View Full Version : Fuel guage/ sending unit troubles

09-18-2009, 07:26 PM
I purchased a used sending unit that appears to be working. And I have a new gas guage Stewart Warner Part# 1562647 that was in the glove box. I don't have a working ohm-meter but this is what I have done using a test light:

I have accesory power at the Gas gauge terminal closest to the driverside of the 66 Daytona (Blk. wire/wht.tr.)
I register no power at the gauge terminal closet to the passenger side of the car (red wire/ blk. tr.)
When I hook up the new guage to the positive lead or leads suspended in air I get the same results.
No matter what I do the guage needles stay pegged at E. Grounded or hooked to sending unit with plate grounded.

Does it matter which way power flows through the gauge, and how do I know it wasn't hooked up wrong already[?]

I wanted to put my used sending unit in the tank and install the tank but I can't determine if the sending unit is working for sure, but this is what I did:

I hooked the test light up from sending unit plate to ground and put power to sending unit terminal. With float at full position test light filament has an orange glow, as you move the float torwards empty position, the glow decreases to nothing.

Do you guy's think I can assume the sending unit is ok[?]

I also put power to the sending unit wiring (red wire/blk tr.) and it seems to be OK, (bright test light at sending unit terminal.)

Any advice about the gauges, what else should I try[?]:(


66 Daytona all stock

09-18-2009, 08:01 PM
Based upon test light filament has an orange glow I'd go ahead and try it. Remember to run a bonding line from one of the sender mounting screws to a good ground.

Tom Bredehoft
'53 Commander Coupe (since 1959)
'55 President (6H Y6) State Sedan
....On the road, again....
'05 Legacy Ltd Wagon
All Indiana built cars

09-21-2009, 06:44 PM
Under tech tips, specs and data, there is a link to www.studebakerparts.com for wiring diagrams. It should show you how to hook the gauge and sender up.

(your diagram should be similar to the following)

The gauge should have three contacts. One is for positive, one is for negative and the last one for the wire from the sender. The sender will have two contacts. One is the sender wire to the gauge. And the other is ground.

The ground is a common ground. So in a test scenario, it should be connected to the sender ground, the gauge ground and the power supply.

DC electrical power flows negative to positive.

Power flows from the power supply to the gauge negative terminal and the sender. then through the sender ( variable resistor ) and on to the gauge, where it varies the current allowed to pass through the gauge. And then back to the power supply.

Your test light is working like a gauge. But, if you had the gauge out, you could have tested the entire system on a bench. That comes in handy as sometimes a gauge adjustment is needed.

63 Avanti R2, 4-Speed, 3.73 TT
Martinez, CA

09-21-2009, 09:45 PM
Actually the guage only has 2 terminal posts, and I do have a gauge in a box that came with the car. Other members told me the gauge should be grounded through its body. But it looks to me like the gauge is only held to the cluster by the 2 terminal post nuts which screw down on an insulator, apparently the back of the gauge body must be held against a metal part of the cluster grounding the gauge? My concern is that there are no markings to determine which post should be hooked to positive power. Seeing as it wasn't working I don't know that it is currently hooked up correctly.

DC power really flows from neg. to pos.?!

The sending unit only has 1 terminal and grounds through the tank.

I have ran full 12 volts through the guage (the one out of box), it takes some time before the needle starts to move but when it gets to about 1/2 it sticks then pegs all at once. I tried cycling power to get it to run to full but still sticks or pegs all at once. After removing power stays pegged for short time then drops back all at once.

I did try getting the gauge to work in circuit with sending unit, no success. I also bought batteries for my meters, the sending unit seems to work fairly well.

I couldn't get any movement out of the gauge in the dash.

Considering drilling the rivits on the guage to see if I can figure out why it hangs. Maybe it is hanging the whole time and that is why it won't work with reduced power in circuit with sending unit????

If I do I will post a picture of the gauge guts.


66 Daytona all stock

09-22-2009, 05:28 PM
It sounds to me like the gauge is bad.

It makes sense that you wouldn't have a 3rd terminal for ground on the car. I was basing it on the Avanti - which is why I pointed to a wiring diagram.

That being said, my gauge was sticking just like yours. There are/should be slots on the back that will enable you to place a screwdriver in the slot and slit it ever-so-minimally to get it to another location. Then it shouldn't stick.:

\ / <--- from the back of the gauge. The left slot controls the full limit. The right controls the empty limit. Moving the left one slightly to the right stopped mine from sticking. The gauge is still slow though.

63 Avanti R2, 4-Speed, 3.73 TT
Martinez, CA

09-22-2009, 10:59 PM
The gauges are SUPPOSED TO BE slow. They are "thermo' gauges that use a bimetal strip, deformed by heat from tiny heating elements, to move the pointer. They are slow, BY DESIGN, so that the gauge unit doesn't wear itself out following the excursions of the float as the fuel sloshes in the tank. I've watched electromagnetic (fast) gauges in British cars flicker all over the place as the fuel sloshes. Close to useless, they are.

The dash unit itself MUST be grounded. It normally gets ground from being mounted in the metal instrument cluster, which is in turn grounded by being mounted in a metal dash structure, which is grounded by being bolted to the body. Now it's possible that the body is not properly grounded, but you would be having trouble with lights, too, if that were the case.

Which terminal is which? If the temperature gauge works, then find the comparable terminals to the ones on the temp gauge. One will be "accessory", and one will be sender. If one terminal has 2 wires on it, that will be "accessory" because the "sender" stud should never have 2 wires on it.

Or, you can power the gauge up on the bench. Connect the body of the gauge to battery negative with a black clip lead. BRIEFLY connect a red clip lead from battery positive to one of the 2 studs. If the gauge needle heads for Full in a big hurry, disconnect it, and mark that stud "sender". If nothing happens, or if the pointer deflects a tiny amount towards "Empty", mark that stud as "accessory". If nothing happens with either stud, then you have a bad gauge.

Incidentally, battery polarity really doesn't matter with these. The guts of the gauge are really tiny heating elements, and heating elements normally don't care about polarity, and will work fine on AC, too. Ask your kitchen toaster! But we use the "conventional" polarity just to avoid confusion.

If the gauge passed this test, and you've satisfied yourself that you've identified the studs, you may as well connect the sending unit, and bench-test it as well. Hook the terminal of the sender to the "sender" stud on the dash unit, and hook the body of the sender to the body of dash unit. Hook battery negative to the body of the dash unit, and battery positive to the "accessory" stud. Wait for 20 seconds or so. The gauge should show a reading that corresponds to the position of the float. Move the float through its range, and see if the gauge follows. Follow fmarshall's advice to tweak the gauge's end-points if it has problems at the top or bottom of its range. Be sure to allow at least 20 seconds after making any change for the gauge to respond. They ARE slow. Plenty fast enough to follow the change in tank level as fuel is consumed, but frustratingly slow if you are testing or adjusting.

If the gauge pointer goes down a bunch, up a bit, down a bunch, up a bit, and does so repeatedly as the float is lowered, you are seeing the hallmark of a worn-out sending unit. I have posted a repair for this here on at least one occasion, and I'm not going to repeat it again. Use the search function at the top of the page.

Gord Richmond, within Weasel range of the Alberta Badlands