"Duo-servo" brakes
Posted: Wed Jul 23, 2025 8:53 am
Like many of us, I'm fascinated by the improvements made to motorcycle brakes by our predecessors.
I've read most of the wisdom on the forum about making old brakes as good as they can be, and I've even tried a couple of careful improvements/modifications. Top of the list is, of course, to ensure that everything is in the best condition, with cables, linings, lubrication and so on all tickety-boo.
But here's a topic I've just come across: the Bendix "Duo-servo" system. It's been commonplace in the rear drums of things like US pick-ups, and it's still out there in the four-wheeled world, hidden away where glossy discs and brightly-painted calipers wouldn't get noticed.
As I understand it, the idea is for the leading ("primary") shoe to be fully floating, and the trailing ("secondary") shoe to be anchored adjacent to the operating cam, so that as the servo effect (sometimes referred to by the American truckers as "wrapping", which describes it quite well) comes in on the leading shoe, it presses on the end of the trailing shoe, which then provides a similar servo effect.
To achieve this, the pivot has to be removed, and the shoes at that end have to be faced with some sort of flat plate/shim, so that the servo force can be transmitted from leading to trailing. The trailing shoe would be anchored at its far end, by the cam. Thus the servo "wrapping" effect is carried all the way round the circle, from cam to anchor.
Here's a sketch I found from someone (not DBDBrian, I know) who has modified a Gold Star brake in this way. It's a bit fuzzy, but just about readable.
One question is why, in the sketch, the trailing shoe anchor - the only fixed point - is placed half-way along the shoe rather than at its end by the cam. Thinking about this, I reckon it might be good to have the shoes loosely connected to the brake plate with a couple of those palm-tree-shaped links as they are in old car drums, but the trailing shoe would certainly need a robust anchor to stop it pressing against and probably distorting the cam.
Having deleted the pivot and the associated disc which holds the shoes to the brake plate, that's another thing to consider for ensuring predictable consequences when the brake is applied. Perhaps a post just inboard of the shoes, with a small plate holding them as the pivot disc used to.
My conversation with the aforementioned Gold Star expert Brian suggests that this could well be a mite too aggressive for a motorcycle, and I'm interested to know if anyone has actually experimented with this sort of setup.
Effectively, it seems to be a twin leading shoe brake with only one cam!
Just askin'.
Best wishes to all,
Pete.
I've read most of the wisdom on the forum about making old brakes as good as they can be, and I've even tried a couple of careful improvements/modifications. Top of the list is, of course, to ensure that everything is in the best condition, with cables, linings, lubrication and so on all tickety-boo.
But here's a topic I've just come across: the Bendix "Duo-servo" system. It's been commonplace in the rear drums of things like US pick-ups, and it's still out there in the four-wheeled world, hidden away where glossy discs and brightly-painted calipers wouldn't get noticed.
As I understand it, the idea is for the leading ("primary") shoe to be fully floating, and the trailing ("secondary") shoe to be anchored adjacent to the operating cam, so that as the servo effect (sometimes referred to by the American truckers as "wrapping", which describes it quite well) comes in on the leading shoe, it presses on the end of the trailing shoe, which then provides a similar servo effect.
To achieve this, the pivot has to be removed, and the shoes at that end have to be faced with some sort of flat plate/shim, so that the servo force can be transmitted from leading to trailing. The trailing shoe would be anchored at its far end, by the cam. Thus the servo "wrapping" effect is carried all the way round the circle, from cam to anchor.
Here's a sketch I found from someone (not DBDBrian, I know) who has modified a Gold Star brake in this way. It's a bit fuzzy, but just about readable.
One question is why, in the sketch, the trailing shoe anchor - the only fixed point - is placed half-way along the shoe rather than at its end by the cam. Thinking about this, I reckon it might be good to have the shoes loosely connected to the brake plate with a couple of those palm-tree-shaped links as they are in old car drums, but the trailing shoe would certainly need a robust anchor to stop it pressing against and probably distorting the cam.
Having deleted the pivot and the associated disc which holds the shoes to the brake plate, that's another thing to consider for ensuring predictable consequences when the brake is applied. Perhaps a post just inboard of the shoes, with a small plate holding them as the pivot disc used to.
My conversation with the aforementioned Gold Star expert Brian suggests that this could well be a mite too aggressive for a motorcycle, and I'm interested to know if anyone has actually experimented with this sort of setup.
Effectively, it seems to be a twin leading shoe brake with only one cam!
Just askin'.
Best wishes to all,
Pete.
As most race singles, used in anger, use disc brake front brakes, for ease of maintanence and better braking. Alternatively, the multi-shoed 4LS/8LS drums are used, but the maintanence and setup of these is not practical on a regularly used road machine and they add to the unsprung weight of the front end, when compared to a floating disc brake setup.