On cold start, very rough/dying idle after fuel pump install

[Woodsy0]

As stated, I just installed a VR-4 pump. Because of a fuel fitting mishap and an extremely rusted-over car, this was accompanied by a brand new fuel pump assembly. I transferred over the o-ring and the other two pieces that came from the top of the stock pump as per VFAQs instructions, and torqued everything down. I used -6 AN lines as well (I know they're not needed, the fuel line needed repair so I figured it'd be better than buying OEM) and torqued them down to specs. Ran the fuel pump relay to check for leaks both at the pump and at the rail, no leaks whatsoever.

However, on cold starts, the engine idles like crap, and if I'm not careful to blip the throttle at idle or when I disengage the clutch, the car dies. However, the problem totally disappears after ~5 minutes of operation...

Any ideas? I know on the VFAQ, it states that if the o-ring is kinked or torn (I did forget to lubricate the ring prior to installation) that "the car will be hard to start after sitting a few minutes, as the kinked/torn O-ring will allow the fuel to bleed out of the feed lines. The symptom will be that the car will take several seconds to start after sitting more than a minute or 2, but starts right up if you turn it off and then immediately start it back up." However, while this is similar to my problem, it's not totally the case, since the car can sit for 15-30 minutes after the initial heat-up and be fine to restart - it's only on cold starts (after the car has been sitting for several hours) that this happens...

Thanks,
Scott

 

[toybreaker]

It's possible your fuel system repairs may have some bearing on this problem. I like your a/n fuel line solution to the corroded fuel line issue these cars can get in the rust belt. . Did you blow thru the hose to make sure it wasn't restricted? Another possibility is some trash got into the fuel filter. These wouldn't ordinarily cause a cold only problem, but it is a dsm p ), so anythings possible...

Tearing the o-ring on the pump is a much more likely suspect. You can test for this with a fuel pressure test. If you can source a fuel pressure gauge with the correct style fittings, you can test fuel pressure at the filter. Check the running pressure and compare it to specs, but mainly what you want to do is leave the gauge on overnite and check the readings cold. If it drops off excessively overnite, the o-ringmay be torn/leaking. The check valve
may also be faulty.

If the fuel pressure is high, disconnect the vacum reference hose to the fuel pressure regulator (put your finger over the refence hose/plug it with a golf tee to prevent a vacum leak) and watch the fuel pressure. With the line off, it should increase, line back on it should decrease. If it doesn't change, the fuel pressure regulator is toast. There may also be fuel in the reference line from a torn pressure regulator diaphragm.

It's possible to ruin the diaphragm with aggressive chemicals sprayed on the fittings at the rear of the car, if they get inside the system, and get a chance to sit in the fuel rail.

When the engine is warm, and in closed loop, it *may* be able to compensate for high/low fuel pressure enough that you only notice the problem cold, and in open loop.

Since this only happens from cold, I would check some other things as well

If you have a datalogger, I would suggest you look at your engine coolant temp. sender.

If you don't have a datalogger, almost all shops have the mt2500 snap-on scanner that will also read the datastream. Using a shop scanner also gives you access to somebody who knows how to intepert the engine sensor readings. Hopefully you know somebody who will hook you up,...beer is very good form of legal tender for this kind of stuff.....

If you do manage to finagle the use of shop equipment, try and get the car there early enough that it can cool all the way down. You're looking for an anomoly that only occurs cold.

I don't know 2g systems that well, but there may also be an intake air temp sensor.
It won't affect the calculated fuel settings as much as the ect will, but it may be enough to affect it somewhat.

 

[Woodsy0]

Thanks for the hugely informative reply. We need more of you guys around the forums to troubleshoot for us dumbasses

I have DSMLink so I effectively have a datalogger, if memory serves I can log IAT, either way, can definitely log Coolant Temp. What values would I be looking for when checking coolant temp and intake air temp values, and what would they hint at?

I agree with you that the culprit is most likely the o-ring, I'm going to put in an order for the right fittings and fuel gauge through Summit, so hopefully I'll have some concrete info to supplement whatever you're having me do with the datalog values and I'll be able to troubleshoot more effectively.

Thanks again.

 

[toybreaker]

I've seen the o-ring issue myself.

My buddy had a pump failure on a road trip, and my used replacement pump o-ring wasn't in very good shape, so we re-used his. Didn't have any vaseline (the recommended o-ring lube), so we just kind of used residual fuel. The pump went on hard, and so we both knew it was a temporary solution...tease: which he drove on for 6 months , before he got enough beer available to accurately troubleshoot this issue... ) Checking the fuel pressure was all it took to nail the problem down.

As far as the temp senders go, you're styling if you got a logger

With everything stone cold, the iat and ect should match.

When you light the car, the ect will increase fairly rapidly, and the iat will hardy change at first. After a bit the car will reach operating temp. The ect will be 180~190ish, (maybe even ~200, but not much more) and the iat will have increased ~some~(sorry, a lot of variables affect the iat)

(I don't want to misinform you, because I don't have a cool guy datalogger, cry: ) so I haven't really tracked my own car much, but *generally* with the hood open, the iat won't change much, whereas, idling, with the hood closed on a hot day, it will change a bunch.)

It's measuring intake air temp, so I think you could just make a reasonable approximation, and be close enough. If you wanted to get all scientifical, I 'spose you could shoot the air can with an infrared temp gun...but I think close is ok, because it's more of a secondary imput for fuel trims. Imho, comparing iat and ect, cold, is enough verification that it's in the ballpark, unless the iat reads elleventy kerbillion * at operating temp.

The ect, on the other hand, is a major player in the calculated fuel trim. The ecu will add fuel to a cold motor, based on ect, and will gradually taper the cold fuel enrichment down, relative to ect input. When closed loop is obtained, the ecu will trim everything up much nicer, using measured feedback from the o2. To some extent, the ecu can trim out the affects of a slightly out of range ect this way.

These cars seem to have issues with the harness wires right at the ect plug, so if your ect numbers are kinky, I'd go there first, before I replaced the ect sensor itself. As the individual wire strands break, one by one, they gradually skew the ect reading to the ecu, but not enough to set the check engine light, or keep the car from starting, just enough to cause issues like this...(It is a dsm after all, and they like to be sure that we're paying the proper attention to them...so they just beg for attention in strange 'n unique ways.. )

One other thing you might want to check is the barometric sensor reading, (also in the airflow sensor.) It should read a reasonable facismile of the barometric pressure for your altitude, on that day, and shouldn't change much as you drive. (unless your commute is uphill/downhill +/-3500, each way, like mine.. ..) It is also more of secondary fuel trim input to the ecu, but it can throw things off enough to get the car to run funky cold, and yet still enable the ecu to trim it out when in closed loop, and warm.

Track your battery voltage as well, a weak battery could cause some funky driveability when the system runs at low voltage, after the car runs for a while, the voltage could recover enough to make everything run smoothly

Good luck, with a datalogger, and your accurate posts/info, I know you're going to solve this problem, and help some other people who read this thread along the way.

 

[Woodsy0]

Thanks again, I won't get a chance to start troubleshooting until this weekend when I limp my car home to my garage, but I'll make sure to post the solution when I find it.

Scott

 

[AL92]

I have similar issue in my 97. It starts ok though most of the time but once in awhile it don't want to start right away and then takes a bit more cranking but it does so far always start. Most restarts are ok and this problem usually is after it is sitting for a bit but not necessarily overnight it seems more random.
I put in a new supra tt pump in mine and they don't fit that great in 2g cars. Also was wondering about the coolant temp sensor .Anyone know the right part number for a oem one?
I got some non oem one from guy at parts counter that turned out was for a non turbo and car overheated very quickly! Lucky it seems to have not done any damage.
Another possible suspects are maybe leaking fuel injector or something wrong with the ecu.
I have dsmlink and a fuel pressure reg but not sure can log the exact event and not sure what am I looking at for fuel pressure check? Its hard to see the reg as the car is cranking over its under the hood. Am installing an in car fuel pressure gauge shortly though so that should do the trick maybe?

 

[Woodsy0]

I think your problem is more likely a torn o-ring at the mating point of the pump and the assembly. Since that problem allows fuel pressure to slowly bleed off, it makes the most sense that that problem would result in extended cranking time (to build back up the lost pressure) before the car starts.

I didn't think the o-ring problem applied to my situation since I never had trouble while cranking the engine, yet the engine stumbled until reaching operating temp.

As an update, however, the problem may have solved itself, since I haven't experienced this problem in the last couple days. I still intend to put on a fuel pressure gauge (delivered today) just in case it starts acting up again. If so, I'll post the fuel pressure values and any applicable info I can gather from DSMLink to see if anyone can troubleshoot.

Hopefully the car will stay docile and maybe cooperate with me for once . That's enough, I wont jinx it, knock wood, thumbs crossed, so on and so forth.

 

[toybreaker]

Hopefully the car will stay docile and maybe cooperate with me for once . That's enough, I wont jinx it, knock wood, thumbs crossed, so on and so forth.

Talk about the kiss of death...

Well, at least you can use this time to gather some everyday numbers from your logger.


I would just look at the datastream during warmup a few times while things are running well. Continue to compare the iat and ect against each other from cold, and watch for patterns in the airflow etc..and you'll be able to spot the anomoly when it starts acting up....

Which will most likely be when it's raining, and you've got a date...