New Boat Fuel Gauge Problem

[Jim Cooper]

Purchased a new boat with dual tanks & one fuel gauge; wiring was unfinished. I purchased an on-off-on single pole, double throw, lighted rocker switch to be able to read each fuel tank on the one gauge. The switch has 4 terminals, one each for the sender wires, one for the gauge feed and one for the light neg. ground (the light positive is supplied from the gauge feed) Connections are straight forward so I know I have the wires connected to each terminal correctly.


Problem: when completely wired, I don't believe I am seeing true readings so I tried a couple scenarios and came up with the following -



Scenario 1 eliminated the switch completely, connected each sender wire independently to the gauge feed. READINGS = TANK1 7/8; TANK2 3/8


Scenario 2 installed switch, connected gauge feed and only one sender independently, did not connect light ground. READINGS = TANK1 7/8; TANK2 3/8


Scenario 3 installed switch, connected gauge feed and both sender wires, did not connect light ground. READINGS = TANK1 FULL; TANK2 3/4


Scenario 4 installed switch, connected gauge feed and both sender wires, light ground wire connected. READINGS = TANK1 PAST FULL; TANK2 7/8


My observations are the reading through the switch only matched the reading without the use of the switch when only one sender wire was connected. My other observation is the readings appear to increase each time an additional wire is connected to the switch.


Any ideas what is going on? The switch manufacture already replaced the switch once and the second device produces the same results.

 

[Bmarler]

it's likely the light is adding resistance to the sender circuit

 

[Bmarler]

should have read more carefully.
how exactly are you connecting the sender wiring to the switch?

 

[Jim Cooper]

backside of the switch are three terminals stacked and one terminal off to the side for the light's ground wire. The manufacturer's schematic shows the three stacked terminals as the middle terminal being the power feed and the top/bottom terminals being the outputs. I have the fuel gauge's power feed going to the middle terminal and the two sender wires connected to the two output terminals. It is just a simple on-off-on switch commonly referred to as an A-B switch.

 

[Bmarler]

backside of the switch are three terminals stacked and one terminal off to the side for the light's ground wire. The manufacturer's schematic shows the three stacked terminals as the middle terminal being the power feed and the top/bottom terminals being the outputs. I have the fuel gauge's power feed going to the middle terminal and the two sender wires connected to the two output terminals. It is just a simple on-off-on switch commonly referred to as an A-B switch.

i think i'd start by measuring the voltage with a good voltmeter to see if the return from the sender wires is different with the switch out of the loop vs. in the loop.
alternatively, you could check ohms if the power is off and the wires isolated.
i think i'd prefer the voltage check though as the gauge is really just a volt meter scaled for the senders.
are these single wire senders?

 

[Jim Cooper]

yes, single wire sender to console, each sender has ground wire also which, I have running to the closest ground block.

 

[High Wire]

Replace the switch with an unlighted spdt.

 

[jgwinks]

As High Wire said. Somehow the switch is changing the resistance on the sender wires. Sounds like you tried manually swapping the sender wires on the gauge and it appeared to read correctly. This is what you want the switch to do so a single pole double throw world work and less chance of it throwing the gauge off. Try running just the senders through the switch, power and light connected directly, and see if that helps. Having the gauge powered and lit all the time shouldn't be a problem. And "stick" your tanks if you can so you know what's really in there. Fuel gauges are notoriously inaccurate, but you can bend thier float arms to get them close, and they should stay as close thereafter. I take it there are no sight tubes?

 

[jgwinks]

Rereading your description of how you connected the switch, I don't think you did it right. The switch is designed to power two devices from a common source but should work in your application as well. Disregard the schematic, it's for a power hook up, not senders. There will be no power connection on the switch in your application. The center terminal is the "common" and goes to the gauge. The upper and lower terminals go to the respective senders. The side terminal is the ground for the switch's internal light. Try it with and without the ground, it could be the problem but I doubt it. If you want the switch to also the power the gauge on and off, you need a DOUBLE throw double pole switch. It would have 6 terminals on the back in three pairs. One side would be power, the other would be senders. I don't recommend doing that, having power and data through the same switch might cause interference.

 

[Jim Cooper]

Rereading your description of how you connected the switch, I don't think you did it right. The switch is designed to power two devices from a common source but should work in your application as well. Disregard the schematic, it's for a power hook up, not senders. There will be no power connection on the switch in your application. The center terminal is the "common" and goes to the gauge. The upper and lower terminals go to the respective senders. The side terminal is the ground for the switch's internal light. Try it with and without the ground, it could be the problem but I doubt it. If you want the switch to also the power the gauge on and off, you need a DOUBLE throw double pole switch. It would have 6 terminals on the back in three pairs. One side would be power, the other would be senders. I don't recommend doing that, having power and data through the same switch might cause interference.

So I do believe I have the wiring correct. I have the two sender unit wires connected to the upper & lower terminals and the gauge wire connected to the middel terminal. I probably confused some folks by referring this was a power line. The gauge feed to the middel terminal is actually providing voltage and it should, right? Afterall, it is what the sending unit is "resisting".

In scenario 3 & 4 I had the light's ground wire disconnected [3] and then connected [4]. Being connected had an effect on the gauge reading but so did having both sender wires connected before I even introduced the light's ground wire effect.

 

[jgwinks]

No, is not DC+ voltage. It's actually ground but with variable resistance compared to the actual ground. DC + and - are applied to the gauge by linking to the + and - terminals on the other gauges. Since the gauge is meant to be in a power circuit, what I think is happening is the switch's ground wire is shorting the gauge wire directly to panel ground, giving it 0 resistance and changing the reading. As designed the switch's internal light gets it's power from the center terminal. In your application this is shorting the gauge wire to ground. Leave the ground terminal disconnected and it should work.

 

[jgwinks]

Look at the senders on the tanks. They should have two terminals, one is the sender wire going to the gauge. The other is system ground. There is no DC+ connection on the sender, you really do not want to be putting DC+ into a fuel tank! If the ground terminal was inadvertantly connected to DC+ I don't know what would happen but it might do some crazy things. There should never be DC+ on the sender wire, at least on any sender I've seen.

 

[jgwinks]

Another thought, are you absolutely sure the wires you are working with are in fact from the senders? Since the wiring was incomplete when you got the boat I wouldn't make any assumptions.

 

[Jim Cooper]

Another thought, are you absolutely sure the wires you are working with are in fact from the senders? Since the wiring was incomplete when you got the boat I wouldn't make any assumptions.

Yes, I had traced that early on! Based on some of your other information, I believe I am going to substitute this lighted switch for a non lighted switch and see what readings I get.

 

[Jim Cooper]

No, is not DC+ voltage. It's actually ground but with variable resistance compared to the actual ground. DC + and - are applied to the gauge by linking to the + and - terminals on the other gauges. Since the gauge is meant to be in a power circuit, what I think is happening is the switch's ground wire is shorting the gauge wire directly to panel ground, giving it 0 resistance and changing the reading. As designed the switch's internal light gets it's power from the center terminal. In your application this is shorting the gauge wire to ground. Leave the ground terminal disconnected and it should work.

Leave the ground terminal disconnected and it should work --- refer to my scenario 3, the bad readings actually start when I connect both sender unit wires to the switch without the light's ground being connected. Scenario 4 I added the ground wire into the mix and had an even greater false reading. I am going to try a non-lighted switch to see if the light, grounded or not, is the culprit.

 

[jgwinks]

I'm still troubled by your statement that the guage to center terminal is supplying voltage. Is there really DC+ on that wire? There should not be. Is it connected to the "S" terminal on the guage?

 

[Bmarler]

I'm still troubled by your statement that the guage to center terminal is supplying voltage. Is there really DC+ on that wire? There should not be. Is it connected to the "S" terminal on the guage?

Even though it’s not a dc positive it can appear to be when it’s not connected to ground. The sender is the dc negative in the circuit.

 

[Jim Cooper]

yes, the connection is between the center terminal and the S terminal post on the gauge. I'll run by the boat this evening and put a volt meter to the wire to verify whether there is DC + running through it.