I put the picture first here, because I spent rather a lot of time measuring mechanical bits and electrical resistance and making this drawing. This is a drawing of the Moss Motors 360-280 fuel sender unit for MGA. This part was previously reported as defective due primarily to loose coil windings that would move around and give erratic gauge readings. Moss has apparently gone to some effort to improve the quality of this unit, and they sent me one of the new ones for my consideration. Without posting lots of pictures, I can say that the quality is indeed greatly improved in the form of electrical coil, wiper arms, and internal grounding. In short, it now works well electrically.

Unfortunately this unit now (or still) has the wrong range of mechanical motion, and the wrong range of electrical resistance. The drawing shows that the bottom limit of motion would be below bottom of the tank, so when the float sits on bottom it is not at the low limit of resistance. Additionally the lowest resistance I could measure was 14 ohms when it should go to zero ohms. With the float on bottom of the tank it would read about 20 ohms (when it should be 0 ohms). At the upper limit of travel it only reaches to about the 3/4 full fuel level, and it gives 97 ohms resistance at that point. The fuel gauge wants to see about 70 ohms for a full tank reading. This unit would read about 76 ohms for a half full tank.

Final result of all this would be gauge reading Full (or pegged past the "F" mark) until the fuel level gets down to about half a tank. On the lower end when the float hits bottom of tank the gauge would read about 1/4, and would never go any lower. So if you get 25 mpg fuel economy, and can go 300 miles on 12 gallons of fuel, here is what the gauge would do. For the first 75 miles the gauge would be pegged above the Full mark. For the second 75 miles the gauge might move slowly from the top peg down to the Full mark. For the third 75 miles that gauge might move from "F" to "3/4". After 225 miles, for the next 65 miles or so the gauge could move from 3/4 to 1/4 indication when it hits bottom and stops moving. Then if you were very observant you have about 10 miles to find a gas station before you run dry with 1/4 still showing on the gauge.

So between half tank and quarter tank it goes from F to 3/4 on the gauge. Then on the short end it would go from 3/4 reading to sucking air at 1/4 reading in about 75 miles. From recent reports on various BBS, a lot of people have apparently experienced similar strange fuel gauge readings. This may explain why.

Mechanical motion of the arm might be improved some by extending the length about 15% and bending the arm more upward to get it to travel fully from top to bottom of tank while matching the full travel of the wiper contacts (shown by the red lines in the drawing). Even then it may come below the Full mark at about 5/8 tank, and may bottom out just above 1/8 tank indication with about 10 miles to go. By fiddling with gauge calibration adjustments (non-original adjustment) you might get the gauge to fall below the Full mark just a little earlier, perhaps after 100 miles of travel, and you may be able to get it to read a little closer to the Empty mark when all the way down.

I have an ingrown philosophy telling me not to modify a perfectly good part (the gauge) in order to accommodate a bad part (out of range sender unit). If the gauge works properly using test resistors, I wouldn't fiddle with it. I still have some hope that one day someone will produce a replacement sender unit that works the same as the original one.

On 11/17/2013, Bill Love in Farmingdale (Long Island), NY wrote:
"My Moss replacement has all the quirks that you describe".  I was recently able to pick up a used Smith sender unit. It works, but was out of a MG T-series. The part number is FT5300/15 while the MGA version is FT5300/20. It appears the only difference is how the rod is bent".

So two years on Moss is still selling this one as-is, knowing full well it does not work properly. If you buy one of these, and you actually figure out how to use it, please let me know how you did it.

Addendum August 8, 2014:
There may now have been some imrovement in this part. Or perhaps Moss Motors UK is now selling a different part than Moss Motors USA

On 8/6/2014, Mike Card in Surrey, UK, wrote:
"I purchased a replacement unit from Moss Motors UK on 2014-04-25, part number AHH5114. Resistance measured 0-ohm and 80-ohm. The float arm needed to be bent upward slightly so that the float would reach the zero stop when the tank is empty. My fuel gauge has been calibrated according to your page FG102 by a friend who has made up a calibration box.
"The Moss sender seems to work well and I have determined the following for my car by careful filling at the pump, starting with a tank less than 1/4 full and taking readings at 1/4, 1/2, 3/4, and F, and again when the pump cuts off when the tank is full (45.4 litre). With my tweaking of the float arm not recorded, they won't be transferable, but shows that the current Moss unit is acceptable. The following are based on the average of two sets of readings. They may be accurate to two significant figures, maybe one.
1/4 - 3.9 litres, 0.9 imperial gallon
1/2 - 13.3 litres, 2.9 imperial gallons
3/4 - 30.3 litres, 6.7 imperial gallons
Full - 42.9 litres, 9.4 imperial gallons
Not a linear result, but at least I have an operating gauge now and know what the gauge markings mean". -- Mike