Ownership / Fuel pump

The fuel pump

The Stag has an SU electric fuel pump and no mechanical one. It ticks when it is working, the rate of the ticking is a diagnosis rather than a noise, it is polarity sensitive, and on many cars there is a cut-out switch on the bulkhead that will stop the engine without warning.

What the Stag actually has

An SU electric fuel pump, mounted on the body rather than on the engine. There is no engine driven mechanical pump on a Stag, and none appears anywhere in the fuel system section of the parts catalogue.

The pump sits on a bracket, insulated from the body by three rubber anti vibration bobbins, part UKC2451. The original brackets, parts 150866 and 151062, are no longer available new and are listed as used items only, which is worth knowing before a restoration reaches that point.

The current supplied part is AZX1311, an SU points type pump to original specification, dual polarity, also listed by several suppliers against the cross reference QFP172E. Rimmer also list the alternative electronic pump AZX1311ALT as fitting the Stag from February 1969 to December 1979, and the same unit serves the MGB, MGC, certain Daimlers and Jaguars, the Jensen-Healey, Rover and Wolseley. The Stag does not have an exotic pump. It has the standard British sports car pump of its era, which is good news for availability.

How it works, and why it ticks

The description below follows Dave DuBois's SU Fuel Pumps 101, updated February 2014, which is the standard published account, and agrees with SU's own description of the pump as an electromagnetic diaphragm type.

The SU is an impulse pump rather than a continuous one. Switch on the ignition and current flows through the points and the solenoid coil. The energised coil pulls an iron disc, and the diaphragm attached to it, towards the coil. That movement creates a vacuum which draws fuel from the tank through a check valve into the pump body.

As the diaphragm moves it pushes a shaft connected to the lower points carrier. At the end of its travel the carrier throws over, snapping the points open. Current stops, the magnetic field collapses, and a spring pushes the diaphragm back the other way, delivering the fuel through the outlet valve. At the end of that return the carrier throws over again, the points close, and the cycle repeats.

That throw over is the tick. SU describe the characteristic tick as the sound of the contact points opening. It is not a fault and it is not the pump struggling. It is the pump working.

DuBois adds one detail that is worth knowing and that the other published accounts do not give: the pump's pressure is established by the strength of the volute spring which sits between the iron disc on top of the diaphragm and the bottom of the coil. If that is right, and the archive has it from one source only, then you cannot raise a tired pump's pressure by giving it more voltage, and a pump delivering low pressure has a mechanical problem rather than an electrical one.

What the ticking is telling you

The pump only cycles when fuel actually leaves the pump body, which makes the rate of ticking a genuine diagnosis rather than background noise. There are three things to listen for.

  • Fast at first, then slowing to an occasional tick. This is correct. The pump is filling the float chambers, then holding pressure.
  • Fast and never slowing. The pump is not filling anything. Look for a leak, a split diaphragm, or a float valve stuck open.
  • Slow from the start, or no tick at all. SU state plainly that poor earthing causes slow ticking, low output and rapid point wear. On a pump deliberately insulated from the body by three rubber bobbins, a poor earth is a more likely fault than it ought to be. No tick at all means no power, no earth, or seized points.

The polarity trap

Rimmer's own listings distinguish between these pumps by polarity: the points type AZX1311 is described as dual polarity, while the electronic AZX1311EL, built to the original design, is negative earth only. That distinction is catalogued fact and worth respecting.

The explanation offered for it comes from DuBois, and the archive has it from that one source. Early SU pumps suppressed points arcing with a swamping resistor, and later ones with a 0.47 microfarad capacitor which, despite appearances, is not polarised. But the AUF 300 and AZX 1300 series replaced that capacitor with a diode, and a diode is polarity sensitive. The Stag's AZX1311 belongs to that family. Other published accounts of SU pumps mention diodes, varistors and zener diodes among the methods used to protect the points, which is consistent with DuBois without confirming the specific claim.

The practical point stands whichever explanation is correct: if a car has been rewired, converted, or had a pump fitted from a parts bin, polarity is worth confirming rather than assuming. A pump connected the wrong way round will not run backwards. It will not run.

DuBois also records that the same series uses a different check valve from the older pumps, a plastic sheet in a valve assembly rather than the simple brass disc of the L, HP and LCS types, which matters when ordering repair parts because the two are not interchangeable.

Points or electronic

The original points pump fails in a predictable way. DuBois identifies two kinds of wear: mechanical, which he describes as slight, and electrical arcing at the points, which he describes as major. In other words the points do not usually wear out mechanically. They burn. SU's own guidance that a poor earth causes rapid point wear points the same way.

That is what the electronic conversion addresses. It replaces the points with a Hall effect circuit which does the same switching without contacts, and it keeps the characteristic ticking, so the diagnostic value of the sound described above is not lost.

  • AZX1311, points type, original specification, dual polarity. The correct choice for a car being kept original.
  • AZX1311EL, electronic, built to the original design, negative earth. Looks right, no points.
  • AZX1311CONV, a conversion kit containing everything needed to convert an existing points actuated SU pump. The option for an owner who would rather keep the pump that is on the car.

The archive takes no view on which is correct for any given car. An owner chasing concours originality and an owner who wants to drive to Scotland are answering different questions.

The service parts

Worth having in one place, because a pump is usually repaired rather than replaced. These are Rimmer Bros numbers taken from their catalogue. They are a supplier's numbers rather than factory part numbers, and the archive has not seen a factory parts catalogue page for the fuel pump.

  • Points assembly: AUB6106A
  • Condenser: AUB6179, superseded, now CZX1004
  • Diaphragm kit, for a split diaphragm: WZX1710
  • Diaphragm gasket: AUB809
  • Joint washer: AUA573
  • Repair kit, points type: AZX1311RK
  • Repair kit, electronic: AZX1311ELRK
  • Mounting bobbins: UKC2451, three required. The replacement bobbin has an M6 metric thread and takes nut GHF212, where the original took the imperial GHF200

Ethanol, and the part nobody thinks about

The pump connects to the fuel lines through banjo unions, part BHH1940, banjo bolts 150926, and fibre sealing washers, part WF524.

Fibre washers and modern ethanol blended petrol are not natural companions, and this is a joint carrying petrol under pressure in an engine bay. Rimmer now sell a banjo conversion kit, RS1562BK, described as replacing the original banjos and bolts and offering longer life and greater protection against E10 fuel.

The archive notes this rather than recommends it. But if a pump is coming off anyway, the washers are the cheapest thing on the car and the consequence of one failing is not a breakdown.

Working through a pump that will not play

A sensible order, drawing on published troubleshooting guidance for SU pumps generally rather than on a Stag specific factory procedure.

  1. The inertia switch, if the car has one. Right hand bulkhead, next to the fuse box.
  2. Listen. Ignition on, engine not cranking, and compare what you hear against the three cases above.
  3. Tap it. A gentle tap with the handle of a screwdriver will often free points that have stopped on a closed cycle. This is a diagnosis, not a repair: a pump that needs tapping needs attention.
  4. Take the filler cap off. A blocked tank vent creates a vacuum no pump can overcome, and the symptom is a car that runs for twenty minutes and then dies. If it restarts with the cap off, the vent system is the fault and not the pump.
  5. Check power and earth at the pump. Twelve volts at the terminal and a sound earth. SU are emphatic about earthing the pump body securely, and about wiring connectors being tight and free of corrosion.
  6. Disconnect at the carburettor end and check whether fuel is actually arriving, which separates a pump fault from a delivery fault.
  7. Test the lines separately if it is not, to find where the blockage is.
  8. Check the inlet side for air leaks. A pump drawing air ticks happily and delivers nothing.

DuBois puts typical pump life at somewhere around 75,000 to 100,000 miles before wear becomes the explanation, which on most surviving Stags means the original pump is long past its first rebuild whether or not anyone remembers it happening.

Sources

  • Rimmer Bros catalogue, Triumph Stag fuel pump and inertia cut-out switch section, for part numbers, the inertia switch location, the pump variants and their polarity descriptions, and the E10 banjo conversion kit.
  • Rimmer Bros catalogue, Triumph Stag fuel system section, for the absence of any mechanical fuel pump in the Stag fuel system.
  • Dave DuBois, SU Fuel Pumps 101, updated February 2014, for the throw over mechanism, the volute spring setting the pressure, the points and electronic types, the diode and polarity sensitivity of the AUF 300 and AZX 1300 series, the check valve types, and typical pump life.
  • SU Carburettors, their own published description of the SU electric fuel pump, for the electromagnetic diaphragm principle, the tick being the contact points opening, and the effect of poor earthing on ticking rate, output and point wear.
  • The Triumph Experience and the MG T Society published accounts of the SU pump, for corroboration of the operating principle and of the use of diodes and other devices to protect the points.

What this page cannot establish

The archive has not seen the Repair Operation Manual's own fuel system section and publishes no delivery pressure figure. A figure of 2.7 lb/in² is attributed to the factory General Specification Data in transcriptions the archive has seen, and is not published here because it has not been read at source. The relationship between the SU designation AUF 303 and Rimmer's part number AZX1311 has not been established, and the archive does not assert that they are the same unit. Two of the more useful claims on this page, that the volute spring sets the delivery pressure and that the diode makes the AZX series polarity sensitive, rest on a single published source; the polarity distinction itself is corroborated by the supplier's own catalogue descriptions, but the explanation is not. The exact factory mounting location of the pump on the car is not established here: the parts catalogue shows a body mounted bracket and three rubber bobbins, and SU pumps are conventionally mounted low and near the tank because they push better than they pull, but the archive has not confirmed the Stag's own arrangement and would be glad of a correction from an owner who has had theirs off. It is not established when the inertia cut out switch was introduced, nor whether federal specification cars differ in any respect of the fuel pump installation. Part availability and supplier descriptions are as listed at the time of writing.

Last reviewed . First published .