Erratic Fuel Gauge

[dkinzer]

Over the last few months I've been having a problem with the fuel gauge on my 1991 LS400. While driving with a relatively full tank, the needle will slowly drift to a quarter or so. After a while, I'll notice that it has moved back up to what is probably the correct level. I pulled the connector on the sender at the tank and cleaned the contacts. For a few days, it seemed that that had fixed the problem but, alas, it has returned.

I've read other threads here and elsewhere that talk of changing out the electrolytic capacitors on the instrument panel circuit board. I don't remember seeing what symptom or symptoms the various posters were trying to correct so I don't have any feel for whether or not that might solve my problem. I suppose it could be the sender unit itself. I did read one post that talked about cleaning the variable resistor on the sender but, again, it didn't say what symptom was being addressed.

So, based on my description of the fluctuating indicator, any ideas on where the problem might lie?

[landar]

Welcome.

I see that you have gotten several replies to this issue from the Club Lexus forum. And it could be the cluster or the sender unit. It will take a little detective work to figure out.

If it were mine, I would probably separate the sending unit from the gauge by disconnecting the sending unit (but leave it in the tank), then 'spoof' the system by jumping a resistor across the open cable connector as if the sending unit were at a specific level. Watch the gauge for erratic behavior and, if none, then the issue is with the sending unit. Else, it is in the gauge assembly unit. I do not know exactly what resistance value to suggest to you but guessing maybe 5 ohms? You could put an ohmmeter on the sending unit and do some measuring to get a rough idea.

Also, do not rule out a bad or corroded connector. I always like to go for the really simple stuff first. Let us know what you find.

[dkinzer]

I would probably separate the sending unit from the gauge by disconnecting the sending unit

That's a good suggestion. I filled the tank and disconnected the sender. There are three terminals on the sender unit and the resistance measured 4.7 ohms between the two closest to each other and about 24K ohms between each of those and the third. Based on information I've read elsewhere, I guessed that the 4.7 ohm pair was probably the variable resistance with the third possibly being a ground.

At any rate, I connected a 10 ohm resistor between the closest terminals on the connector and the gauge now reads almost full. Without the resistor the gauge reads empty. Now, I'll watch for fluctuations as I'm driving around.

[landar]

If you have another 10 ohm resistor, you could put them in parallel to make 5 ohms and maybe get about a half tank reading (hopefully). Just be careful not to run out of gas! Or keep a spare container with you just in case depending upon how much time you need to run the 'test'.

[dkinzer]

[You could use] 5 ohms and maybe get about a half tank reading (hopefully).

Actually, the resistance increases as the tank level falls. If you pull of the sender's connector (infinite ohms), the gauge reads empty.

[landar]

Oh yeah, sorry...should have said to put two 10 ohm in series for a total of 20 ohms and see where the gauge rests. Not sure how linear it will be.

[landar]

I would just like to see your gauge perform properly in several different linear positions (such as approx. halfway) to make sure the movement is not having trouble holding a particular position in its sweep range.