Ok, it takes a bit to stump me…..
But this was pretty high up there…
I have the wema tank panel with membrane switches and 7 zones
3 zones are water tanks, 2 are head, and 2 spare channels
My fuel tank is on N2K, and reads out on all my connected gear.
Anyway, one of the tanks would just go from reading correct status to empty.
On the wema panel, a green led flashes when water tanks are empty, and another one flashes when holding tanks are approaching full.
The way that the wema tank sensors work is that there is a cylindrical float running up and down a stainless tube.
Inside the stainless tube are a series of reed switches, each wired to a resistor.
The resistors span from (approx ) 400 to 35 ohms.
Inside the float is a small magnet which closes a reed switch at it’s level, thereby setting the resistance across the sensor output.
There are approximately 10 resistors, spanning about 40 ohms apart in value.
They also have one fixed resistor which is always in circuit, so the other resistors simply form a parallel circuit with it.
So, at the tanks lowest level, you read maximum resistance, and highest water level lowest resistance.
When the guage would go “wonky”, it would basically range the needle to the “E” position exactly.
And here’s what is weird.
If there was an open circuit either in the boat wiring or the sensor, the needle would drop below “E”, to the meter’s mechanical zero.
But, it was going to the Actual “E” position, meaning the float was at the very bottom of its electrical range.
I disconnected the wiring from the wema panel, and sure enough, read about 400 ohms.
So, at this point it appeared that the float was stuck; which I could sorta believe for a waste tank. But not a FW tank.
I called Wema’s representative, and their opinion was that I fried the tank level resistors. I told them that wasn’t possible, as it was deriving it’s current from their own panel meter. They then just told me to replace the sensor… end of story.
So, decided to pull the tank sensor to figure out what was going on.
Sure enough, the float was riding up and down 100% freely, and there was zero gunk in my tank. I’m very meticulous about that.
So, disconnected the sensor, and hooked it up to my trusty Fluke.
400 ohms, irrespective of the position of the float.
Huh??
I usually have a collection of various magnets aboard, and decided to run a test.
I took the magnet, and ran it up and down the tube.
The sensor worked flawlessly from 30 to 300 ohms.
A little more explanation. Is now in order.
The devices used in float switches are reed relays. They are essentially hermetically sealed bi-metallic strips that can be magnetically Saturated. The two leaf contacts are a small distance apart in the relaxed state, but in the presence of a magnetic field become saturated, and the flux lines flow through them, causing them to move towards each other and close
Reed switches are rated in the billions of cycles, and are used in door alarm
Switches, tachometers, etc. because they move such a small distance, and are sealed in a glass tube, they never fail if used within their ratings.
Reed switches are rated by their AT spec, (short for amp turns on a bobbin coil), which is needed to actuate them.
So,,,
What happened to my water tank sensor???
The long and the short of it is that the magnet lost some of its power.
As reed switches take so little magnetic flux to saturate them, designers of mechanisms that employ them get a little lax on their design specs.
Inspecting the tank float, they put a very tiny tubular magnet in it. And flux reduces on the inverse square law, meaning that the closer you get to minimum operating conditions, the more dicey it gets.
Oddly enough, the tank guage did read on a few of the level positions, but this became less and less frequent.
So, it was evident that my sensor was shot.
And with nothing to lose, I drilled a hole in the sensor float for the small disc magnet I was using in my experiment.
Put in a dab of JB Weld marine
And it works perfectly!!!!!
Thought I’d share this in case anyone else runs into a similar tank mystery .
But this was pretty high up there…
I have the wema tank panel with membrane switches and 7 zones
3 zones are water tanks, 2 are head, and 2 spare channels
My fuel tank is on N2K, and reads out on all my connected gear.
Anyway, one of the tanks would just go from reading correct status to empty.
On the wema panel, a green led flashes when water tanks are empty, and another one flashes when holding tanks are approaching full.
The way that the wema tank sensors work is that there is a cylindrical float running up and down a stainless tube.
Inside the stainless tube are a series of reed switches, each wired to a resistor.
The resistors span from (approx ) 400 to 35 ohms.
Inside the float is a small magnet which closes a reed switch at it’s level, thereby setting the resistance across the sensor output.
There are approximately 10 resistors, spanning about 40 ohms apart in value.
They also have one fixed resistor which is always in circuit, so the other resistors simply form a parallel circuit with it.
So, at the tanks lowest level, you read maximum resistance, and highest water level lowest resistance.
When the guage would go “wonky”, it would basically range the needle to the “E” position exactly.
And here’s what is weird.
If there was an open circuit either in the boat wiring or the sensor, the needle would drop below “E”, to the meter’s mechanical zero.
But, it was going to the Actual “E” position, meaning the float was at the very bottom of its electrical range.
I disconnected the wiring from the wema panel, and sure enough, read about 400 ohms.
So, at this point it appeared that the float was stuck; which I could sorta believe for a waste tank. But not a FW tank.
I called Wema’s representative, and their opinion was that I fried the tank level resistors. I told them that wasn’t possible, as it was deriving it’s current from their own panel meter. They then just told me to replace the sensor… end of story.
So, decided to pull the tank sensor to figure out what was going on.
Sure enough, the float was riding up and down 100% freely, and there was zero gunk in my tank. I’m very meticulous about that.
So, disconnected the sensor, and hooked it up to my trusty Fluke.
400 ohms, irrespective of the position of the float.
Huh??
I usually have a collection of various magnets aboard, and decided to run a test.
I took the magnet, and ran it up and down the tube.
The sensor worked flawlessly from 30 to 300 ohms.
A little more explanation. Is now in order.
The devices used in float switches are reed relays. They are essentially hermetically sealed bi-metallic strips that can be magnetically Saturated. The two leaf contacts are a small distance apart in the relaxed state, but in the presence of a magnetic field become saturated, and the flux lines flow through them, causing them to move towards each other and close
Reed switches are rated in the billions of cycles, and are used in door alarm
Switches, tachometers, etc. because they move such a small distance, and are sealed in a glass tube, they never fail if used within their ratings.
Reed switches are rated by their AT spec, (short for amp turns on a bobbin coil), which is needed to actuate them.
So,,,
What happened to my water tank sensor???
The long and the short of it is that the magnet lost some of its power.
As reed switches take so little magnetic flux to saturate them, designers of mechanisms that employ them get a little lax on their design specs.
Inspecting the tank float, they put a very tiny tubular magnet in it. And flux reduces on the inverse square law, meaning that the closer you get to minimum operating conditions, the more dicey it gets.
Oddly enough, the tank guage did read on a few of the level positions, but this became less and less frequent.
So, it was evident that my sensor was shot.
And with nothing to lose, I drilled a hole in the sensor float for the small disc magnet I was using in my experiment.
Put in a dab of JB Weld marine
And it works perfectly!!!!!
Thought I’d share this in case anyone else runs into a similar tank mystery .
