05 Cobalt fuel problem

[56pontiac]

2005 Cobalt.No fuel. I have an 05 Cobalt.Wife drove home from work.I went out a few hours later to move it.It was a no start. Checked and it isn't getting fuel.Pulled the fuel pump and bought a used one with 13,00 miles (95,000 on ours ) .I also put in a new relay and checked the fuse & put in a new fuel filter.Still no fuel.Tested for current at connection where pump plugs in and there is current.Test light is not real bright.I don't have anything to do a voltage test with. When I turn the key to start,the fuel pump does not run.Getting Any suggestions? Thanks.

[rodeo02]

Originally Posted by 56pontiac

Tested for current at connection where pump plugs in and there is current.... When I turn the key to start,the fuel pump does not run.Getting Any suggestions? Thanks.

So there is NO power ever to the pump with the ignition set to run? Normally with the ignition cycled to run the FP will run, charge the line then cut off.

Joel

[56pontiac]

When I put the key in run position there is current. But when I put the plug for the fuel pump back in, the fuel pump is not running.I know it should charge the pump then shut down.But there is nothing from the FP.I pulled the plug on the pump at least a dozen times to check for current to the pump. Everytime it shows current at the plug terminal.I have a buddy turn the key off, then on.When the key is turned on, the test light lights up.When he turns it off, it goes out.But the light isn't bright when it is turned back on.So there is current to the plug located on the underbody pan behind the rear seat.But the fuel pump will not run. I don't know if it takes the full 12 V to run it. If the test light is right it shows a weak light. I would say the light is about half as bright as when running it right off the battery.I don't have a volt meter to check. Mine is broke. I learned not to slam the tool box lid down when your voltmeter is half in and half out. Thanks

[rodeo02]

Geez I wish I had some more ideas. Does your 2005 have a fuel pressure regulator on the fuel rail? I know most if not all the new GMs use pulse-width modulated control for the FPs now. Basically it's 'variable speed' control of the pump. It's all done thru the ECU and BCM. That might explain the dim test light you were seeing (lower voltage being sent to the pump). There's other 'permissives' that need to be met for the pump to run/continue running. I'm not knowledgeable there.

Joel

[56pontiac]

There is a valve (Similar to a tire valve) on the fuel rail. I'm at a loss on these new cars.I just know you need expensive equipment just to find a minor problem.I appreciate all the help and suggestions. Makes sense about the variable power to the fuel pump giving a dim light. Is there a way to test a fuel pump out of the vehicle ? Thanks.

[rodeo02]

Yep, the book name for them is Shrader valve. It's basically a test and purge point on the fuel rail. You should get a serious spray of fuel out of there with the key to run or engine on. I dunno the specs of what the PSIG should be.

Joel

[J-Ri]

the Schrader valve is for a fuel pressure gauge to be attached to, not to stick things in and see if you get "a serious spray of fuel" While I don't know the spec offhand (I think it's at least high 40s), I know the ecotecs will run, although not well, on as little as 15 PSI.

[56pontiac]

My buddy talked to a retired GM mechanic. He seems to think the fuel pump that came with the car, and the used one I bought are both junk. I have a 30 day warranty on the used one. I just don't know how to test it. The guy at the yard said to hook up a 12V wire to the hot and the ground to a ground. I'm going to take the pump I took out of the car to the junkyard tomorrow and see if they want to test it.If it doesn't work, then I take the other pump out and try it. The car is at a friends house and it's 20 miles each way.So don't want to make a trip if I don't have to.My vehicle I use now is a 1977 Dodge M880 3/4 ton pickup 4 wheel drive with 4:10's. I get 8 MPG with the wind at my back going downhill. That's why it only has 66,000 miles on it. Again,thanks to all who help.It is VERY much appreciated.

[rodeo02]

Originally Posted by J-Ri

the Schrader valve is for a fuel pressure gauge to be attached to, not to stick things in and see if you get "a serious spray of fuel" .

Sure, but it is a quick redneck 'troubleshoot' point!

[56pontiac]

UPDATE - I took the fuel pump that was on the car to the yard where I bought the used one. They bench tested it and it is good. So that would mean the problem is elsewhere in the electrical / computer system.

[J-Ri]

Originally Posted by rodeo02

Sure, but it is a quick redneck 'troubleshoot' point!

You don't need to tell me about redneck testing... I've done 'em all, check out the truck in my sig and avatar (before lift and wheelbase mod), tell me that don't say redneck all over it I'd be lying if I said I'd never done the spray test, but I'm always worried about potential liability (as screwed up as the "justice" system is) from telling someone that and have them do it 3 ft. from a space heater while his buddy lights a cigarette and throws the match under the car... and have them end up owning my house and cars.

Originally Posted by 56pontiac

UPDATE - I took the fuel pump that was on the car to the yard where I bought the used one. They bench tested it and it is good. So that would mean the problem is elsewhere in the electrical / computer system.

Just to clarify the electrical measurements, current is amps and is measured with a meter either a clamp-on inductive meter or an inline meter. This must be done with the circuit intact because you are measuring the amperage that the pump is using. Voltage is electrical "pressure" and can be checked with the pump unplugged, it sounds like that's probably what you were checking when you said you had current to the pump?

Check and see if the fuel system has a regulator on the fuel rail or not. I have an '04 Cavalier which does have a regulator and return line, but I think all Cobalts regardless of year have a returnless system that uses the pulse-width modulation that Rodeo02 mentioned. These must be pretty good cars, I havn't seen one except for a battery problem, so I'm not familiar at all with them. If it is returnless, try unplugging the fuel pressure sensor (what tells the PCM what duty cycle to command). I would assume that would make the pump run at some "should work most of the time" duty cycle. If that makes the pump run, chances are the sensor is reading a pressure over what should be there, so the PCM is not commanding the pump on. Not all returnless systems are pulse-width modulated, some have the regulator in the fuel pump module.

You'll be miles ahead if you buy a new DVOM, especially one that has a frequency/duty cycle measurement on it. Also, if you didn't check for a good ground to the pump, do that before you do anything else.

EDIT: I got to thinking about it, and if you replaced the fuel pump relay, then most likely it is not pulse-width modulated. A relay is just a magnetically operated switch, which is incapable of switching as fast as necessary to control a motor's speed (and would weld the contacts together in no time). It's not impossible that the relay is just a safety measure to be sure that power can be cut when needed, but the output to the pump would have to come from a transistor inside the PCM.

[56pontiac]

Originally Posted by J-Ri

You don't need to tell me about redneck testing... I've done 'em all, check out the truck in my sig and avatar (before lift and wheelbase mod), tell me that don't say redneck all over it I'd be lying if I said I'd never done the spray test, but I'm always worried about potential liability (as screwed up as the "justice" system is) from telling someone that and have them do it 3 ft. from a space heater while his buddy lights a cigarette and throws the match under the car... and have them end up owning my house and cars.



Just to clarify the electrical measurements, current is amps and is measured with a meter either a clamp-on inductive meter or an inline meter. This must be done with the circuit intact because you are measuring the amperage that the pump is using. Voltage is electrical "pressure" and can be checked with the pump unplugged, it sounds like that's probably what you were checking when you said you had current to the pump?

Check and see if the fuel system has a regulator on the fuel rail or not. I have an '04 Cavalier which does have a regulator and return line, but I think all Cobalts regardless of year have a returnless system that uses the pulse-width modulation that Rodeo02 mentioned. These must be pretty good cars, I havn't seen one except for a battery problem, so I'm not familiar at all with them. If it is returnless, try unplugging the fuel pressure sensor (what tells the PCM what duty cycle to command). I would assume that would make the pump run at some "should work most of the time" duty cycle. If that makes the pump run, chances are the sensor is reading a pressure over what should be there, so the PCM is not commanding the pump on. Not all returnless systems are pulse-width modulated, some have the regulator in the fuel pump module.

You'll be miles ahead if you buy a new DVOM, especially one that has a frequency/duty cycle measurement on it. Also, if you didn't check for a good ground to the pump, do that before you do anything else.

EDIT: I got to thinking about it, and if you replaced the fuel pump relay, then most likely it is not pulse-width modulated. A relay is just a magnetically operated switch, which is incapable of switching as fast as necessary to control a motor's speed (and would weld the contacts together in no time). It's not impossible that the relay is just a safety measure to be sure that power can be cut when needed, but the output to the pump would have to come from a transistor inside the PCM.

The only thing on the fuel rail is the Schrader valve. Near the fuel tank is a small tank that is a fuel tank pressure sensor. I've done the push the valve test also. On this car all I got was a very small psssh and literally a drop of gas.When the car first wouldn't start the first thing that crossed my mind was fuel pump. Had the wife turn the key while I stood beside the car and listened. We both heard the fuel pump cycle. So I had her do it again. This time nothing. We tried it at least a dozen more times, and still nothing. I have a wiring diagram and see there is a FUEL PUMP RELAY CONTROL. It shows this in the rear of the car.This is located in the PCM Along with the generator terminal output. I bet that item won't cost more than 4 or 5 grand !! I am going either Saturday afternoon,or Monday afternoon to see what I can do on the car. Thanks again for the help, always aprreciated. - Me -My wife

[rodeo02]

Great info J-Ri. To the OP; Are you getting any codes at all? Do you have a CAN compliant code reader to check? Is it possible the fuel pump won't fire due to not seeing a signal from the crankshaft position sensor or the likes? Could you have a chewed wire somewhere?

Joel

[56pontiac]

Originally Posted by rodeo02

Great info J-Ri. To the OP; Are you getting any codes at all? Do you have a CAN compliant code reader to check? Is it possible the fuel pump won't fire due to not seeing a signal from the crankshaft position sensor or the likes? Could you have a chewed wire somewhere?

Joel

I don't have a code reader. Can't ruleout a chewed wire some place. I checked the wires that are visible. Because of the poor design of a lot of items on this car I can't rule out anything. I hit a racoon 3 years ago.Over $800.00 in damages. Because there is no support, or protection, under the radiator it took it and the a/c condensor out. Not realizing a racoon would do that damage, by time I got home the car overheated. That cost me the head gasket.I haven't fixed it yet. Used a sealer that has held up good so far.But the check engine light came on a few times due to bad gas. If anyone around the PA,Ohio, W VA area uses the same gas, they will have problems. I won't mention the Gaz Stazions name though. I eat there, but never ever will buy the gaz again. I got gas at another station and the light went out and the rough running cleared up.But anyhow the check engine/ service engine light came on around 75,000 miles and has stayed on. I figured with the head gasket problem it might have been the O2 sensor. I get in the catch 22, if I buy a code reader, no money for parts. But I have never been called smart.Probably, no - should have, bought the reader and went from there. Problem is we moved my Mom in with us because she has Alzheimers. So I had to quit work to care for her.My wife works and that barely pays the bills.Other than the normal change oil and filters (air, oil) the only other thing I did was change the spark plugs.

[rodeo02]

Whoa dude you've got ALL kinds of stuff going on. So the 'coon thing happened when the car was basically new? That sucks. FWIW, none of them have any protection up/around the radiator anymore. Overheating in an all aluminum engine is major bad news, but if it happened 3yrs ago and a dab-o-goo has held you off since then, I'd think you're good to go. The only odd-ball 'chewed wire' concern I'd heard of in regards to a delta chassis was with the vehicle speed sensor. It got munched, basically the car wont shift, dash no workie, a mess. If you've pulled the battery a bunch of times and the car hasn't run since, I'm not sure if you'd get a good code read now anyway. I'd still try anyway. I know the ~$40 OBDII with CAN from ChinaFreightTools works like a champ (Harbor Freight Tools).

Joel