Hard warm start and rich at idle - 3.2

[Hunt3R]

UPDATE: 6/1/2013
I tested for spark when cranking during a hard warm start by pulling a plug wire and putting an extra spark plug in it and grounding the base of the plug to the engine with a copper wire. First I drove the car then parked and came back around 45 min's later and the fuel pressure was still at about 31psi. While I had someone else crank the car, I saw a bright blue spark on the plug that appeared to pulse every second, then after about 5 seconds the engine started and the spark was constant. I shut the engine off, waited a little while to trigger the hard start issue again, and then repeated the test 2 more times with the same result each time. Is that spark pulse while cranking expected? If so, I assume this means the hard warm start issue is fuel related...

UPDATE: 5/5/2013
Hard warm starting still a problem...

I removed my charcoal canister last week but have no replacement so I left the hose coming from the fuel expansion tank just hanging in the wheel well to vent to the air and I plugged up the airbox attachment where the hose would've come back from the charcoal canister.

The first warm-ish start that I attempted after removing the canister the car started right up as soon as I turned the key! I thought this must've been the problem all along and I've finally fixed it. But then the next day it happened again and after driving more this past week it's still happening 100% on warm starts. I have no idea why it was "fixed" on that first warm start when this hard warm start problem has always been 100% reproducible, but again everytime since then I'm cranking for 5 sec's before it starts, same as before.

It seems like the charcoal canister may have had something to do with it. Perhaps the line from the expansion tank is clogged, the carbon pellets were spilling into that line too. Or perhaps the expansion tank itself is faulty? It really seems like my problem is vapor lock in the lines, although I hold fuel pressure as expected after shutting off the engine.

Has no one else with a 3.2 ever experienced warm start issues like this? No one has any experience with the evap emission system?

A/F mixture definitely seems spot on now though, once the car is started it drives perfect, idles excellent, pulls strong, and no more black smoke or rich smelling exhaust.

UPDATE:_4/28/2013
I think I may have fixed my A/F mixture. It had been a while since I tried using my DVM on the O2 sensor so b/c I recently fixed more vacuum leaks I decided to try again and this time I actually got a good reading. I have the mixture screw approximately 12 full turns out, not coincidentally this is where it was set before all of this started, and my DVM reading at the O2 is now fluctuating from 0.3-0.8.

I also may have figured out the hard warm starting. The vacuum hose that runs from the charcoal canister in the rear-passenger wheel well attaches to the back of the airbox and I had recently discovered it was disconnected. I double-checked it today and discovered it's so old that due to stretching it to reach the airbox it has cracked, so I decided to pull it and replace it.

Well when I disconnected the hose from the charcoal canister a bunch of charcoal pellets came pouring out of the hose like it was clogged with them and I can see them at the edge of the canister ready to pour out. I assume this is not good and I need to buy a new charcoal canister in addition to new vacuum hoses. Bentley says "a faulty evaporative emission control system can cause hard warm starting" so I'm crossing my fingers this is it. Anyone have any experience with this?


Original Post:_4/21/2013__________________________________

The Subject:
1986 911 Carrera 3.2
100k mi. I've driven her for 4k of those since purchased 10 months ago. Car was stock and very clean and had very few miles put on it over the past decade. It has generally always run great with a couple minor issues I'm struggling to fix...

Warm Starts:
Since I've owned the car, after I have driven and then stopped for longer than 10 minutes, it takes 5-10 sec's of cranking before the engine starts. However after she's sat overnight or for a day or more then the engine will start right up as soon as I turn the key, and if I just make a quick stop, say less than 10 minutes, then she will start right back up. This hard warm "ish" start seemed rare at first but eventually progressed to 100% of the time.

Base Air Fuel Mixture:
My idle mixture seems off, I think it's always been too rich and after I made a few modifications and fixed some vacuum leaks it only got worse. I took the car to Porsche of Orlando but they no longer have the tools to measure the exhaust gas. The guys at Stratton Motorsports told me they're getting a gas analyzer soon and I can use it, but I don't have one so I've been making base A/F mixture adjustments sorta blindly and never getting it perfect, close, but not perfect.

Here's a video of my warm "ish" start issue. The car has been sitting for 2-3 hours here so it cooled down but takes like 5 seconds to start and bounces a couple times, I think the idle bounce is strictly due to the hard cranking, too much fuel in the lines perhaps from cranking so long? Because that only happens on these hard warm starts.
Video of warm_hard_to_startup | Flickr - Photo Sharing!

[Hunt3R]

Same video but different link b/c the Flickr hosting seems to be very slow:
Cranks good but won't start for 5-10 sec's - YouTube


History:

• On a cold start, after a few minutes it would develop a slightly choppy idle, sometimes fluctuating between 750-800rpm, until fully warmed up where it would hold a steady idle around 950rpm.
• Replaced catalytic conv. with Dansk euro premuffler. Could actually smell the exhaust now and it smelled very rich at idle and started putting some white smoke out on cold starts, clearing up quickly, but previously I had never seen any smoke out the tailpipe.
• Installed K&N filter, nothing changed with issues but it started whistling thru the intake on moderate acceleration, so I cut off the airbox lid which got rid of the whistling and now I can hear the engine scream at full throttle.
• Installed a Steve Wong chip for 93oct and premuffler/sport exhaust. Idle rose too high with new chip to 1300, adjusted it down to 880. Same issues still exist but now right when starting up the car the idle bounces up and down maybe 5 times but then settles down and holds steady for the rest of the drive. Gas mileage and throttle response improved a lot.
• Tried to get reading from O2 sensor with DVM but when I first tried it showed a rich reading, bouncing from 0.8-1.2V, so I adjusted the base mixture screw out 1 full turn and it just fell out of the AFM. So I screwed it back in all the way and backed it out 2 turns (supposedly the factory position) but this made it run very rich and burned my eyes and nose with some black smoke out the tailpipes but my DVM reading said lean @ 0.0-0.2V. So I changed the O2 sensor, but that did not help. I started making base idle mixture adjustments and resetting the base idle speed too and then I'll drive for a few days to test it out. As I adjust the mixture screw out, it's obviously leaning the mixture b/c the car chokes less, holds the idle steadier, drivability from a stop improves, the black smoke out the tailpipes goes away, and the warm idle hunting at startup will slowly improve and bounce less as I lean it out further, but the hard starting is never affected by these adjustments and I've been doing this for months now. My DVM reading from the O2 always says lean even if I screw in the mixture screw all the way in where the engine won't even start up it's so rich. And if spray carb cleaner into the intake the idle rises but my reading doesn't change, still says lean, maybe my DVM isn't sensitive enough.
• Installed a B&B muffler, no change with any issues but man does the car sound awesome now.
• Replaced the CHT sensor because my DVM said the resistance was too low, barely registered on cold engine. Thought this might be causing a slightly rich mixture but nothing changed with the new sensor and I get the same reading with my DVM on the new sensor. Like I said, I think my DVM sucks.
• Replaced fuel filter and fuel pressure went up a couple psi from 37psi to 39psi running. I learned that this is too high and I discovered that the vacuum hose that runs from the top of the fuel pressure regulator to the top of the fuel pressure damper has a T-connector where they meet behind, and plug into the throttle body and it was disconnected. After I rerouted it because it was stretched too much to stay on the nipple, my fuel pressure then dropped to about 30psi running, Bentley says it should be 29psi. When I stop the engine the pressure rises to 32psi. Bentley says it should hit 36psi but it does not. Bentley also says it should hold above 14.5psi for at least 20 minutes after shutting off. It does, and it takes nearly an hour to drop that low.
• Checked all vacuum hoses and retightened some clamps and found 2 more disconnected hoses, the one from the evap system that plugs into the back of the airbox, and the short hose that connects like a U to 2 nipples on the back of the throttle body right above that T-connector from the FP reg/damper. Fixing these vacuum leaks has only made the base idle mixture richer it seems and I've had to unscrew the mixture screw further and further.
• Tried disconnecting injectors one at a time while running, the idle sounds slightly rougher and black smoke starts coming out of the tailpipes and the fuel pressure rises a couple psi. Plugging them back in doesn't seem to help much, the pressure goes back to normal but I still had black smoke coming out until I turned off the engine and restarted. It was hard to ascertain any differences when unplugging each individual injector as the first one I tried unplugging messed it up and it didn't really recover.
• Swapped DME fuel pump relay a couple times but nothing changes.
• Tested the functionality of the idle position switch, it checks out.
• Tested and cleaned the ICV, it seems to be working fine.

Where I Am Now (updated):
Currently I have the mixture screw approx. 12 full turns out and after fixing vacuum line issues my mixture seems to be perfect. The car holds the idle steady and smooth and runs great and pulls strong and is very drivable and has no smoke coming out the tailpipes and no more rich smell.

The warm hard starting issue still exists. Any suggestions where to go from here? Can bad injectors cause this? They're one of the only things I have not thoroughly tested. My records show at least some injectors were checked, cleaned, and/or replaced about 15k miles ago, but that was also about 15 years ago. It seems like a vapor lock problem but I hold fuel pressure as expected after shut down. Removing the defective charcoal canister seemed to have fixed the warm start issue at first but then it started happening again 100%. I removed the canister and have the hose from the fuel expansion tank venting to the air in the wheel well but the carbon pellets were spilling out into all 3 hoses so maybe they somehow got back up in the line from the fuel expansion tank and it's clogged?...

[DRACO A5OG]

if I may suggest, assuming your o2s is operating properly, it will only give you a rough reading and thecar must be hot when you attempt to get a reading on the dvm.

to be absolutely accurate, you need a LM1 or LM2, try to borrow one.

in your video, I see sub dips and hunting. it also should have dropped to 880RPMs after 45 seconds. in my experience the culprit lies in the improper AFM adjustment then you must set base idle then check AFM again. recheck base idle.

Jim

[Hunt3R]

Car has been fully warmed up to +190F every time I've tried to read the O2 with my DVM and make base mixture and idle adj's. As I said in my post above I'm planning to use a gas analyzer when my friends at Stratton Motorsports get one sometime in the near future, but until that happens I have been unable to find one. Even Porsche of Orlando did not have one.

The idle would've soon dropped and held at 880rpm but I didn't record video long enough. I have become a self-proclaimed expert at setting the base idle speed, I've literally done it hundreds of times now. The idle will get no better at the current mixture setting, but the sub dip may get better if I richen it, however the idle bouncing on startup will get worse if I do that. This video was actually a good example of about the best it gets, it only bounced twice instead of 5-10 times and it started after only 4 seconds of cranking when it usually takes at least a couple sec's longer.

I think that there must be some part failing that I've not discovered. I'm suspect of the fuel pressure regulator since my running pressure is 1psi too high and my pressure at shutdown is 4psi too low.

[JAR0023]

Did you check the fuel quality switch on the back of the DME? I'd be looking to swap my AFM with a known good one if it was me. -J

[Hunt3R]

Yes the FQS is set to position 0, the stock setting.

I would love to swap my AFM out if I knew someone who has one. However I've tested the AFM per Bentley, the plug from the DME has just over 4V on pin 3, Bentley says it should be approx. 5V. I noticed it also has the same voltage on pin 1.

And on the AFM side the connector pins 2-3 show some resistance and as I push the vane inside with a pencil the resistance increases. The vane moves freely yet has good spring tension on it. It appears clean inside too.

Testing the intake air temp sensor on pins 1-4 I get resistance, but not what Bentley says I should, but then my DVM never does, like I said above I changed out the CHT because my DVM showed low resistance and then I got the same reading on the new CHT, waste of money. I should prolly buy a better quality DVM, I bought the cheapest one they sold at Autozone. But I am getting consistent readings at least, so I'm not so sure the AFM has an issue. I did find an old post though where a guy with a 3.2 had a similar rich mixture issue and he said replacing the AFM fixed it...

[Chuck.H]

First off, get a DVM you trust -- I seem to recall seeing Flukes at Sears, I highly recommend them.

Second, I wouldn't be too concerned about fuel pressures that are off by a few PSI - it's not that critical IMHO...

Third, since you mentioned vacuum hose problems, I would check everywhere for vacuum leaks - that power brake booster atrocity on the left intake runner is a good place to start.

Fourth (I really didn't mean to number these) don't get hung up on the O2 sensor, it isn't used for starting the engine. Most important sensors for this are head and air temp.

If you want to make the drive up to Gainesville, I have a spare AFM and DME you're welcome to try swapping in.

Chuck.H
'89 TurboLookTarga, 350k miles

[Hunt3R]

Thanks I may have to take you up on that offer, its about a 2hr drive for me but I go that way on occasion. I have checked all vacuum hoses very thoroughly though, that's when I found the others disconnected. They all appear to be snug and in good shape now, the brake booster hoses too. One thing I noticed after fixing the vacuum leaks was my oil pressure gauge had always been kinda twitchy and now it's much more steady.

[Hunt3R]

Updated up top with latest findings...

[Hunt3R]

Updated at the top with latest findings...

I really think my hard warm starting is due to a fault in the evaporative emissions system but Bentley is very thin in describing this system with no information at all on how to test or fix problems.

[Hunt3R]

Updated at the top with latest findings...

I appear to be getting a good spark while cranking so I assume that means the hard warm start must be fuel related.

[Hunt3R]

edit... not FIXED!!

Unfortunately I really don't know what fixed it. I was finding that if the gas tank was near empty, running with the gas light on, then the problem improved. Sometimes when I turned the key it would try to start but choke a little and crank maybe for only 2 seconds and then start up. As soon as I would put more gas in it the hard warm start would become 100% again.

During an undercarriage inspection at a local Porsche shop, Zotz, they found that my belly pan was rusting as if the battery above had spilled acid but there was no sign inside the trunk compartment of that ever happening. Perhaps it had been cleaned up well inside but the damage soaked in deeper than the previous owner realized.

So anyway I had Stratton Motorsports replace the belly pan which requires taking apart the entire front end and removing the gas tank, etc to cut out the belly pan and weld in a new one. During this process they did something that fixed the hard warm start issue. They said there was some crud in the line before the fuel pump but otherwise didn't notice anything. Perhaps that crap was slowing down the fuel flow to the pump during warm starts?

[Hunt3R]

Of course as soon as I declare it fixed it comes back again. All week I had been driving everyday and after letting the car sit for different amounts of time and it started right up each time. I must've started it 20+ times over the past week.

Well last night it started cranking 5-6 seconds again before starting. So after taking the gas tank out either something took nearly a week of driving to get back into this hard warm starting state again AND/OR it really does have to do with the fuel level in the tank because I had been driving on a nearly empty gas tank all week when it was starting up perfectly, then Thursday evening I filled up with 1/2 a tank. After this I started it back up warm 3 more times without issue except for the 3rd time it choked a little and took a couple seconds, and actually the 2nd time I had stopped for less than 15 mins, and now everytime since I'm right back where I was again, cranking for 5-6 seconds before finally starting up. I'm betting it will again improve once I run the tank down towards E...

[Hunt3R]

Never did update my thread here that I bought an LM1 A/F meter and have my base A/F ratio set right where Steve Wong told me to set it for the chip, fluctuating from 14.2-14.4, and I've had it set this way for a while now. Of course I reset the base idle to 880 afterwards too.

I also did a valve adjustment, spark plugs, rotor, etc late last year, but it didn't affect the hard starting issue. After this however, is when I started to notice that if the gas tank is near empty then it would sometimes miraculously start up right away.

Yesterday I decided to investigate the fuel evaporation system since I've been suspecting it. I found that the hose to my fuel expansion tank is clear and I can pump air thru it as well as thru the expansion tank, however the hose from the top of the tank that runs back into the car is missing so my expansion tank just vents in the front fender. I tried pumping air from the hose in back from the carbon canister (which I removed) but it seems this hose is clogged so perhaps the previous owner opened the top of the expansion tank on purpose. Regardless, it does seem like the fuel evaporation system is at least working as it is and couldn't cause the hard starting so I'm ruling it out.

Today I decided to investigate whether the problem could be due to vapor lock so once I got the car warmed up I parked for 1hr. Then I pulled my DME relay and in the socket it plugs into I jumpered pins 87b and 30 to run the fuel pump for about 30-40 seconds. I could of course hear the pump running and my pressure was about 32psi. This is also the pressure it was already sitting at before I started running it as it had been holding at 32psi since I turned the engine off 1hr ago. I then plugged the DME relay back in and cranked the car and it still took 5-6 seconds of cranking before it started.

So being that I hold fuel pressure for a long time after shutting down, and even running the fuel pump to clear out the lines before starting didn't help, then I guess this rules out vapor lock due to whatever multiple possible causes like leaking injectors.

I think I'm going to start focusing on the fuel pump now... however I don't know what I expect to find wrong with it as I do get fuel pressure immediately upon cranking the car. This is what my fuel pressure does while cranking during a hard warm start: 1986 Porsche 911 3.2 fuel pressure during hard warm start - YouTube