Tango Mike
Well Known Member
I'm the original poster of this thread and wanted to add a follow-up for any RVers who might have to deal with this issue, because it might save someone a lot of time and aggravation.
In retrospect, I think that the fuel sender in the right tank was defective, and that started this whole process. The new sender checked out fine before installation, and I bent and cut the Shepherd's crook exactly as shown in Van's instructions. One item of note is to ensure that the float is positioned in the Shepherd's crook so that it extends forward toward the front of the tank in both the left and right tanks. I don't know if that makes much difference, but the builder hadn't done that with the original sender.
With analog fuel gauges and resistance senders, I'm used to the fact that the fuel quantity indication doesn't initially decrease proportionately with fuel use. In effect, the gauges are lying because they show full tanks when I know that's not the case after burning some fuel. I accept that as a limitation in part because I'm a lot more interested in accuracy at lower fuel levels where it really counts.
My mistake was in assuming that with the G3X digital gauges, I would be able to accurately measure fuel all the way up to full tanks. This may be common knowledge to builders, but I didn't realize that the float reaches its maximum physical travel with a nominal 15 gallons. It isn't capable of measuring any more than that.
So when I began calibrating the tanks, I expected the gauges to show 21, and I fooled myself into thinking that this was possible by not paying careful attention to the fuel calibration procedure, which says to wait 2 minutes after each incremental addition of fuel to let the sensor settle down. I think that with the left tank, I prematurely calibrated the 15-gallon data point, which provided me with a sensor reading between the 10-gallon and 20-gallon reading.
Then on the right tank, when I saw a 15-gallon sensor reading close to what I had used to calibrate the left tank and started to accept the value, I happened to notice that it slowly continued to change until it read exactly the same as the left tank did at full.
The big question: Why is the right tank sensor reading at 15 gallons the same as the left tank reading at 20 gallons?
The answer: Because I unintentionally fooled the system into thinking that there was a sensor reading between 15 and 20.
The new plan: Start again with an empty tank and calibrate in 1-gallon increments up to the point at which the addition of fuel does not result in a change in sensor value, and then cancel that final calibration point so it doesn't cause the G3X gauge to wander at the top end.
With full tanks, the gauges will read about 3/4 full until fuel used causes the quantity to reach the last calibration value. From then to empty, the gauges should be as accurate as the G3X system, a resistance fuel sender, and my care with calibration can make them.
Tosh
In retrospect, I think that the fuel sender in the right tank was defective, and that started this whole process. The new sender checked out fine before installation, and I bent and cut the Shepherd's crook exactly as shown in Van's instructions. One item of note is to ensure that the float is positioned in the Shepherd's crook so that it extends forward toward the front of the tank in both the left and right tanks. I don't know if that makes much difference, but the builder hadn't done that with the original sender.
With analog fuel gauges and resistance senders, I'm used to the fact that the fuel quantity indication doesn't initially decrease proportionately with fuel use. In effect, the gauges are lying because they show full tanks when I know that's not the case after burning some fuel. I accept that as a limitation in part because I'm a lot more interested in accuracy at lower fuel levels where it really counts.
My mistake was in assuming that with the G3X digital gauges, I would be able to accurately measure fuel all the way up to full tanks. This may be common knowledge to builders, but I didn't realize that the float reaches its maximum physical travel with a nominal 15 gallons. It isn't capable of measuring any more than that.
So when I began calibrating the tanks, I expected the gauges to show 21, and I fooled myself into thinking that this was possible by not paying careful attention to the fuel calibration procedure, which says to wait 2 minutes after each incremental addition of fuel to let the sensor settle down. I think that with the left tank, I prematurely calibrated the 15-gallon data point, which provided me with a sensor reading between the 10-gallon and 20-gallon reading.
Then on the right tank, when I saw a 15-gallon sensor reading close to what I had used to calibrate the left tank and started to accept the value, I happened to notice that it slowly continued to change until it read exactly the same as the left tank did at full.
The big question: Why is the right tank sensor reading at 15 gallons the same as the left tank reading at 20 gallons?
The answer: Because I unintentionally fooled the system into thinking that there was a sensor reading between 15 and 20.
The new plan: Start again with an empty tank and calibrate in 1-gallon increments up to the point at which the addition of fuel does not result in a change in sensor value, and then cancel that final calibration point so it doesn't cause the G3X gauge to wander at the top end.
With full tanks, the gauges will read about 3/4 full until fuel used causes the quantity to reach the last calibration value. From then to empty, the gauges should be as accurate as the G3X system, a resistance fuel sender, and my care with calibration can make them.
Tosh
