Fuel system issue question - '93 Shadow 3.0

[Wylie6G72]

Hey guys,
Got A new fuel tank, pump and sender (both Delphi) installed last Wednesday. Got it back with a little less than a quarter tank, drove around and ran a bunch of errands on Thursday. Ran great. It had about an eighth of a tank in it when I headed back home so I stopped and filled it up.

Got home and discovered a slow drip under the right rear corner of the tank. Determined it was coming from the sender or the pump. Stuck some clear tubing down the filler tube and siphoned most of the gas out, down to about 1/8 of a tank, which stopped the drip, and just left the car untouched until this morning.

Attempted to crank it this morning and it will not start. Cycled the key multiple times, can hear the pump run, but it won't fire. Added some gas where it now shows a quarter of a tank. Still will not fire.

The ECU has a stored code 21 "oxygen sensor." Don't know how long the code's been there, but I replaced the oxygen sensor a few weeks ago. It may not like the new one, but that's not causing a non-start. Cleared the code.

Pulled a spark plug and cranked it over, she's got spark.

When I pulled the pressure hose off the end of the fuel rail a few weeks ago to replace the hose, there was lots of built up pressure and a decent amount of fuel was released.

I just pulled the same hose today, nothing came out. There's not even fuel resting in the end of the hose. Gauge is registering a quarter of a tank of fuel. Should be more than enough to submerge the sock on the pump. No fuel came out of the hose when cycling the ignition, or cranking.

Pulled the ASD relay and powered it with 12 volts, the relay clicks and has continuity. ASD circuit tests at 12.1 volts with the ignition on.

Bad fuel pump?

 

[drtyler]

Is it possible that you siphoned too much fuel from the tank?

Since you can hear the pump running and the pump is new and was just working, it seems unlikely that it could die so quickly.

 

[Wylie6G72]

Is it possible that you siphoned too much fuel from the tank?

Since you can hear the pump running and the pump is new and was just working, it seems unlikely that it could die so quickly.

I considered that, but I think there's a problem with the pump if the car won't start on a quarter tank of gas. I could fill it up to half and see what happens, but if it's still the pump then I have a half a tank of gas to siphon instead of a quarter. Not a huge problem in the scheme of things, I would just hope that a brand new pump would perform better than this.

 

[Chris142]

When you stuck a hose in it to siphon you knocked something with the pump off. A wire or a hose. I did the same to our can am.

 

[dishdude]

If it's leaking the tank had to come down again anyway, might as well just drop it and see what's going on.

 

[Lubener]

A cleaning of the tank and a new filter should always be done with a new pump. Didn't see you mention that. Running down to 1/8 tank on such an car with probably the original tank is not a good idea.

Yes, the new pump is probably bad. Just hearing it prime doesn't mean there is sufficient pressure.

 

[Wylie6G72]

I slept on Drtyler's comment and went out this morning and poured more gas into the tank. The gauge registered a hair below half. The car fired up. Thank you for putting that bug in my ear.

Lubener, sorry if I wasn't clear, the tank, sender, and pump were all replaced. The filter has 4,000 miles on it. I ran it down to 1/8 of a tank on the new tank, not the old.

My question now is this: Do you think there's some sort of issue going on that the pump couldn't prime itself with a quarter tank of gas? I've run a car out of gas once in my life, ironically a 93 Plymouth Sundance, and I do remember having to pour two 1 gallon cans of gas into it before it would fire. That was nowhere near half full. I feel like I've still got an issue here.

It's not normal for me to run a tank lower than a quarter, so I'm not worried about getting stranded somewhere. I just did this in an attempt to make sure everything was in good working order with the new equipment, and to get as much gas out as possible, as it was returning to the shop to address the leak.

Since the pump can't be put in sideways or upside down, I don't know what the issue may be. Just doesn't seem normal to require so much gas to prime.

 

[D60]

Is the sender correct? I was just complaining about a Delphi assy that didn't read fuel level correctly from Day 1.

IOW is it possible you siphoned to basically E?

 

[drtyler]

Is the sender correct? I was just complaining about a Delphi assy that didn't read fuel level correctly from Day 1.

IOW is it possible you siphoned to basically E?

That was my line of thinking. Maybe siphoned out more than realized, or perhaps the gauge not reading correctly. Pump is probably fine (as far as new aftermarket pumps go), but it just didn't have enough fuel.

In the first post, he mentioned siphoning fuel to 1/8 tank. That's pretty low, especially if the car isn't level. Replacement fuel pump may not be quite as long as the original, who knows.

@D60 needs to start a new company specializing in good quality aftermarket fuel pumps. I'll be first in line to buy one.

 

[D60]

@D60 needs to start a new company specializing in good quality aftermarket fuel pumps. I'll be first in line to buy one

It wouldn't make it. Everyone buys the lowest price on Amazon and thinks they're clever because they "saved" money

Also OP if you're back in there I'd figure the ohm range of your sender and check it at empty and full, and look for a smooth-ish progression along the scale. I wonder if this uses Chrysler's oddball 30-240?? (can't remember) range? Not that it matters, I just find it interesting how manufacturers choose different ranges of resistance and some are inverse.

 

[Hohn]

Man, a Shadow with the Mitsubishi 3.0 is a time machine for me. I'm delighted to read there's one still being cared for and in service.


Loss of prime is always an air ingress issue. The problem is that it's difficult to distinguish between losing prime and never having achieved it at all. Those are two separate fault trees with different diagnostic procedures.

 

[Hohn]

Dang, check out the reply generated by Grok AI:

Keep in mind, this is just from what it can see without having a BITOG account:

Hi [Original Poster's Username],Thanks for the update about the recently replaced fuel tank. That’s a big step, but it sounds like you’re still having issues. Could you share more details about the symptoms (e.g., hard starting, hesitation, stalling, or loss of power) and whether the issue started before or after the tank replacement? Here are some thoughts tailored to your 1993 Shadow 3.0L (Mitsubishi 6G72 engine) that might help narrow it down:

1. Fuel Pump Installation: Since the tank was replaced, the fuel pump (or its strainer) might have been disturbed or improperly installed. Double-check that the pump is securely mounted and the electrical connections are tight. If the pump was reused, it could be weak—these can fail on high-mileage vehicles. Test fuel pressure at the rail (should be ~35-40 psi, check your manual). A gauge is cheap to rent at most auto parts stores.
2. Fuel Lines and Connections: During the tank swap, fuel lines or fittings might have been kinked, loosely connected, or contaminated with debris. Inspect the lines from the tank to the engine for leaks, pinches, or blockages. Ensure the quick-connects are fully seated.
3. Fuel Filter: If the filter wasn’t replaced with the tank, it could be clogged, especially if the old tank had rust or sediment. The filter is usually under the car near the rear axle. Swap it with a quality unit (Wix or Mopar) to rule it out.
4. Tank Venting/EVAP Issues: A problem with the tank’s venting system (e.g., clogged rollover valve or EVAP canister) could cause pressure or vacuum issues, leading to fuel delivery problems. Check the vent lines for blockages, especially since the tank was recently handled.
5. Injector or Regulator Issues: If the tank replacement stirred up debris, the fuel injectors or pressure regulator might be affected. A quick test is to check for fuel in the regulator’s vacuum line (a sign of a bad diaphragm). For injectors, try running a bottle of Chevron Techron or Lucas Fuel System Cleaner through a full tank to clean them up.
6. Other Checks: Since it’s a 1993 model, pull any OBD-I codes (use the key dance method or a basic scanner) to see if sensors like the MAP, TPS, or even ignition components are contributing. Sometimes ignition issues (plugs, wires, or distributor) can feel like fuel problems on these engines.

 

[Wylie6G72]

Is the sender correct? I was just complaining about a Delphi assy that didn't read fuel level correctly from Day 1.

IOW is it possible you siphoned to basically E?

Anything's possible I suppose. I siphoned until it dropped below a quarter, and it looks like that's a no-no. I've never had any problems driving because I rarely, if ever let my tank get below a quarter.

Not sure what to do if the new sender isn't sending the correct current to the gauge. My temptation is to siphon all the gas out, pour in exactly 7 gallons and see if it doesn't land on 1/2.

 

[Wylie6G72]

Man, a Shadow with the Mitsubishi 3.0 is a time machine for me. I'm delighted to read there's one still being cared for and in service.


Loss of prime is always an air ingress issue. The problem is that it's difficult to distinguish between losing prime and never having achieved it at all. Those are two separate fault trees with different diagnostic procedures.

They're a rare sight on the roads in Dallas. You have a better chance of seeing Lancers, Shelby Chargers, and Omni GLHs. Guys at the gas station like to start up conversations about it. Younger people have no idea what it is.

 

[D60]

Anything's possible I suppose. I siphoned until it dropped below a quarter, and it looks like that's a no-no. I've never had any problems driving because I rarely, if ever let my tank get below a quarter.

Not sure what to do if the new sender isn't sending the correct current to the gauge. My temptation is to siphon all the gas out, pour in exactly 7 gallons and see if it doesn't land on 1/2.

Well the only problem with that is many manufacturers don't exactly follow a linear fill indication rate. I guess since your gauge is direct-read, there's no PCM toying with the readings, but still the shape of the tank can mean it reads higher or lower depending upon fill level.

Imagine you put a float in a cone shape-- is halfway vertically up the cone half-full?

I think for fuel gauges the best you can hope for is full when it's truly full, empty when it's truly (close to) empty, and no jumping of the gauge in between. Fords would notoriously start to jump around just sitting still -- some GMT400s, too

If it's not reading correctly you can try adjusting it by lightly bending the arm, but this won't help if you don't see near the max ranges of spec when on the bench (ie one end of the scale is likely 30 or 240 and vice versa). I could be wrong on my numbers but the point is the range is the range.

 

[D60]

Also IIRC 2nd Gen Rams -- which were darn close to the era of your Shadow -- used a very different method for their fuel gauges. I almost think it was varying voltage rather than a varying ground?

I suffer from CRS. But the point is do a bunch of research before you tear out your sending unit again based upon this internet rando.

 

[Wylie6G72]

Also IIRC 2nd Gen Rams -- which were darn close to the era of your Shadow -- used a very different method for their fuel gauges. I almost think it was varying voltage rather than a varying ground?

I suffer from CRS. But the point is do a bunch of research before you tear out your sending unit again based upon this internet rando.

This wouldn't be the first time on this car that I was forced to just learn about the problem and adapt to it rather than beat my head against the wall trying to solve it.

 

[Hohn]

They're a rare sight on the roads in Dallas. You have a better chance of seeing Lancers, Shelby Chargers, and Omni GLHs. Guys at the gas station like to start up conversations about it. Younger people have no idea what it is.

View attachment 288784

I remember driving a V6 manual trans variant of one of them thinking holy cow this thing really scoots.

Of that era, the most memorable car to me was the Dodge Spirit R/T. My Dad's buddy had one-- it REALLY moved. 224hp in essentially one of the last k-cars.


Then there were all the Shelby one-offs Chrysler did. Omni GLHS, etc.

This was, after all, the 1980s Chrysler that gave us the Dodge Dakota CONVERTIBLE! Yes, a convertible pickup.

Lord, I love being a Gen X'er.

 

[Wylie6G72]

I remember driving a V6 manual trans variant of one of them thinking holy cow this thing really scoots.

Of that era, the most memorable car to me was the Dodge Spirit R/T. My Dad's buddy had one-- it REALLY moved. 224hp in essentially one of the last k-cars.


Then there were all the Shelby one-offs Chrysler did. Omni GLHS, etc.

This was, after all, the 1980s Chrysler that gave us the Dodge Dakota CONVERTIBLE! Yes, a convertible pickup.

Lord, I love being a Gen X'er.

View attachment 288796

A buddy of mine in Wyoming bought a Shelby Charger GLHS when it was a pretty new used car. Man that thing was fun. Took a lot of V8 iron by surprise at the stop lights. Growing up with that generation of cars, and now driving a 2024 Kia Forte, It's hilarious how much power a 4-cylinder puts out these days. My '93 Sundance duster, bought new off the showroom floor, was a V6 5-speed. Still my favorite car of all the ones I've owned.

 

[Hohn]

A buddy of mine in Wyoming bought a Shelby Charger GLHS when it was a pretty new used car. Man that thing was fun. Took a lot of V8 iron by surprise at the stop lights. Growing up with that generation of cars, and now driving a 2024 Kia Forte, It's hilarious how much power a 4-cylinder puts out these days. My '93 Sundance duster, bought new off the showroom floor, was a V6 5-speed. Still my favorite car of all the ones I've owned.

Sundance Duster-- that was the car I drove. Like an overpowered go-kart, what a blast of a car. :sigh: