- This topic has 5 replies, 4 voices, and was last updated 10 years, 5 months ago by
.
-
Topic
-
ok, a bit of background here. When I bought this car, the fuel gauge never read empty. When I pulled the fuel sending unit to investigate, it seemed obvious that it was a patched together piece, with crimp connectors, and various screws and hardware holding things together, so I figured someone had patched in a resistor and float arm from another car, and it was simply not compatible. I left it be, relying on the low fuel warning light to keep me out of trouble, and that was fine, until . . .. the low fuel warning light stopped working. Once I established that it was the sender, not the bulb that had failed, I ordered a new sending unit from Honda.
Everything was right in the world for just about a year, THEN . . . the low fuel warning light came on non stop. Of course, by the time I found my sales receipt from Honda, my one year warranty had expired, so once again, I bought a new sending unit.
And, again, everything worked fine, this time for almost two years. Now, just a few weeks ago, I noticed that the fuel gauge wasn’t dropping all the way to E before the low fuel light came on. Over the course of a couple of tanks, I am now at the point where the gauge barely makes it to 1/2 before I am out of fuel. I have verified that the gauge is capable of moving down to the E mark, by disconnecting the sender and letting the gauge drop (not quite to E, didn’t want to break anything).
According to the manual, the resistance on my sending unit should be about 3 – 4 at full, and around 105 or so empty. The numbers I am seeing are 3 – 4 full, and about 40 empty, with a nice smooth transition between them.
Since the gauge drops to E with the sending unit unplugged, I am sure there isn’t a problem in the wiring between the sending unit and gauge, the problem has to be in the sending unit. Any ideas? I haven’t dug into the unit yet to see how the resistance is created, though it doesn’t appear to be the wound up resistance wire I am used to, looks like the wiper runs across a bunch of foil pads on a circuit board. I didn’t see any obvious problems with those foil pads, so I am assuming my problem is on the other side of the board, under the cover.
-
Replies
-
with GM cars a simliar phenominon happens, before the fuel pump in the fuel sending unit fails, sometimes tens of thousands of miles before the fuel gauge seems not too be sure what the fuel level is, the fuel gauge will wave wildly all over, as if the fuel level indicatior is just reading the fuel as it sloshes around…
it is caused buy corrosion apparently on the “variable resistor.” on the fuel level indicator, and the ethanol in the fuel is largely too blame in many cases, replacing the entire fuel sending unit is the recommended fix too the issue, i have seen two GM cars with this issue, my 2000 Buick Century Custom, replacing the fuel sending unit fixed it, and my friends 1997 Buick Riviera, and also replacing the fuel sending unit after failure fixed the fuel gauge waving wildly.
now with that being said the only thing i can think of is the fuel sending unit is faulty myself, as replacing it seems too fix the issue, the issue just returns after about a year making me thing the ethanol in the gasoline like with the GM fuel sending units is eating something in the fuel sending units, have you tried different brands of fuel sending units? possibly try a higher quality brand fuel sending unit?
It could still be something in the wiring since your previous sending unit exhibited similar symptoms. I think if nothing else it might not be a bad idea to check the wiring between the cluster and the sending unit just to know for sure. There may be a short that’s actually damaging the sending units, OR there could be a connector issue as suggested. Corroded connectors can cause increased resistance which would give false readings at the gauge. More info on solving electrical issues here.
http://www.ericthecarguy.com/faq/solving-automotive-electrical-problems
Good luck and keep us posted.
I thought about corrosion, but that would lead to higher resistance, not lower. My issue is caused by too little resistance. When I disconnect the sender from the wiring (infinite resistance, no connection at all) the gauge drops toward E as it should. I have repeated this test over and over, at times when I know I have hardly any fuel in the tank, and the gauge is reading very high, perhaps as much as 3/4 tank, as soon as I disconnect the wiring at the sending unit, the gauge begins to drop.
For this reason, I am pretty certain it is something between the sending unit, and the small bit of wiring attached to it, and it has to be a short of some description, not corrosion. The gauge moves up toward full with lower resistance. Indeed, when I measured the resistance at the sender, my results were a bit less than half of what they should have been, with only 40 Ohms or so at the empty position where I should have had around 105 – 110 or so.
I guess I’m going to tear into the covered section of the sender and see what is up there with the resistor section of the unit, as that just about has to be where the problem lies. I will let you all know what I find when I get in there.
And, for what it is worth, I’ve never found any aftermarket supplier of fuel sending units for my car. Both of my brand new units were genuine Honda parts ordered from Majestic Honda, go figure.
Well, we have a result, and to be honest, I almost don’t believe it myself.
I pulled the sender out again, and measured resistance. 3.5 full, only about 30 empty, no where near the correct 100 plus.
I then measured resistance from the 12v in lead (red) of the sender connector to the resistor, moving my test lead along the resistor, and surprisingly, it had an excellent range, going from just over 3 all the way up to about 105.
Next up, I took a piece of thin cardboard, and used it to isolate the wiper from the resistor. Testing from the red lead to ground, I was seeing a 30 Ohm resistance. After much close up visual examination, I determined that there was no visible way current could be getting from the resistor over to the ground, so I cut the red wire about halfway between the connector at the top of the sender assembly, and the resistor.
Testing between the red wire where I had cut it, and ground, through the resistor assembly, I saw no continuity. Testing between the other side of the red wire (the side connected to the terminals at the top of the sender assembly) to ground, I was seeing 30 Ohms, meaning I had a short with 30 Ohms of resistance in the connector, where it passes through the metal cap.
At this point, I was determined NOT to hand Honda another hundred plus dollars for a replacement, so I grabbed my pocket screwdriver (which I bought ONLY because I saw how often Eric reaches for his!) and used it to gently pry up on the plastic terminal housing, then tested again. This time, my resistance was 50 Ohms, going in the right direction. A bit more prying, and I had 0 continuity. I then ran the end of a small screw between the terminal housing, and the metal cap to hold that gap, and filled in all around / under that terminal housing with epoxy.
Finally, I crimped, and soldered the red wire back together, and protected it with two layers of heat shrink tubing, then installed the sender back into the car. Low and behold, my fuel gauge responded by moving from Full to just above half a tank, a big move in the right direction.
Guess I need to run the tank down to empty now and make sure the gauge responds, but it looks like I’ve got the problem solved at this point, and it definitely was not in a place where I would have expected a failure to be.
- You must be logged in to reply to this topic.