I'm really frustrated right now, and totally dumbfounded!

A couple of weeks ago, my idle suddenly shot up to ~2000rpm, and my fuel level gauge stopped working. This happened suddenly and with no warning.

After a bit of troubleshooting, I also determined that anytime the fuel pump is being commanded to run, my TPS readings go nuts and fluctuate between 10-35% @ idle with the throttle closed.

Here's where things get really funny... There are three wires that go to the fuel level sender in the tank, one of which is a ground. If I disconnect either of the other two (one is the 5v reference, the other is the signal return), the TPS problem goes away, and the car idles fine. Obviously the fuel level gauge will not work when I do this. If I remove the fuel pump relay fuse in the fusebox in the engine compartment, and manually jumper the fuel pump, the TPS readings are normal, car idles fine. Still no fuel level reading on the gauge, and ATAP reports ~4-5% fuel level regardless of the level in the tank.

Some other things I've tried which have no effect:
Disconnected wire from PCM to fuel level gauge (at PCM). No change.
Disconnected fuel level signal wire @ PCM, and ran a "test" wire from the PCM back to the tank. No change.
Disconnected TPS sensor connector from TPS sensor. No change.
Verified good grounds on all PCM pins that are supposed to be grounded.
Visually checked engine harness (completely removed it from the car).

Does anyone have any ideas? I'm at the end of my wits here, and hate guessing how much fuel I have in the tank. It's not an option to buy an aftermarket fuel level gauge, since the stock sending unit's resistance does not match any value needed by the various aftermarket gauges. Most 1965 and later GM vehicles are 0ohms empty, and 90ohms full. The stock sender seems to be ~30ohms empty, and ~250ohms full. Autometer has a "universal" gauge that would almost work, but it is 33ohms empty, and 240ohms full. I could use that gauge, but it would show empty when full, and full when empty. At least I would have some kind of idea about my fuel level, though. <img border="0" title="" alt="[Embarrassed]" src="gr_emb.gif" />

I'm totally out of ideas here. I'm beginning to wonder if there is a problem in the PCM itself. If so, I don't care to replace it to solve the problem. I'd rather have a backwards aftermarket gauge, or no gauge at all.

Again, any suggestions here?

 

TTT

 

"The stock sender seems to be ~30ohms empty, and ~250ohms full. Autometer has a "universal" gauge that would almost work, but it is 33ohms empty, and 240ohms full. I could use that gauge, but it would show empty when full, and full when empty."

Did you mean the universal unit is 33ohms full and 240ohms empty ? I'm trying to see how that would read opposite.

<small>[ September 16, 2002, 03:18 PM: Message edited by: 98blackSS ]</small>

 

BTW, I also suspected possible ECM related problem as I started reading this post. How did you verfiy grounding to the pump and level sensor?

 

Oops. Yes, the Autometer is a 240-33ohm. 240ohm empty, 33ohms full.

My pump is external, so it's easy to verify the ground. <img border="0" title="" alt="[Wink]" src="gr_images/icons/wink.gif" /> All of the original wiring to the fuel level sender burned up in my fire, and I rewired it from scratch. The wires are still in perfect shape up to the point where they disappear to the top of the tank. I don't think that they could be the problem anyway, since I've got the strange issue with the TPS. Three things must be true for the TPS problem;
a)The fuel level sender wires are all connected.
b)The PCM is commanding the fuel pump to run
c)The "fuel pump" fuse is installed in the fusebox (remember that I can jumper the fuel pump and not see the TPS problem.

Under no circumstance does the fuel level gauge work, and ATAP reports ~5% fuel level.

On a somewhat more positive note, I found out that '89-up Ford senders are 16-158ohms. I could use a 89-up Ford gauge. It would show full at ~1/2 tank, and never quite make it all the way to empty. At least it would be readable, though.

 

Very strange indeed. I'm wondering if the universal type gauge could be opened up and have some connections reversed (maybe the actual indicator wires inside) to make it work in your case. I'm not familiar with how these fuel level gauges operate internally though. This would probably be as close to functional as you could get. I'm very curious as to whether a portion of your ECM is fried but like you said, you don't want to swap that out. Probably don't blame ya..

What I am wondering though, are you sure that the fuel level sensor is still working in the hopes of even using an aftermarket gauge? And if so, what is the voltage reading off of it? For full and empty? At very worst case, I could build you something (it wouldnt be pretty mind you) that you could use as a gauge as long as the voltage is linear from empty to full. I'll see if I can read up some on how most aftermarket fuel gauges work, although I'm pretty sure I know without looking..

 

Well... I got an email from someone who had suffered some similar problems with his car. I'm going to check out what he recommended tomorrow and see if that helps.

In the meantime, I dropped the rear end and the tank out of the car this afternoon, and verified all of the wiring was correct. The sender is still 100% functional. I now know the exact values also. It's 40ohms empty, and 245ohms full. If it comes down to me installing an aftermarket gauge and I can internally modify the universal Autometer 240-33ohm gauge to read backwards, it should be pretty darn accurate. If not, I'll try the late Ford 16-158ohm gauge, and at least have some sort of visual reading of my fuel level.

Like I stated above, ATAP was reporting 5% fuel level. I pumped right at 7 gallons of fuel out of the tank before I started. Obviously not right, since my car doesn't have the optional 140 gallon tank . <img border="0" title="" alt="[Wink]" src="gr_images/icons/wink.gif" />

I'll keep chipping away and eventually figure something out. <img border="0" title="" alt="[Cool]" src="gr_images/icons/cool.gif" />

 

Definitely looks to be a problem with the ECM then. I can't believe GM would reverse the 240-33 scheme on our cars..boneheads. I wonder if the sending unit is a standard type fit in the tank.. Might be another path if they aren't expensive, although I've never taken one out and can imagine it's probably a bitch to do.

I'm guessing the way that these sending units work is that they are basically just a potentiometer with 12v on the high side (240ohms to gnd) and ground on the other (around 30-40ohms between wiper and car gnd) with the 3rd wire (signal return) being the wiper on the pot. As fuel level in the tank becomes full, the sender is reading high (around 10v?) and close to 0 at empty. It seems these autometer gauges work on reversed logic assuming a compatible sending unit would have voltage and ground reversed from our scenario. I wonder if the ground wire for the sender is actually grounded to the tank or car. I'm thinking it may be possible to reverse supply and ground for the sender and get the correct logic we need for the autometer gauge to work properly. Just a thought. Might be able to ohm out the sender ground wire to car/tank ground and see if it's open. If so, it may just work... Of course, this is assuming that the sender is as simple as I'm thinking. If there is a more complicated circuit within the sender, then this theory would not be possible.

<small>[ September 17, 2002, 06:27 AM: Message edited by: 98blackSS ]</small>

 

"If I remove the fuel pump relay fuse in the fusebox in the engine compartment, and manually jumper the fuel pump, the TPS readings are normal, car idles fine."
Just reaching here....you didn't fry any wiring/contacts at/in the fusebox itself, did you?
Might take some more work with a volt/ohm meter.
Good luck.

 

Not that I've been able to find... Everything looks fine. This whole mess started out of nowhere. The car had been fine since getting it back together more than a month before this started. All of a sudden, the idle jumped up and the fuel level gauge stopped working.

 

98blackSS, the level sender is even more simple than you think... Three wires go to into the tank. One is a ground wire, which connects to the main "arm" that the stock pump and everything else is attached to. It doesn't directly have anything to do with the actual sender. The other two wires go directly to the sender itself (one wire to each side). The sender is nothing more than a float attached to an arm. The arm is attached to a potentiometer. Depending on the fuel level the resistance varies from the aforementioned 40-240ohm value. There is absolutely no way to reverse the sender so it reads the opposite way. I tried for well over an hour last night, and couldn't come up with anything.

I think it has an "oddball" value because it's actually fed through the PCM to the gauge. The typical 0-90ohm GM sender was a direct connection to the fuel gauge.

All I would have to do to make an aftermarket gauge work would be to ground one side of the sender, and attach the other side to the gauge (on the sender terminal). The other two connections on the gauge are ignition (+12v) and ground. All the gauge does is really measure voltage drop across the sender line (which is varied depending on the position/resistance of the float/sender).

If I can internally reverse the Autometer universal gauge, it will work. Then I can just do away with the entire stock insturment cluster (like I've been wanting to do anyhow), and install Autometer mechanical gauges (oil pressure, water temp) and electric voltage, fuel level, and maybe even a speedometer.

While I was putzing around with the car today (it's all put back together again now), I did pull the carpet up and check all of the wiring out to the main harness under the hood. I'm almost totally convinced it's something in the PCM, and I think I'll give up at this point and go for an aftermarket gauge...

 

Hmmm. It still sounds like a fried wire/contact to me. I know that's the hardest thing in the world to trace if there's no visible damage,takes a lot of time to ohm out / check voltage the wiring. Just for S's and Grins, why don't you get some spray type contact cleaner, remove all fuses/relays out of the boxes and use liberally,even the inside fuse box? The only reason I suggest that is that I understand you had a fire problem-if so, some wiring could have got hot enough to temporarily short and FUBAR'd either the wiring or any contact points. Also, if the wiring got hot enough somewhere, it could have changed the resistance (added resistance) which could be crucial on a 5V feed. Again, just throwing out ideas. <img border="0" title="" alt="[Sad]" src="gr_sad.gif" />

 

Interesting stuff. Let us know how it turns out. Hopefully it'll be right on...