[chumley]

I have a 96 gcv w/ 3.8l and was operating decently although the past 6 months have had random issues with stalls at any given time (highway, city, stop or anywhere between) and would pull only P0340 cam pos sensor. This might not happen for a month or might happen every couple of days.Well, on my way to work the other day it died and wouldn't start. It actually did start but would run as a vehicle or machine would run when starting after being flooded (although it wasn't as there was no gas odor) very rough and low rpm but as soon as you touch the gas it would die (this was when the engine was hot). Later after work as I was waiting for the the tow truck to pick up, it wouldn't start at all. When I got home i checked compression - OK so timing chain not jumped. Coil pack tests ok. Had a spare (not new cps) I installed, still no start. Am looking for info on what to use as gap for cps when no paper attached - but can't find any (I notice people have .050 for crank pos sensor but no info for cam pos sensor). Can't seem to find too much on testing the crank pos sensor either. I did some checks per shop manual, with coil pack connector off and metering between B+ and ground - it says when you turn ignition on it should show voltage and then it should go away if sensors are ok - mine shows voltage when ignition is on but it doesn't go away. Has anyone seen this?

Thanks

 

[GLHS60]

I had the same problem with basically the same van and the problem was the cam sensor connector had a loose pin problem.

Thanks
Randy

 

[chumley]

Thanks for the suggestion. I will check that out.

 

[chumley]

Sorry to take so long, I have been tied up the last couple of nights. I checked the connectors/pins and they are OK. I think I may have been barking up the wrong tree as I was following the camshaft sensor sensor code that was pulled the last time it died and wouldn't start (although this had happened before and the van immediately started and ran fine right after it stalled). After doing the rounds on the electrical side, spark seems to be good, I figured I would pop the cap on the fuel rail and press the valve to make sure I was getting fuel - well a tiny bit of gas dribbled out just after cranking, so I am thinking fuel delivery issues. I tried banging on the tank, it started, ran for about 15 seconds then stalled. If it started again, as soon as you step on the gas it would also stall (again pointing to lack of fuel).I tried this banging on the tank a few times, it started two out of three times so nothing real conclusive except lack of fuel. I can't really hear the fuel pump when you turn the key on either. I pulled the relay under the hood for the fuel pump and metered it, I am getting power to the relay. I also changed the relay with the one beside it and no change. Is there a good way to test the fuel pump before I pull the tank or in reasearching this it appears there is a regulator - could it be that? Any good ideas for proving out the fuel pump/regulator?

thanks

 

[ImperialCrown]

[sub]You should have a good 12 volts+ at the fuel pump to ground with the engine cranking/running and good (~49 psi) regulated fuel pressure measured with a pressure gauge. [/sub]

 

[chumley]

Thanks ImperialCrown, I will give that a try. How would you know if a filter was plugging or the regulator not working properly?

 

[ImperialCrown]

A restricting fuel filter would start giving leaner mixtures under high fuel demand conditions like hill climbing or rapid accelerating. It should still allow enough fuel for an idle and other low fuel demand conditions.
The filter 'sock' inside the tank is the first line of defense in keeping the fuel supply clean. The fuel pump must have filtered fuel, as any debris passing through will destroy it. A 13 year old vehicle might have some debris sitting on the bottom of the tank and hopefully it doesn't get ingested and lock-up the pump. Multiple fuel pump failures usually mean that the tank has to come out and the inside cleaned or the tank itself replaced.
The fuel pressure gauge on the fuel rail fitting should show ~49 psi +/- 5 psi that would show that the fuel pressure regulator is doing it's job.

 

[chumley]

Thanks for the explanation ImperialCrown. I was poking around last night and again this morning. Some very strange stuff going on. I pulled off the connector on the fuel pump to see if I was actually getting voltage there (because I haven't been hearing too much humming from the pump when you first turn on) and no voltage on the green lead which I believe (from the shop manual) is the hot lead. I jimmied up some leads and strung from the battery + and - to the fuel pump and as best as I can tried to force the pump to operate - nothing.
I pulled the relay in the fuse box for the pump - I had previously swapped this with another to no avail. I did not hear the relays clicking when I pulled out or put in - I know enough about electronics that the relay should be hauling and clicking initially if it is getting power. So next I metered the pins in the socket for the fuel pump relay in the fuse box and this is what I found:
1/ There is power sitting there waiting to be sent to the fuel pump on the make contacts.
2/ I connected the coil side pins of the relay across the battery and it clicked and worked. I also checked the make contacts for continuity after the relay hauled and the relay is fine.
3/ I connected the relay (out of the vehicle and while the ignition was on) across the -ve ground side of the battery post and what is the hot +ve side of the relay socket and the relay hauls.
4/ I connected the relay (out of the vehicle and while the ignition was on) across the +ve hot side of the battery post and what is the ground -ve side of the relay socket and the relay does not haul
5/ However when I put a meter across the same ground side of the relay socket and the same hot side battery post it shows voltage.

So just to recap no voltage at fuel pump, relay is good but won't haul when it should, and I have conflicting evidence of why the relay is not working because everything meters ok (as far as all socket positions) however when the relay is in, it doesn't work. It is making very good contact and is not corroded at all.
This is very bizarre - any suggestions?

Thanks

 

[valiant67]

You won't normally have voltage at the fuel pump. It will come on for a second or two to prime the pump, but after a couple of key cycles even that stops. However, once the computer sees the engine is spinning then there is power applied to the pump. I would assume the relay is only powered when the engine is rotating as well except for that brief blip when the key is turned.

 

[chumley]

Thanks valiant67. I will get my assistant (son) to give me a hand turning it over to see if I get a momentary shot of voltage. Still bizarre that the relay is getting voltage but not working and the relay proves out OK. Just wondering if the computer is a problem? Don't really want to try replacing that unless I can prove that is the issue.

 

[John Wood]

For a normally working fuel pump, jumping the contact side of the fuel pump relay (i.e. #30 & 87) with the ignition switch in the run position will start the pump and it will run continuously.

Identify the spade lugs with those numbers and put a jumper in the relay socket corresponding to those numbers. Turn the ignition switch to run. If you hear nothing, go bang on the tank with your hand. If the pump starts running, you can probably start the car and move it to where you want it to be to change the pump, or drive it to a garage. Be sure to remove the jumper when you are done.

BTW, I did that exact method on my 96 Grand Voyager when the fuel pump became intermittant.

 

[chumley]

Thanks John. I metered across the hot pump green lead and the black ground lead on the pump and checked for resistance to see if the pump was open. It was open - no resistance or continuity. Will pick up a pump and change it out. I will update the post after completing that. Is there an easy way to get the plastic threaded locknut off the tank without breaking it - I am assuming if one taps the ribs they will probably break.

 

[ImperialCrown]

The fuel pump nut tool looks like:
http://www.etoolcart.com/gastankfuelpumpremovaltoolloc-940.aspx
A parts store may have one for loan if you leave a tool deposit and buy the part from them.
You can try a very blunt flatblade or chisel edge and tap it off ccw down near the base of the nut.

 

[John Wood]

For that price, I'd just buy it. It wasn't readily available when I did mine so I just made the tool from square tubing. The thing that made it tough was getting the measurement just right. Here is a picture of my tool and a modified filter tube that I used to pump out the tank with an intermittant pump that worked during the pumpout process.

 

[chumley]

Thanks again IC and JW, you guys are a wealth of knowledge. Nice tool JW - I have cobbled up some home made tools myself in the past. One more stupid question, I have the tank dropped down but have two more vent lines to disconnect (the ones with the weird ribbing on them - one goes on top of the tank above the area where the pump goes in and the other runs from the back of the tank to towards the front top of the tank). How do those come off, they aren't connectorized or don't appear to be able to twist off - I didn't want to try twisting off and break them. You guys know how those come off?

Thanks in advance

 

[ImperialCrown]

There are some vapor/valve lines that are permanently heat-shrunk onto the connector fitting ends and don't come off. They are serviced with the fuel tank assembly.

 

[John Wood]

It is possible to replace the pump by only lowering the tank and not messing with the vent lines that go up to the filler tube. I've done it both ways. Otherwise, you will need to unsnap the lines at the filler neck and the filler neck will need to be lowered with the tank. There is a cross brace that goes between the two vent lines when they are attached, that is why I just lower the entire tank and filler neck and avoid trying to disconnect those lines to get the tank out. The plastic lines can be brittle so work carefully.

 

[chumley]

Thanks again JW and IC. It's funny that in the shop manual and any other info I can find, they simply say remove vent pipe at the tank - sounds pretty easy, but you guys are correct, there appears no easy way to remove them as that's what I thought as well - they appear to be one with the valves that are in the tank.

 

[ImperialCrown]

You only need to remove the inboard support strap bolts and lower the inboard side of the fuel tank to change the pump.
If you want to remove the tank from the vehicle and slide it out from underneath for interior cleaning, then the filler tube hose and quick-connect lines can be released. Just squeeze the oblong sides of the oval ring at the connection that will unhook the retainer and you can then separate the line.
To reinstall the line, just push it back on until you hear the latches click. Tug it apart slightly to ensure that it has latched.

 

[chumley]

Thanks again IC and JW, I had hoped to get this done on Sunday but had company come over for the day. Monday had work and TKD class after, but I started working on it late and put in a few hours. Had some trials and tribulations but got the tank out, changed the pump and pretty much got it buttoned up except for a few details and shut it down as it was around 12 am. I just got a chance to got out earlier this evening and finish up. It took a while to prime up but she started up and seems fine, I still have to put the air breather and few other items but the pump definitely works and we appear to be back in business.
Thanks for your assistance.