Many Austin Seven owners have written about this subject for decades, especially cross-shaft twist and deflection, cable lengths, and how to ‘balance’ the brakes. It may help to review what's already been written and try to draw it together as a summary. Here’s what a very long online search discovered - and a distillation of what has been written. It concentrates mainly on the simple coupled brakes system used from mid-1930 until around 1932 but also applies, in general, to the earlier uncoupled and later more complicated system fitted in later years:
Correctly adjusting the early coupled brakes on an Austin Seven takes more than raising all four wheels and setting the brakes to come on in a satisfactory sequence. The brakes can appear to look beautifully balanced with a light application in the garage, yet pull to one side when you press the pedal hard on the road. The difficulty lies partly in the operating mechanism's flexibility and the various adjustments that have have been made to let the shoes just begin to touch their drums. Unfortunately, the forces generated during a heavy-application stop are quite different to those on a light stop.
On the standard right-hand-drive arrangement, the characteristic inequality favours the offside rear brake. The nearside rear receives its operating effort from the other end of the cross shaft, and so is more affected by the shaft’s torsional deflection and flex in its mounting brackets. Consequently, the familiar remedy is to give the nearside rear an initial lead, allowing the inequality to increase as pedal pressure rises. The discussion in Motor Sport, May 1950, specifically identifies cross-shaft twist as a cause of unequal rear braking
A particularly revealing owner’s account appears in the Austin Seven Friends discussion “adjusting brakes”, in March 2014 Hedd Jones describes setting the rear brakes on his RL Saloon, which retained the early unsleeved cross-shaft. With the brakes adjusted to pull evenly under light pedal pressure, a hard application caused the car to veer towards the middle of the road. He therefore gave the nearside rear a lead, accepting a tendency to pull left under light braking in exchange for a straighter stop under heavy braking. His experience illustrates the limitation of adjustment alone: a fixed difference in cable setting can compensate for flexibility at a particular load, but cannot necessarily produce perfect balance at every load on the pedal.
Two separate movements need to be distinguished. The first is twisting of the cross-shaft itself. As the load increases, the shaft’s operating arms no longer keep the same angular relationship they had under negligible load. The second is deflection of the supporting brackets that allows the shaft’s offside end to move back and the nearside forwards. Both affect the movement delivered to the cables, and both are frequently described, over simply, as “cross-shaft flex”. However, fitting braces to stiffen the bearing brackets do not eliminate torsion within the shaft, while of course cable adjustment removes neither cause.
This explains the limitations of the customary raised-wheel test. Very little braking effort is required to stop a wheel being turned by hand. Although the test is useful for establishing clearance between shoes and drums (which should be the very minimum), detecting drag and setting the order in which the brakes begin to operate, it does not mean that an equal braking effort will apply to all wheels when the pedal is pressed much harder. Again, equal initial contact and equal braking under load are not the same thing.
The problem of the cross-shaft bearing brackets is described particularly clearly by Jim Holyoake in “Brake Cross-shaft bearing lubrication”, November 2015 When the pedal was pressed hard, he observed the offside end moving forward and the nearside end moving backwards. He restrained the offside by fitting a ⅜-inch BSF bolt through the chassis cross-member, with its head bearing against the bracket that holds cross-shaft bearing assembly. A separate restraint prevented rearward movement at the nearside bracket. His car was an uncoupled-brake 1929 RK, and, after these modifications, he noted a considerable increase in braking power. Another contributor, Press, recalls using an offside stop bolt and a short angled nearside strut in the 1950s.
Austin subsequently tackled the torsional inequality with a concentric tube over the offside portion of the cross-shaft. This tube was connected to the brake pedal and carried pedal torque to the centre before transmitting it through to the shaft itself. The 1950 Motor Sport article notes post-1933 cars had this arrangement and describes how lever seizure or interference could undermine its operation. Identifying the mechanism fitted to your Seven is therefore essential: an early plain shaft and a later sleeved assembly should not automatically receive identical adjustment. After decades of repairs, substitutions, and bodging, the car’s year of manufacture is not evidence of what lies beneath.
The ability to finely adjust the length of the rear brake cables is especially useful where the early shaft remains in use. A turnbuckle, or rigging screw on the nearside cable, makes small changes easier to achieve and repeat, but it is an adjuster, not an automatic compensator. Shortening the nearside cable advances that brake’s application; lengthening the offside cable delays it relative to the other. Either can establish a nearside lead, although the remaining adjustments must then be checked.
The Friends discussion “rear brake cable adjustment adaptions”, April 2021 examines turnbuckles, the now-unavailable cables with threaded ends and the difficulty of making fine changes with those units where the cable can be pulled through the connecting end. On page three contributor dickie65 describes using a rigging screw on the right-hand rear cable, leaving enough slack to accommodate greater losses on the left. Thus, the underlying principle supports giving the nearside brake a lead, but it does not require fitting the turnbuckle on that side. The same contributior warns that an unloaded static setting can be misleading and that the handbrake may not reproduce the rear balance obtained with the pedal.
The choice of adjuster also matters. On page four of the discussion there is a tremendous amount of useful information on the various kinds of cable ends and how they can be bought, or made, and fitted. Cliff Ringrose explains that the goal is to make an accurate setting easier to achieve, not to improve an already satisfactory system simply by adding a turnbuckle. A properly made adjustable cable end offers another way of obtaining the necessary fine control. Whichever arrangement is used, the cable connections and the means of retaining the adjustment are part of the braking mechanism and must be suitable for that duty.
A useful practical starting sequence appears in “...to get good brakes?”, April 2024 Steve Jones describes raising the car level with all wheels free, setting the front brakes to begin rubbing at two handbrake notches, the nearside rear at three and the offside rear at four. This deliberately establishes the order: front brakes, nearside rear, offside rear. The owner subsequently reported a favourable result. However, the car concerned combined a short-wheelbase coupled chassis with the superior semi-Girling front brakes. The handbrake notch counts are therefore an experienced owner’s method for that installation, not a factory specification for every 1930–1932 Seven. The discussion also records Robert Leigh’s preference for adjustable cable ends over turnbuckles, for he connected the latter with additional 'softness' in the system.
Front-to-rear balance introduces another variable. Braking torque pushes the early front axle and radius-arm arrangement backwards and so reduces front-brake application—the effect commonly called “negative servo”. A satisfactory relationship between front and rear brakes with the car stationary may therefore change during a stop. This is why the Dorset Austin Seven Club’s “Austin 7 Brake Adjustment” recommends that the pre-Girling front brakes operate slightly before the rears, so they work together under heavy braking. The article also stresses inspection before adjustment and the usefulness of a brake-testing machine such as a 'Ferodo Tapely' (which are often advertised cheaply on eBay). Its detailed shoe-shimming suggestions should be treated separately, since some such methods are disputed by other experienced owners.
The published advice is not entirely consistent. A reproduced 1936 Austin Magazine article, “Braking Efficiency” describes the fronts as initially operating slightly later than the rears. That instruction concerns a later production arrangement and should not be combined with advice for an early unsleeved shaft. The article does, however, provide clear evidence of Austin’s threaded rear-cable ends and the intended method of making small individual adjustments without twisting the cables.
Neither cross-shaft bracing nor careful cable setting can overcome faults at the wheels. Ray White’s account brake cross shaft hangers?”of his 1930 coupled-brake Seven brings together flexible shaft supports, the need for nearside rear lead, front-axle movement and incorrect assembly of operating components. A pull under braking should therefore prompt examination of the complete system. Unequal lining condition, loose rivets and shoe pivots, worn cam bushes, contamination by oil, wear, or incorrect operating geometry, etc., can all cause unequal braking even when the cable movements appear satisfactory.
Modified cars require particular care when combining components. In “Converting a 1932 RN to semi-Girling brakes” Roger Bateman describes the difficulty of achieving satisfactory operation with later brakes and long operating levers while retaining the RN cross-shaft. He reports resolving the problem by reverting to shorter RN-type levers. The experience demonstrates that parts which work well in their original installation do not necessarily form a correctly matched system when transferred individually to an earlier chassis.
A sound investigation therefore begins by identifying the cross-shaft and brake components, rectifying wear or stiffness, and observing the mechanism under substantial foot-pedal pressure. The aim is to distinguish shaft twisting from bracket bracket flex and bearing clearance before deciding how much adjustment can reasonably achieve. The raised-wheel test then provides an initial setting, including any necessary nearside rear lead, followed by checks at progressively greater braking effort.
A suitable brake tester can help quantify individual wheel effort, while controlled low-speed testing in a closed area checks the car’s behaviour as a complete vehicle. Older accounts describe comparing wheel locking on grass or gravel, but simultaneous locking on a loose surface does not precisely demonstrate road balance. Likewise, a satisfactory handbrake test does not establish satisfactory balance under a hard pedal application.
The characteristic early-Seven difficulty is that the whole braking mechanism changes its effective geometry as the driver asks more of it. Giving the nearside rear a preliminary lead can compensate for the resulting offside advantage; accurate cable adjustment makes that compromise manageable, and properly engineered cross-shaft bracing can reduce the contribution from flexing brackets. The final judgement must be made under meaningful braking load. Four wheels agreeing politely on axle stands is a good start, but they haven't been asked to stop the car on the road. Remember, always be able to stop in the distance you can see to be clear.