to the repair. It showed a perfect read. He told me to not be surprised if the fuel mileage went up and asked me to log it.

I get it that if any sensor is changed, reset. Maybe the bleed off of the residual energy is so that one doesn't have to wait so long.

I'll go back to what I said earlier in this thread (paraphrased). It's not what you know, it's what you don't know. I don't know enough to work on these cars. Well, the electronics AFA engine management anyway.

[Brillo]

When I had my SC in Tony's shop, I found that he is extremely thorough and takes no short cuts. I like that.

I haven't had my 986 there yet, but it soon will be going in for the first time in check over.

[Zeke]

It wasn't until I picked up the car that Tony, my wife and I figured out what happened to cause the issues. Late in Dec an in-duh-vidual backed into the car and we had the bumper repainted. The bodyshop did this off the car. When we picked up the car it had a noise that sounded like a stuck lifter, but I couldn't tell where it was coming from.

No need to complain to the bodyshop, so we drove it home. The noise seemed to go away. I'm going to go out on a limb here and say that the computer was trying to learn some new parameters to adjust. Meanwhile, the CEL bulb was out so codes were not indicated.

My wife managed to get a couple of weeks out of the gas tank and the fill up was the first since retrieving the car from the shop.

Now, as I understand it, the valve was not working properly because a hose and been pulled off behind the front wheelhouse inner liner (it looks like a bodyshop error). So, I guess it got raw gas from the filler neck right down where it connects (wherever that is).

That shut the car down, but by the time it reached Callas, it would start. It did not run well though. So, the saga began. In addition, the oil fill tube was also breached. It was a matter of coincidence and circumstance that may never be repeated.

I think Tony reset the computer to shorten the diagnostic time so that he did not have to drive the car the miles it takes for the thing to learn with parts replaced and hoses connected.

Now it's certain that I don't understand every little nuance about this episode and don't have all the terms exactly right. But, it seems highly circumstantial and I have to think it all started when the bumper was removed/replaced.

Something to think about. The car has run perfectly since.