Hi, all

I'm trying to track down a problem I have every spring. All winter long the car will start amazingly quickly, but once the overnight weather gets above around 50F, the first start of the day can take a good ten seconds of cranking. After that the car runs fine and will restart fine.

I finally got around to checking the fuel pressure since I've read about fuel pump going bad on this forum, but I don't think that's the problem after all. This is a 97' with the old style return line.

• Turn key on with engine off, pressure immediately goes up to ~42 psi and stays there during engine start
• After 15 minutes it only drops to 39 psi
• Liquid gas spurts out of the valve even after sitting overnight, so doesn't seem to be vapor

I pulled the vacuum line off the fuel regulator, and I didn't spell any overwhelming smell of gas. Maybe just a whiff.

Scanned the input air temperature and engine coolant temperature, and they were sane and the same.

MAF reading looks reasonable and responsive while idle.

The only odd thing I see in my scanner is both LT fuel trims sitting at -5 with engine off and both LT and ST fuel trims going down around -10 with the engine running at idle.

Anyone have any thoughts on what to look at next? Do negative trims tend to impact starting, and what about warm weather might make a car harder to start cold but not harder to start warm?

 

Coolant temp and intake air temp sensors would have been my first guess. Since those are reading normal (ambient) from what you said, and your rail pressure is good, that makes it a tricky read.

Mass air flow not reading correctly from the hotwire? You might try a simple MAF clean for starters.

 

I went ahead and cleaned the MAF sensor, and then I recorded specific numbers. Glancing around the forums, these numbers look too high:

• Idle, 660 rpm gives 5.9 g/s
• 1200 rpm gives 8.3 g/s

A post such as this one shows idle giving 3.7 g/s.

I've had negative fuel trims on this car for a while, so maybe the MAF has been giving wrong readings for a while.

I also noticed on one of the O2 sensors not being as active as the other, so I suspect that as well, but the forums say the computer monitors those for malfunction and would have thrown a code. I have only the normal P code, not even anything pending.

I figure I'll replace the MAF.

 

Volkris,

As you seem to have most bases covered, I had another thought that it could be the EVAP purge valve.




It's purpose is to manage the fumes from the fuel tank, and when faulty is known to cause hard starts. When it's warm, you'll have more fuel vapour in the tank. Our fellow Texan Forum member had a pain of a job because of it. Link to his thread below. Rev Sam had a good YouTube video on it too.
https://www.jaguarforums.com/forum/x...highlight=wife

Just another thought, as I had this problem on other cars, try opening the fuel filler cap and stick a screwdriver etc down to open the inner flap, and let the vapour out for a few minutes. No smoking nearby. If car then starts - then there's the problem.

 

DavidYau,

Awesome, thanks! It'll be something to try before I have to do more research about this surprisingly expensive MAF sensor.

 

Update: no solution yet.
I opened the gas tank with a big gush of vapor, and let the tank sit open for a few minutes. It was still a hard start.

A new observation is that unplugging the MAF sensor made no change in starting, not that it necessarily would have.

My plan is to order a new MAF sensor based on the elevated idle flow readings compared to others I see in these forums, and keep on opening the tank in the mornings. Maybe the car will learn new trim levels without a faulty purge valve putting vapor in the manifold or with a more accurate MAF reading.

I supposed I could test this with a hard reset, but man I hate reprogramming my radio.

 

Update: I think DavidYau's suggestion was actually correct!

For my first test, I just went out when it was already 80 degrees and opened the fuel cap. There was a gush of vapor, but the car still didn't start easily.

BUT on the second test I left the gas cap off overnight, and the car started right up. Third test I left the gas cap on but just loose, and it still started right up.

So I'm thinking vapor was leaking through a valve over the course of hours, overnight, and building up somewhere. Just opening the cap didn't release that buildup, as if there might be a checkvalve involved.

I still need to understand the EVAP system better. My early XK8 might have the key valve located in the engine bay instead of in the fender as in Reverend Sam's video, or maybe they are two different valves.

Anyway, just wanted to write this down for future readers before I forget and never get around to it. The leaving the cap off overnight turned out to make a critical difference.

 

Anyone know if the purge valve does double duty as a pressure relief valve?

I bought a replacement purge valve off EBay that was listed as new but seems obviously used. I installed it, but the car had the same hard start.
Now I'm faced with the question of whether the used replacement is also bad.

The 2nd edition of the workshop manual says to check the valve for internal leaks by drawing a vacuum on the output line. The old valve definitively fails this test, and the replacement seems to pass. However, on both valves I can blow air throughfrom input to output, and I don't know if that's also a sign of a broken valve.

Is this how the valve is supposed to work? If there's positive pressure behind the valve, should it be allowing vapor to go into the engine even when closed?

I'm thinking the canister vent is supposed to make sure it's at atmosphere, so there shouldn't be positive pressure behind the purge valve. This could be a safeguard in case the canister or its vent is plugged. Or maybe the valve is just broken.

My hard starts seem related to pressure in the gas tank, so now my theory is that cranking the engine opens the tank pressure valve which sends a slug of vapor to both the purge valve and the canister. It's supposed to exit through the canister, but instead these purge valves are letting it flow into the engine. I just don't know if that's because of bad purge valves or bad canister venting.

Any thoughts on this are appreciated!

 

Well,

Did you check the solenoid worked (clicked open & closed under 12V) when you blew through it?

If you’re confident the replacement purge valve is good, then it’s must be a bad hose somewhere. Did you ever see a P04xx DTC?

The purpose of the EVAP system is to use the fuel vapour from the tank and route it to burn.

However there is another scenario - if there’s a leaking hose between the valve to the intake, it’ll introduce a gross air leak which is after the MAF. Therefore, it’s not too much fuel, but too lean to start. Counter intuitive I know.

When you had an easy start was the fuel cap open or closed?

You’re almost there. Keep at it.

 

No, I'm not confident that the replacement is good. That's what I'm asking in my last comment

The replacement purge valve blocks the engine from sucking, but it doesn't stop the tank pressure from blowing. I don't know if that's the expected function or not, but I could imagine reasons for it being intentional.

I have no codes, but I do have negative fuel trims, which makes me think it's not an air leak. The easy/hard start doesn't change with the gas cap on or off during starting, but it does change with whether the cap was on while the car was sitting. In other words, what matters isn't whether the cap is on or off now, but whether the tank had been allowed to build up pressure, even if the pressure has been released already.

And now I'm going to take a close look at the canister atmospheric vent.

 

Another update, another bad part!

Finally following Technical Bulletin 310-02 on hard starting, I followed the directions for testing the Tank Pressure Control Valve, (TPCV), aka Rochester Valve. As soon as I pulled the vacuum line off I could smell the gas vapors. It couldn't hold any vacuum at all. I believe this vacuum line goes directly to the intake manifold, so it provides a direct path for gas vapors to get to the engine unmetered.

 

So far this looks solved, and the car has been starting easily for a few days.

Following this technical bulletin I ended up testing the tank pressure control valve (TPCV) and found it open from the vacuum line directly to the tank side.
Both my purge valve and my tank pressure control valve seemed broken. Replacing the purge valve didn't fix it, but replacing the TPCV as well did.

It's a really strange effect that releasing pressure on the tank before starting didn't help anything, so I'm thinking gas vapors leaking through these valves were building up somewhere, but how or where I can only speculate.


TSB 310-02

 

Volkris,

Thanks for following up on your thread and the TPCV solution.

I suppose with the age of our cars, we are going to see more examples of these valves failing.

Decades ago, As a youngster, I had an old Honda scooter and the rubber valves/diaphragms seemed to go one after the other, until my Dad and I ended up replacing all of them. Loved that little bike.

Enjoy the happy motoring.... until something else goes wrong! LOL!