Welcome to our resource for MG Car Information.
MG MGA - Fuel gauge sender calibration.
|I am just about to try to calibrate my fuel gauge/sender to read a more accurate zero, after running out returning from the Le Mans Classic weekend, fortunately right alongside a petrol station in the French countryside! Could someone who has done this please inform me - if I get the tank fuel level exactly at the pickup pipe level, ie. just running out, can the sender unit be un-bolted without fuel pouring out, ie. is it positioned high enough up the side wall of the tank. |
|You can do that Bruce, I have done the same thing.|
Whether you should do this is another matter. (health and safety etc :-) I would drain the tank.
At the very least, stick some Duck tape over the sender unit hole whilst it is out, switch the ign off and either do it outside or in a very well ventilated space.
I have just bought a new sender unit from NTG as I was so fed up of my own optimistic fuel gauge which never showed below half full.
I have checked the resistance of the coil and amazingly, it is exactly as Barney says it should be (from memory 70 ohms)
I plan to test wire it up to the gauge in the next couple of days and I will let you know if it works.
|The sender unit is not adjustable, except to rewind the resistance wire or to fix an open circuit. Do not remove it from the tank unless you first determine it is faulty.|
Sender resistance should run from 0-ohms empty to 70-ohms full. If you just ran out of fuel, it should be 0-ohms (float arm sitting on bottom of tank).
Jack up, remove RR wheel, disconnect wire from sender unit, and measure resistance from the connector stud to ground on the sender housing. If it is very near zero ohms when empty, the sender is likely okay, and the gauge should go to "E". If you ground the sender wire on the tank (or on the sender housing), the gauge should go to "E". If the doesn't work, then ground the sender wire on the chassis frame, and the gauge should go to "E".
The sender unit is grounded on the tank, and the tank should be grounded to the chassis. The tank is (originally) mounted in rubber lined straps and the filler pipe is connected with a rubber hose. The tank was originally grounded through the steel pipe running from tank to fuel pump (and the pump has a grounding wire. If the metal pipe has been cut and reconnected with a rubber hose, then the fuel tank would not be properly grounded. In that case, add a new grounding wire from one of the sender unit screws to a nearby tab on the frame.
If you ground the sender wire to the frame, and the gauge still does not go to "E", then check/clean the snap connector in the engine bay near the starter switch. Look for the Green wire with Black stripe. The electrical diagram for the fuel gauge is here:
Note that the fuel gauge must also be grounded. There should be multiple black wires from the harness connected to the mounting bracket on the gauge.
|Thanks Colyn and Barney. However, I fear I may have caused some confusion with my question! Let me re-phrase it. |
The sender unit is a good old original unit in perfect working order. I recall measuring its resistance some time ago and it was to spec. The "re-calibration" I was referring to was not by altering or re-winding the resistance wire ( even as an electrical engineer I would not wish to tempt fate by such a move) but to bend the float arm to get it to read zero more accurately when fuel level is just at the level of the pickup pipe. I think that I had previously, foolishly, made the gauge read zero with the tank drained completely by the bottom drain hole rather than to read zero with fuel at the level of the pickup pipe, ie. at a real practical zero!
So if that is clear (!) my question is just to know if the fuel level will be below the sender unit when just at the pickup pipe, so petrol does not flow out when the sender is removed. Then I can do a trial and error adjustment of the arm to read zero, or thereabouts. Currently it reads 1/4 tank full when it runs out!
|Bruce, I have three spare sender units, two with the original metal float and a repro. plastic float type.|
With each of the floats in the lower limit position the distances from the bottom of the tank hole to the top of the float are :- 23mm, 24mm and 27mm. So the corresponding fuel level is an inch or more below the tank hole when the sender units are bottomed.
Incidentally, I had occasion to drain my tank some time ago and as I didn't fancy unscrewing the tank plug I uncoupled the fuel pipe near the rear carb. and used the pump to empty the tank. It's surprising how long it takes especially when it has to be watched carefully, around ten minutes per gallon pumped out!
When my tank was empty the gauge read below the E and it took about 1/2 to 3/4 gallon to get it to the E mark which I think is a good safety margin. I wouldn't like the gauge to read empty when the tank was empty. I think most cars have some safety margin..........................Mike
|The simple answer to your original question is, "Yes".|
Just trying to suggest that if it ain't broke don't try to fix it. Always do the diagnostic work first. What is the resistance of your sender unit when the tank is empty?
If you previously had the gauge reading zero with empty tank, then the sender must have been registering 0-ohms. When fuel level is low enough to run out of gas, the float will not be floating. It would be sitting on bottom of the tank with not enough buoyancy to make it lift any at all. If it was zero ohms once, it will still be zero ohms. If the arm is formed so as to hold the float higher so it doesn't touch the bottom of tank, it will still be sitting at bottom of stroke and 0-ohms.
Measure the sender unit resistance before you take it apart. If the arm was formed to position the float BELOW bottom of tank, then it would never get to bottom of stroke, and you might have more then 0-ohms when empty (and it could help to bend the arm upward). But if you find 0-ohms, then that is not your problem, and nothing you do to the sender unit would get the gauge reading closer to "E". When sender resistance is 0-ohms, and the gauge is not reading "E", that's a problem with the gauge or wiring, not the sender unit.
|Thanks Mike and Barney|
You both confirm that fuel is below the bottom of the tank hole when the car runs out of petrol, so I can remove/play about with the sender if needed, with fuel at that level without spillage. That's reassuring.
Measured a few things as Barney suggested.
a) sender connected and tank absolutely drained dry (ie. below where it would normally stop running) - gauge reads exactly on the zero/empty mark.
b) disconnected "positive" wire from sender and measured 11.4 ohms from its terminal to ground. Somehow the arm is being held up?
c) earthed positive wire to ground and gauge reads a little under zero/empty (pointer goes to midway between the empty mark and the "E" letter, ie. a small amount below the empty mark).
d) open circuited the "positive" wire (ie. infinite resistance) and gauge reads full (above the "F" letter and on the end stop).
e) from my running out of fuel experience in France, the car stopped with the gauge reading just under a quarter full.
I shall leave it for now and sleep on it! Seems to me I need to take a look at the sender and make sure it can get down to zero ohms and isn't stuck part way down, or some other funny thing!
Any other thoughts welcome. Start again tomorrow. Need to get the car mobile for Friday Goodwood Revival!
|About five years ago Moss Motors produced a new sender unit that first appeared to be quite nice, except it had the wrong resistance range and wrong mechanical range. See here: http://mgaguru.com/mgtech/electric/fg115.htm|
I believe they have more recently corrected the resistance range, because I installed one recently and it works. The mechanical range (angle of the arm) might still be wrong. If it lands on bottom of the tank before hitting bottom of travel of the sender, then it might show the 11 ohms you see when empty. That might be corrected by bending the arm upward a bit so bottom of travel will coincide with bottom of tank.
|Thanks Barney. Agree with the diagnosis. Will be trying an upwards bend later today and let you know the result. |
|"Confused of Shoreham" reporting back! Took sender out - model no. FT5300/20 - which is an original as per Barney's Guru site. All in nice condition, good free movement through full travel, but resistance minimum now down to 1 ohm, as it should be, not the 11 ohm as I previously reported. Maximum resistance 80 ohm. In other words, seems nothing wrong with it!|
Cleaned it up, re-used the nice Peter Edney rubber gasket, float not bottoming out in the bottom of the tank as originally suspected, connected up and now indicating, correctly, a little below the gauge empty mark. Lowered fuel level in carb float chambers so that the pump would be looking, unsuccessfully, for fuel in the fully drained tank. Slowly added a couple of litres or so of fuel to the tank and the pump picked up and stopped, having reached the minimum fuel level. The gauge now reads a perfect, microscopic amount below the empty mark.
So, all good, nothing needed adjusting and my original 1/4 full indication when running out of fuel appears to have gone away. I can but run it and see how it goes in future. Can only assume the float arm must have stuck a little above minimum level but the reason for that is unclear.
|Just to update you all on the new tank sender unit that I bought from NTG.|
I test fitted it yesterday whilst I had the car up on jacks fitting new rear brake drums and shoes.
I connected the unit to the wire to the gauge and earthed the sender unit body.
It seemed to be a big improvement on my previous unit which showed full when the tank was full but then rapidly dropped to the half full position and stayed there even if the fuel ran out.
When I lifted the float fully, the needle on the gauge swung over to approximately the 7/8th full position.
When I fully lowered the float the gauge showed empty as it should.
So I decided to fit the unit using the cork gasket that came with it and a smear of Hylomar.
So far it hasnt leaked and the gauge shows full with the tank filled up.
Time will tell if the gauge shows empty when the tank is actually empty but I am reasonably sure that it will.
However,I will probably still check my odometer readings after filling the tank until I am more confident with the gauge readings.
|Bruce, I hope you don't mind my final update about my new sender unit.(from NTG)|
I have just done over 500 miles in the car since I fitted the new unit and it is almost perfect.
The gauge still doesn't quite show full when the tank is actually full but I ran it down to show empty and there was only about 4 pints in the tank.
The best part is that there is actually 2 gallons in the tank when the gauge shows 1/4 full and this has proved to be consistent too.
So at last the fuel gauge is a useful instrument and I am delighted with it.
Did you manage to fix your set up Bruce?
This thread was discussed between 06/09/2016 and 01/10/2016
MG MGA index
This thread is from the archive. The Live MG MGA BBS is active now.