AC Aceca Door Latches and Strikers: How They Work, Why They Fail

How the latch works

The latch is a rotary type. A toothed or star shaped wheel sits in the door on a shaft, alongside a pinion, and engages a striker or receiver plate on the body pillar. It works in two stages: a safety catch first, then the full latch. That two stage action is what matters when you come to read a fault, because a door can be held on the first stage and look shut while not being shut.

The equivalent mechanism on the MG A Fixed Head Coupe is documented in detail and worth understanding, and it is the latch an owner has now fitted to an Aceca. There the star wheel and the pinion sit on a common shaft with flats machined at each end, a double D. The wheel and pinion are hardened; the shaft is not.

This applies to the Aceca and the Greyhound rather than to the Ace. The Aceca has proper push button exterior handles where the Ace has small doors with no exterior handles at all and leather pulls, which is why sources about the Ace are silent on the subject.

Reading the symptom before condemning the part

A door that shuts but sits proud, as though it wants pushing further in, is usually the latch stopping on the first stage and not going fully home. Five things cause that, and only the last of them needs a part.

• The trim holds the door off. A door seal that is thicker or firmer than original can stop the door travelling far enough to reach the second stage, and so can door cards, the inside trim panels, that are too thick. When this fault was discussed on the club forum in 2008, the owner chose to replace non-original Furflex on his car before buying anything else, and the owner who has since fitted a replacement latch strongly recommends looking at the seal and the door cards first. Both cost nothing to check.
• The striker sits too far out, so the door cannot travel far enough to reach the second stage. This is the commonest cause, and it is the only one for which a confirmed cure is published anywhere.
• The pawl return spring has gone soft, so the pawl catches the first notch but not the second.
• The grease has hardened. One documented case on a Jaguar XK150, where the handles had become sluggish and sticky as though the springs were weak, was cured entirely by stripping, degreasing and relubricating.
• The second stage notch is worn, or the star wheel has lost its positive grip on its shaft and drifts out of index.

Look first at the door that has done the most work. The owner who fitted the MG A latch had a car built with left-hand drive and later converted to right-hand drive, and the latch that had failed was on the passenger door, because for most of the car's life that had been the driver's door. On a converted car, the worn latch may not be on the side you now get in.

If the door can be persuaded to catch by turning the star wheel round a position or two, that points either to uneven wear across the teeth or to a wheel that has drifted from its rest position, so the striker meets it in the wrong place. Which of those it is cannot be told with the latch in the door.

The archive's own car has this fault, and it is being attended to. A photograph of the shaft ends and the centre of the wheel will be added here when the latch comes out, because that is the evidence nobody has published.

Before touching anything, mark round the striker so it can be put back. And check the door is not simply dropped on its hinges, because adjusting a striker to mask a dropped door forces the door to work against itself and does not last. That is covered on the door fit page.

One caution before you blame the wheel

A restorer in Bradford, comparing Series 1 Jaguar E-Type latches of the same family, holds that movement at the star wheel tip of up to about three millimetres is an original design feature rather than wear, and is not the reason doors fail to close. Play on its own is therefore not a diagnosis. The striker position and the pawl spring come first.

The MG A Fixed Head Coupe question

The club magazine ACtion, in September 2004, stated that many of the door parts can be found in the MG A Fixed Head Coupe. The relevant MGA Coupe part numbers are AFH3894 for the right hand latch and AFH3895 for the left, with strikers C218L and C218R.

That claim circulated for more than twenty years. It was raised again on the club forum in 2008, when a member reported that catalogue photographs appeared to show identical latches, and again in 2013, and in all that time nobody reported buying one and fitting it. In October 2026 an Aceca owner told the archive that he had done exactly that, using an MG A Coupe latch bought from Rimmer Bros. His warning is that the cars are hand built, so fitting one can be a challenge: expect to offer it up, check it and adjust rather than bolt it straight in. He describes the fault as common on cars of this age, and the cure as a combination of things found by trial and error, but on his car the latch needed replacing regardless. It is one owner's account, and the archive would be glad to hear from anyone else who has done it.

Two practical warnings for anyone ordering one. The MGA Roadster latch is a different component entirely, cable operated, for a car with no exterior door handles, and at least one supplier lists those parts as suiting both body styles, which appears to be wrong. And the MGA pages themselves record a Coupe found fitted with a 1950 to 1954 Ford latch of the same family whose star wheel indexing was wrong and had to be reground, which is a warning that in this family of latch, looks the same and is the same are not the same thing.

The Rover P4 claim

A second lead circulates and does not survive examination. A member recorded in 2008 that the exterior chrome door handles interchange with the Rover P4, and suggested that the two cars might therefore share latches as well. In the same post he cited the 1962 AC parts list, which gives the Aceca's door latch/locks as 3/4875 for the left and 3/4876 for the right. The P4 point is a claim about handles. The suggestion that the latches might also match was offered as a possibility and supported by nothing.

The P4 is a four door saloon. A handle is a separately bought pressing, and sharing one with a two door coupe implies nothing about the mechanism behind it. No Rover P4 supplier lists a latch at all, and no technical description of the P4 latch could be found. The archive records the handle claim as reported and the latch extension of it as unsupported.

The idea has a second root. In 2013 a member recalled having once found a United Kingdom reconditioning service for these latches, and thought it might have been connected with the Rover P4 community. He described the memory as vague and produced nothing further. That is the whole of it.

Repairs: what worked, and what did not

The principle that runs through every successful repair is the same. Replace a worn friction or peened joint with a positive mechanical one.

• A professional latch rebuild is the only route with a clean confirmed success. An owner of a comparable American car sent his away, had it rebuilt, refitted it and confirmed it working after a road test.
• On a Morgan, a worn peened joint on the latch was drilled and tapped and replaced with a threaded and Loctited component. The owner reports it still sound after fourteen years.
• On a Jaguar E-Type, an owner replaced the bush and spot welded the shaft to the dog after filing off the riveted end, and reports the star wheels turning freely with no wobble.

One further method is worth setting out, because it addresses this exact wear rather than a comparable fault on another car. A restorer posting on the club forum in 2013 described the two things that go on these latches as the star or gear wheel becoming loose on its shaft, and wear on the shaft and its bushing. His remedy, for anyone with access to a lathe or mill, is to make up a new shaft oversize in diameter to take up the wear, having first cleaned and lubricated the mechanism and checked the springs. He posted photographs of the work. The archive records it as a method described by someone who does this for a living, and notes that nobody in that discussion reported having carried it out and found it good.

A quicker idea was put forward on the forum in 2008: refit the star wheel turned half a turn on its shaft, so that unworn teeth meet the striker, or swap the latches left for right to move the wear. Nobody reported trying either. The latches are handed, so a straight swap from one door to the other is unlikely to work. Turning the wheel on its shaft is closer to the re-indexing described below, and how far it can be turned depends on the shape of the shaft, which nobody has yet recorded for the AC part.

And one repair is on record as having failed.

• Building the worn star wheel back up by welding a strip onto it was tried on a 1940s American car, and the owner who later inherited that car describes the repair as unreliable and the door as prone to popping loose. It is the obvious thing to try and it is the one thing anyone has reported not working.

A note on the state of the evidence. The 2013 discussion was started by an owner whose left hand catch was being held shut with tape and a tie down strap, and whose driver's side was going the same way. He was given four routes: a supplier, a modern conversion, a reconditioning service and the oversize shaft repair. He closed the thread saying it was time to pull the latches out and inspect them, and promised to report back. He never did. The first account the archive has of an Aceca latch being replaced came thirteen years later, from an owner whose latch needed replacing and who fitted the MG A Coupe part described above.

Illustrated procedures for this mechanism, covering drilling and pinning the shaft, welding, re-indexing the star wheel against the pinion, and the pawl return spring, are published at mgaguru.com in its MG A Coupe technical section. Whatever the origin of the AC part, the techniques transfer. Re-bushing has been done with a flanged slide bearing turned to suit, at a cost of a pound or two.

Getting a part, or not

There is no new or remanufactured latch sold as an Aceca part, and no second hand supply. The practical route is the MG A Fixed Head Coupe latch, AFH3894 right hand and AFH3895 left hand, which one owner has fitted to an Aceca using a latch bought from Rimmer Bros. Order the Coupe latch, not the Roadster one, and allow for fitting work on a hand built door.

For anyone who would rather repair the original, an MG parts specialist in Ipswich, trading as MGbits, offers a reconditioning service for the MG A Coupe latch on an exchange basis, and sells a spindle and bush repair kit that addresses shaft and bush wear rather than replacing the whole latch. Their listing for the reconditioning service currently shows as unavailable, so it is worth asking whether they will recondition the latch from an Aceca.

Brown and Gammons, the MG specialist, were checked in 2013 and found to list lock plates but not complete door mechanisms. That is the same pattern every other supplier shows: striker and lock plates can be had, the mechanism cannot.

The body side of the latch is the easier half. In 2008 a United States AC specialist could supply the receiver plate that mounts on the pillar, and the club's own spares service advertised an Aceca door striker and plate in ACtion that September. It was the mechanism in the door that nobody could source until an owner turned to the MG A Coupe part.

On the modern conversion, the product named on the forum in 2013 was the Autolock bear claw latch set, ordered from Summit Racing in the United States, at 52 dollars a set at the time. The member who found it described it as requiring panel work and as appearing to be the best and safest solution available. He was offering an opinion rather than an account: nobody in that discussion fitted one. Everything the archive says elsewhere on this page about bear claw conversions still applies, in particular that no account has been found of anyone doing it on a coachbuilt aluminium car.

AC Heritage at Brooklands hold full catalogue factory drawings and manufacture parts on original AC tooling from Thames Ditton. They are the likeliest holder of an authoritative answer about what this part is, and as far as the archive can tell nobody has asked them.

A UK specialist in these latches was named on the club forum in 2013 and appears never to have been followed up. The club will have the details.

A modern bear claw latch conversion is sometimes used on other cars. It is confirmed working on a steel bodied Ford, fitted at the dovetail positions, after seven years. The archive has found no account of anyone doing it on a coachbuilt aluminium car, and it requires rigid adjustable steel mounting plates at both the door and the pillar, which on an ash framed door is a significant piece of fabrication. It is recorded here as an option that exists, not as one the archive can point to a precedent for.

The safety point

The current DVSA MOT inspection manual classes a hinged door that is likely to open inadvertently or not remain closed as a dangerous defect, which is a more serious category than a major defect. Acecas are exempt from testing as vehicles of historic interest, so no tester will ever apply that wording to one. It remains the standard an insurer or a court would reach for, and it is the reason a latch that only holds when the wheel happens to sit in the right place is worth attending to promptly rather than living with.

What is not established

• Who manufactured the latch fitted to the Ace and Aceca. No maker's stamping, no British patent and no catalogue page has been traced. Wilmot Breeden is the obvious candidate, as a Birmingham firm supplying door locks to almost every British manufacturer, but the one design of theirs the archive could document, on the Standard Vanguard, is a cam and pawl rather than a star wheel and pinion. Bloxwich can be ruled out; their recorded products are commercial vehicle and container door gear.
• Whether the MG A Fixed Head Coupe latch fits every Aceca without modification. One owner has fitted one; the cars were built by hand and he warns that fitting can be a challenge.
• Whether AC parts list numbers 3/4875 and 3/4876, given in the 1962 list as the door latch/locks, are the latch mechanism alone or a latch, lock and handle assembly.
• Whether the Ace and Aceca latches differ from one another, and whether the Greyhound shares either.
• What the original manufacturer's part number for the AC latch was.
• Whether anyone currently offers reconditioning of this latch. None advertises it.
• How many Acecas have had the latch replaced or repaired. Two club forum discussions, in 2008 and 2013, both ended without a result, and the one account the archive has is the owner who fitted an MG A Coupe latch in 2026.

On the AC part, two threads on the AC Owners' Club forum, of September and October 2008 and of January 2013, both of which ended without any participant reporting a result, together with the club magazine ACtion of September 2004 as quoted on the forum. On fitting the MG A Coupe latch, and on the door seal and door cards, first hand information from an Aceca owner given to the archive in October 2026. On the mechanism and its repair, the MG A Coupe technical pages at mgaguru.com, and single first hand accounts from owners of a Morgan, a Jaguar E-Type, a Jaguar XK150 and two American cars. On the roadworthiness standard, the DVSA MOT inspection manual. Supplier and trading information from those firms' own listings and from Companies House. No individual is named here.