A while back my buddy Tom asked if I'd like a red cube fuel flow transducer for a bit of "forensic examination", as he had recently replaced one which wasn't working right.
Of course I said "Oh yeahhhh"
The manufacturer has designed the device so it cannot be dismantled for overhaul, at least not without a lot of special effort. First, the allen screws on the cover plates are installed with a thread locker, and being long, small diameter screws, most just twist off before the locker will let go. Second, the electronics are potted. Third, one wire is routed and potted in such a way that removing a cover plate breaks the wire, and it can't be repaired. In this case I didn't care about repair, so brute force methods were employed. Here we go:
There are cover plates on both side of the cube. One side has two stacked plates, and the other has a single plate. The single plate side just covers an potted electronics bay. I did not investigate the electronics.
On the other side, removing the topmost plate exposes the outer face of this machined assembly. The potted groove contain a wire:
The inside face contains a jewel bearing in the center, and an LED (or a receiver, don't know which, see below). You can also see the broken stub of the previously mentioned wire. Removing this assembly from the cube body breaks the wire, which was potted in both the cube and the plate after they were assembled. Clever.
Inside the main body we find another bearing, and a matching LED (or receiver). You can see the (broken) potted wire at about 5 o'clock:
There are two ports for fuel flow, arranged tangent to the circle:
Suspended in the circular cavity between the two bearings is a plastic rotor with 20 teeth and four impellers:
The "teeth" break the light beam between the LED and receiver, 68,000 times per gallon, as the rotor spins in the flow:
When the cube is installed between a boost pump and the engine driven pump, most builders report an excessive fuel flow indication with the boost pump running. I don't think any of us fully understand the flow mechanics with precision, but looking at the rotor, it's not hard to see how a little bit of pulsing in the flow might result in very small reversals of the wheel rotation, which would cause extra counts.
That's all folks.
Of course I said "Oh yeahhhh"
The manufacturer has designed the device so it cannot be dismantled for overhaul, at least not without a lot of special effort. First, the allen screws on the cover plates are installed with a thread locker, and being long, small diameter screws, most just twist off before the locker will let go. Second, the electronics are potted. Third, one wire is routed and potted in such a way that removing a cover plate breaks the wire, and it can't be repaired. In this case I didn't care about repair, so brute force methods were employed. Here we go:
There are cover plates on both side of the cube. One side has two stacked plates, and the other has a single plate. The single plate side just covers an potted electronics bay. I did not investigate the electronics.
On the other side, removing the topmost plate exposes the outer face of this machined assembly. The potted groove contain a wire:
The inside face contains a jewel bearing in the center, and an LED (or a receiver, don't know which, see below). You can also see the broken stub of the previously mentioned wire. Removing this assembly from the cube body breaks the wire, which was potted in both the cube and the plate after they were assembled. Clever.
Inside the main body we find another bearing, and a matching LED (or receiver). You can see the (broken) potted wire at about 5 o'clock:
There are two ports for fuel flow, arranged tangent to the circle:
Suspended in the circular cavity between the two bearings is a plastic rotor with 20 teeth and four impellers:
The "teeth" break the light beam between the LED and receiver, 68,000 times per gallon, as the rotor spins in the flow:
When the cube is installed between a boost pump and the engine driven pump, most builders report an excessive fuel flow indication with the boost pump running. I don't think any of us fully understand the flow mechanics with precision, but looking at the rotor, it's not hard to see how a little bit of pulsing in the flow might result in very small reversals of the wheel rotation, which would cause extra counts.
That's all folks.