Fuel Level Sender

Status
Not open for further replies.
Joined
May 10, 2005
Messages
2,731
Location
Toronto, Canada
Never seen this type before, maybe others are familiar with it.

I just bought this as replacement for a 2006 Freightliner M2. The rod is non-magnetic stainless steel. The float rides on the rod and is magnetic, it will attract a screwdriver held next to it. On my analogue multimeter, the two leads have 28 ohms across them when float is all the way up (full tank) and the resistance increases linearly to 255 ohms when float is moved down. so it looks like there is a potentiometer housed inside the rod and the float moves the wiper.

The senders I have encountered before have the pot exposed to fuel.

The two leads coming from the instrument panel and going to the sender have 12v across them when not connected to the sender. When I hook up my analogue multimeter on the 250 ma DC scale across the leads from the gauge, the short-circuit current is only 30 ma. So the source impedance is 400 ohms, considerably larger than 255 ohms. The circuitry inside the instrument panel processes the signal from the sender and drives the stepper motor for the gauge.

000_0180.jpg
 
I've seen something similar in aircraft. It uses capacitance to determine the liquid level.
 
Maybe this one uses capacitance too. The varying resistance when the float is moved however suggests a potentiometer inside the rod.

I still do not completely understand this sender and the fuel gauge. The Freightliner FSM offers no info on the workings of the sender.

I have not installed the sender yet. Right now the drivers of the truck just look into the tank to determine how much fuel is left. When I install it I will hook up a scope to see what kind of voltage is applied to the sender. If it is straight DC then obviously it does not use capacitance.
 
Last edited:
What you are most likely dealing with internally is a series of magnetic micro switches that add or subtract resistance (or capacitance) as each switch closes or opens. They are used in a lot of rugged equipment where total isolation from the liquid fuel is desirable. Think Tanks (as in Military) Tanks (as in Acid) and Tanks (as in tractor trailer tanker trucks).

Rated explosion proof in most instances.

You might be able to confirm this action by connecting the meter and watching as each switch opens or closes. I'll bet it's not a smooth reaction, but progresses in steps as the magnet travels along the rod.
 
Thanks, I will test that out.

What was really puzzling during my initial testing was, after seeing that the short circuit current was 30 ma, I shorted the two wires coming from the instrument panel and there was no reaction from the gauge. After the initial sweep at KOEO, the needle stayed at empty with no movement at all whether the wires were shorted together or not. Whether it was zero ohms or infinite ohms the needle stayed put. We have another truck, a 2004 very similar to this one and I ran two jumpers from the sender of that truck to the instrument panel of the subject truck and the gauge read 3/4 tank, which was the level in the tank of the 2004 truck

I theorized that the signal processing circuitry in the instrument panel expects to see the resistance to be within a certain range, say 20 ohms to 270 ohms and that is why there was no reaction from the gauge to zero or infinite ohms.
 
Yes, the sending unit most likely forms one leg of a Whetstone Bridge circuit. Or something similar. Hard to tell these days with digital devices.

A little more info came to me in my sleep last night. The last one of these I've seen opened (from an M1 tank) had all of the switches mounted on a printed circuit board. Long and narrow to fit into the stainless tube. Each switch had a resistor in it's particular circuit. If all switches were open, the full resistance of the assembly was seen at the leads. As the magnet rose along the rod, each switch would close, thereby taking all the resistors below it out of the circuit. That way, the higher the fuel level, the lower the resistance of the assembly.

Keep in mind, all of the resistors MAY be the same value, but depending on the shape oof the tank, they may vary that to compensate. In a straight walled vertical tank, that would be appropriate, (all the same) but otherwise fuel content would vary by the shape. In other words, fuel may be at half LEVEL, but not necessarily at half VOLUME.

Tanks, John
lol.gif
 
You were right!

Hooked up a DMM and slowly slid the float along the rod and the resistance varied in 18 discrete steps on the 16 in long rod, so the switches were placed just under an inch apart. Here are the values in ohms, rounded off to the nearest whole number
255, 240, 225, 198, 176, 154, 141, 128, 116, 103, 94, 84, 75, 65, 56, 46, 37, 28.

An analogue multimeter provided an interesting display, with the needle jumping to the next value every so often as the float was slid along.

The 27 ohm jump at the 225 ohm mark is an anomaly. I double checked the resistances and the big jump is there. The tank is a plain rectangular cube (wikipedia calls it cuboid), so it is not the shape of the tank dictating the big jump. Perhaps the switches were not evenly spaced out at that point due to design constraints.

I will cut open the old sender when I take it off. If I do it soon enough I will post a picture.
 
Quote
Perhaps the switches were not evenly spaced out at that point due to design constraints.


Possible bad contact or a resistor drifted off value, it happens.
At least now you know what you are dealing with.
 
Just talked with my Brother. He runs the maintenance shop for Freightliner in Rocky Mount NC. He said that empty should be 33 ohms and full 300 ohms.
He also said that on an M2, everything is J1939 datalink. So you MUST let the truck go through the sleep cycle by closing all doors and windows for 3 minutes. The ICU will then get the wake up signal when the drivers door opens. This should clear the active codes. They won't clear without the ICU going to sleep.

He also said the bulkhead module is the Master, the chassis module the slave, and that all switches are smart switches with binary codes.
Hope this helps.
 
Yes, the vast majority of the electrical system uses J1939, Freighliner calls it Multiplexing.

I don't think the fuel level signal is multiplexed. I don't remember seeing it among the inputs or outputs of the BHM (BulkHead Module) or CHM(Chassis Module), nor do I remember seeing it in the data stream. The open circuit in the fuel level sender did not set a code.

Most of the switches are smart switches but there are some exceptions. Systems that have not been integrated into the Multiplexing have to use the old fashioned on-off swtiches. In our situation, with moving company trucks, multiplexing does not have a provision for the box interior light. When Upfitters install a box on the back of the truck, they have to install a non-smart switch on the dash and run a wire all the way from the dash switch to the lamps on the ceiling of the box. I expect snow plow switches, dump box switches etc that are not installed at the factory have to be non-smart switches.
 
Last edited:
Here are two photos of the failed sender.
As you can see, the old one was a potentiometer type sender. Strangely enough the failure was not at the actual pot but at the top where the connector is.. I have not broken the connector apart to see the actual failure. Connector pins look fine.
The new micro switch type sender was surprisingly inexpensive, at $80 Cdn.

IMG_0173.JPG


IMG_0172.JPG
 
Last edited:
Status
Not open for further replies.
Back
Top