My Mirage’s gauges show accurate measures only when I have about 15usg left in them (which is 1h of holding time). I use the totaliser and switch based on what was consumed, to keep the difference to less than 10usg. Piper has upgraded the system twice in 20 years, but none of the upgrades have fixed the issue. Very annoying since the gauges in the Meridian work great.
I usually start by writing down my total and split fuel, e.g. 50+50=100. Then I switch at 40+50=90, 40+30=70 etc. I also write down the time and place I do the switch, so if I lose my avionics I can keep track based on time and fuel flow.
A totaliser is usually much better than the fuel gages and, starting from a known initial fuel load, can provide accurate readings for a long time. But there are three major traps:
- Start with wrong values and everything from then on will be dangerously wrong (I posted an account some months ago on this forum how that happened to me once with a very lucky outcome).
- If you have a fuel leak somewhere upstream of the sensor that feeds the totaliser with data, you will never know about it until it is too late.
- And as you already wrote, balancing fuel between tanks can be difficult with a totaliser alone. Timing and accurate records are required which can be a big and unnecessary distraction from flying.
One other problem comes to mind: Your unstable fuel readings can be the result of a worn cable between the fuel sensor and the gage. Parts of these cables run close to the fuel tanks through areas of your wing that may contain fuel vapour. Producing sparks with a worn wire inside there may end your flying career in a flash. I recall one indident of a Piper Navajo where one wing blew up on the runway (luckily for the occupants) because of a worn strobe light cable.
I was going to say what What Next has just posted.
It is one thing to have the “traditional crap GA fuel gauges” but windly fluctuating readings, or unusual readings, could be due to an electrical issue or a fuel system issue, and neither of those can be tolerated.
That is why, on the flight to Lausanne, when the fuel totaliser stopped working, I aborted the takeoff, had the cowlings off and checked for leaks and verified that the fault was in the transducer only.
You never quite know for sure the full extent of any problem, which is why e.g. I think it was patently stupid for BA to fly that 747 all the way to the UK after an engine shredded itself on departure. I know it was IAW their AOC manual but it is still stupid, They could have had a fuel leak and an electrical issue, caused by the disintegrating engine. Look at how much damage that A380 got…
It isn’t hard to make good gauges (the TB20GT has them) but they cost a few k which is perhaps 3x the cost of the crap ones.
Thank you all. It is good to see the spectrum of advice available. It does help to develop good judgement.
You should not switch tanks every 30 minutes but rather after 30 minutes first and then 60…60… Less switching, same result.
In thr SR22 i fly mostly by the fuel totalizer, but of course every now and then i check it against the mech. gauges.
In the Warrior i have a FS-450 and do the same. Very precise.
And: i always depart with full tanks for longer flights.
If it’s your SR22, get it fixed. The Cirrus has very accurate fuel sensors.
The later Cirrus G3 and G5s do have the new accurate gauges but the gauges on earlier models are useless and IMHO it is dangerous to rely on them. I am aware of one fuel exhaustion forced landing where an instructor (not a CSIP) got this wrong and damaged the aircraft quite seriously when landing on a road.
The new gauges can be retrofitted but I believe there have been some problems with the early retrofit installations and it is quite expensive.
The totaliser in the Cirrus is very accurate and, when used in conjunction with it, the Tank 2 app that Roger mentions is one good way of addressing this problem.
Personally, I will always try to fly with full tanks or, if W/B considerations dictate, with not less than tabs or tabs plus a known quantity that I have personally supervised going in. I then plan to land with not less than 20 gallons left although I do occasionally end up a bit under that.
For me the idea of taking unnecessary fuel for a tour of Europe is just a waste of payload and performance so I travel with the optimum fuel for the trip.
The DR400 is very good for fuel managment because it has four tanks three of these have low fuel indication, the two wing tanks have a (safe) hour at 75% power with a 5 min warning of becoming empty, the centre tank has a 20 min warning of empty with the AUX tank emptying into the centre tank.
Good fuel planing has the tanks being used in the order Centre, AUX, left wing, the right wing carrying the diversion and contingency fuel.
With all the hard fuel management points at which there is no doubt what fuel quantity is on board there is little doubt when the engine will stop, I am about to fit a Shadin Miniflo so that I can get a better idea of the finer points of fuel managment and carry less fuel on longer trips.
For me the idea of taking unnecessary fuel for a tour of Europe is just a waste of payload and performance so I travel with the optimum fuel for the trip.
Payload yes, performance no. In our aircraft, weight (MTOM vs light) has virtually no impact on cruise speed and fuel burn.
When taking off from airports at say 6000 ft DA you may need to carefully consider the effect of weight on takeoff performance and obstacle clearance, assuming you don’t want to die. In the subsequent reduced rate climb you may also have an issue with clearing still higher terrain, deviating, circling etc, but it’s true that the additional fuel burn integrated over the extended climb period does not greatly affect cruise range in a subsequent long distance flight.
Not everybody lives where the land is flat for miles in every direction 
On some shorter local flights in mountainous terrain, time en route is more affected by rate of climb than by a substantial difference in speed. A 120 kt Pitts biplane can seem very ‘fast’… in the sense of getting there first.
That aside, I stick my tanks on one plane, with great accuracy, and then fly time on tanks. On the other I stick the tanks when possible (only possible if they are relatively full), now use an accurate fuel flow totalizer, and also fly time on tanks. On neither plane do I pay great attention to the gauges, and would only do so if I noticed them reading empty unexpectedly.
The gauges in my old Cessna 150 were quite unreliable. They were ok in the lower half but basically stayed on full too long and then went to half real quick.
On the Mooney, I was quite astonished to find out that the gauges actually work pretty well. The only issue I have with those is that they are in lb while the quantity indication on the fuel tank caps are in USG and theShadin fuel flow/ totaliser is in liters. All from before my time and not much can be done about it. I use the recommended fuel schedule which is run first hour on one tank, then switch to the other until it runs dry and the go back to the first tank.
