This is a look at a Sky Arrow but the main point here is the earthing and common problems. Firstly we had a look at the aerial and found the ohms to be off the scale. After further investigation the earthing of the antenna was the issue as the securing bolts for the antenna where shorting with the carbon fibre fuselage and its ground plane was disrupted due to this.

Here is a photo of the aerial installation on Johns’ plane. You can see the ground plane under the fuselage

When the aerial was disconnected from the fuselage and only the ground plane the ohms fell into place.

*GOLDEN NUGGET* Carbon fibre is electrically conductive.

I remember a Sky Arrow builder found smoke coming out the rear of the fuselage while he was building his kit as he had accidentally created a 12 volt short. Bit of a wake-up call!

John has kindly given me permission to publish his article he wrote for the LAA. This is a typical example of how errors can be created with incorrect earthing. This subject will be widely discussed in the near future and crowning the earth is vital to keep the Coms path free of RF and the aircraft as a whole. Enjoy Johns article and apply the knowledge to improve your aircraft.

4 June 2017

Cockpit worries:
Fuel disappearance in flight?:

For some years now the fuel gauge on the Sky arrow G-BXGT has presented the pilot with many devious readings. The readings have had no value in determining the real fuel status. Reliance on the time honoured dip stick/stopwatch technique has been utilised. It came to pass that the electronic gauge, fitted by the previous owner in place of the VDU needle type round analogue gauge, became difficult to read anyway . The LCD display was unreadable in bright light conditions and the unit eventually developed an internal fault which caused it to become very smelly and stop working at all. A more up to date version of the gauge was subsequently fitted whose display was a great improvement. However it still gave unreliable readings in flight. Some time was spent attempting a system recalibration and the correlation between the dipstick readings and gauge was erratic to say the least. The fuel sender came under suspicion . The VDU sender was suspect.

The sender utilises two separated vertical strands of resistance wire stretched between two non conducting endplates which are then shorted together by a contact riding on a float. (Resistance is low when tank is full and high when empty).The whole assembly is enclosed in a partially ventilated cylindrical cover which prevents the normally turbulent “sloshing” fuel from having undue effect on the vertical position of the float. It transpired that the shorting contact had become displaced in the moulding in which it was held leading to an unreliable contact between the wires. The fault was cleared and the subsequent static calibration was achieved with reasonable correlation with the dipstick. Not entirely satisfied that the sender unit was playing ball it was decided to change it for a new one. (subsequent forensic microscopic analysis of the old unit showed failings of the shorting bar and geometry of the wires which caused sticky operation of the float. The noted wear may have been caused by the jostling of the float by the disturbance of the return fuel stream impingement.)

The in flight performance was still unreliable but it was noted that the gauge became seriously unreliable toward the lower quantity end of the measurement around 15 Litres. This at the very level, and below, at which one would wish to have accurate data. The readings were wild and of no use at all. What a puzzle.

Whilst sitting in the cockpit one day in the merry merry month of May —(mmm), surmising, and looking at the gauge, the reading was the same as the dipstick. At the lower fuel end of the dip stick. By chance the electric fuel pump was then switched on. The gauge reading lost over 10 Litres !!!!!. After more thought the action of the fuel recirculating system became an immediate suspect.

The return fuel coming into the fuel tank may be upsetting the reading by causing turbulence in the tank which somehow upsets the float in spite of the tubular shield.
The return line and tank fitting wa s inspected and the connection welded into the tank for the return line was a short radial distance from the sender unit, pointing down the side of the sender tube. Furthermore the control jet, which is necessary for the operation of the recirculating system, was part of the welded nipple on the tank top

The jet, it was thought, would increase the velocity of return fuel descending into the tank and stir in air and vapour upsetting the float. Accordingly, to check the theory, a baffle was placed under the jet to disperse the returning fluid. {see picture of tank top plate with baffle}

Sitting in the cockpit the gauge was switched on and left to settle- the pump was then switched on and lo and behold the reading was only 3 litres short. Clearly there had been disturbance from this source. However the reading was still short and this had to be accounted for.
(We really have got into this now!!)

With the pump off again the static reading was still correct. (the engine driven pump would also have produced the return jet effect without the electric pump operating)

It was whilst re-sealing the tank top it was noted that the “earth return” (0 volt line) for the sender unit was also connected to another device THE ELECTRIC PUMP !!.
The pump draws huge peak currents which causes a sizeable volt drop during the peak along the quite long length of the common or 0volt line . It would not necessarily show on any ordinary volt meter because the meter would not react fast enough to show the pulse and in any case they are not calibrated to cope . This means that the gauge electronics was compromised by detecting this volt drop in the return line. The line is over 2.5 Metres long to the common point in the nose cone. The spurious voltage would appear to the electronic gauge as an increase in the detected resistance in addition to the increasing resistance of the sender as the fuel level decreased.

The return line for the sender was separated from the pump line and a dedicated return line introduced down to the common point in the aircraft nosecone. Subsequent operation of the pump now has no effect on the readings and the effect of the return jet has no effect. So prior to landing it is now possible to check the fuel state.

Flight tests now show a gauge which is now certainly more believable and accords with the dipstick . Large changes in attitude do upset things but the readings return sensibly after a short delay . The delay is electronically introduced to aid reading stability.

It all goes to show that a multiplicity of bits of bad design or lack of intersystem operating consideration can add up and lead to bad system performance or no sensible performance at all and huge inconvenience which detract from a good day out. It also costs huge time and effort sorting out the combined systems failures not to say the cost especially if expert professional advice or input has to be utilised. The physical exercise is not inconsiderable, jumping up and down on a Sky Arrow to visit the fuel cap and the tank top several times an hour can bring on faintness and a serious desire to go to the Pub! . To LAA members it’s just a bit of intellectual exercise.

One of the main considerations is the COMMON MODE problems with the wiring which shows up regularly with the advent of ever increasing sophistication in the electronic system department . Fuel Gauges , Instruments, Radios with squawks and whistles induced by the modern Transponders again looking for huge peak currents of short duration and aerials too close for the good of sensitive receivers. Bonding of anodised instrument panel metalwork has also been a recent problem Anodising is a good insulator and one cannot rely on the rivets to do the bonding job alone since the predrilled holes are also anodised !. the cure: spot face some fresh drilled holes, both sides, and install ragged lock washers and nut and bolts.

John Goodale

Please join the Chat Frequency if you would like to discuss issues like this so that the knowledge can be gathered in one place.


No responses yet

Leave a Reply