frantik

Well i tried a different set of cantis on a 700c bike and they work GREAT! I was using modern Avid Short 4 brakes but replaced them with old school Deore MT62 cantis. It's possible to lock the wheels with these bad boys, though I tend to set mine up so that you don't lock up since that doesn't slow you down as quickly

/retraction retracted

jimmuller 

How do you adjust brakes so they don't lock up? Set them so that the levers bottom out? Regardless of how you do it, it necessarily means giving up some braking power. If you can peg the performance to be right at the limit of adhesion then fine, but if you err on the side of too much braking then you have to learn to modulate them by hand anyway. And if you err on the side of too little, you are indeed giving up that much braking power.

I'd rather have more than enough than not enough. But that's just me.

frantik

i set them up so that they don't lock up until you've almost bottomed out.. this way you get more travel to modulate how much power you want.

I do get that they are a pain to set up, and it seems like the modern "low profile" ones kinda suck.. so i could see why a lot of people don't like them

toytech

You can vary the way the act and work quite dramatically by playing around with angles and straddle length too, I set all mine up (and all brakes for that matter) as close as possible without rub. Modulation is in your fingers

frantik

yeah this is one one reason i like cantis.. lots of adjustability. once you get the hang of setting them up, it's not that much longer, though obviously if you're a manufacturer, every second adds up

robatsu

I'm wondering if the liability of the bike producer is reduced a bit by the use of out of the box calipers rather than a cantilever setup with which they are very involved.

frantik

with "z-link" cables it's not much different in terms of installation really.. but those zlinks dont give you good adjustabilty

noglider 

If a brake can skid my wheel then it's powerful enough. If it requires me to squeeze very hard, it's no less powerful enough than a brake that can skid the wheel with a light touch of a finger. This is what a lot of people don't understand. As long as I'm strong enough to get maximum braking, the brake is good enough for me.

If you switch from a dual-pivot or a V brake to a single-pivot sidepull, you might think the sidepull sucks and barely stops. But that's because you're not squeezing hard enough. Just squeeze harder, and you'll stop just fine. Of course, if you're not strong enough to squeeze that hard, then a single pivot sidepull is not for you. That's why they invented brakes with more leverage. Leverage in a caliper brake is merely the amount movement of the brake arm for a given amount of cable pull.

noglider 

I would never set up a brake that lets the lever bottom out on the handlebar.

jimmuller 

+100,000 That's exactly the point I was trying to make.

Whether the lever hits maximum braking at 1/3 of its movement range or at 9/10 of its movement range makes no absolutely difference in braking. Having the lever closer to bottoming out just means you might not actually be able to hit maximum braking if, heaven forbid, the cable stretches or slips, or the pads wear away. Which is another way of saying you have voluntarily given up safety margin for absolutely no gain.

frantik, do yourself a favor and tighten you cables!

fietsbob

Once the Asian companies started making cantilever brakes with pad height adjustments possible
the mass producing frame making companies were able to pass through parts
where the bosses were not as accurately placed, as they needed to be before , with Mafac Brakes,
so they were in all sorts of 'Oh thats close enough', boss sitings.

80's, road vs Cross
Had 2 screwed and glued AlAns, same fork blades, but, the longer ones with the (Modolo) cantilever brakes on them
vibrated wildly when the brake was applied , but the road fork with the brake bolted through the fork crown,
this was not a problem

IthaDan 

As I understand it:

--> The amount of power that a canti affords would be suicidal in a pack. Similar logic mandates fixed gears in mass start track events to keep riders moving roughly the same speed.

--> The consequences of a swapped (neutral support?) wheel having a smaller rim are FAR less dire with a caliper [where the pads barely make contact with the rim] vs cantilevers [where the pads dive into the spokes].

Couple those two with the aforementioned aerodynamics, a drive towards lighter and lighter frames, and a quicker "quick release" than a cantilever, and we're getting towards a pretty airtight case.

not_jason

I don't buy the argument that the industry has suppressed the cantilever brake for the sake of controlling the market. And part of the reason is that there are just too many exceptions. You may not see road racing bikes with cantis, but you do see plenty of touring bikes with them. And of course, they are pretty much the only thing you ever see on cyclocross bikes, which bleeds back onto the street in the form of cross bikes converted for road use. Pretty much any bike with flat bars runs V-brakes, even the relatively fast ones. I think that there is plenty of reason to suspect that if it made good sense to do so, manufacturers would be putting them on fast road bikes too. My suspicion is that they just aren't necessary. Whenever I ride a new bike with a well tuned dual pivot brake, I'm amazed at how much stopping power it has, and all with an incredibly compact, easily tuned, aerodynamic set of brakes. As much as I enjoy the cantilevers on my gran turismo, they aren't exactly elegant or convenient to set up. Nor do I feel like they really outperform my other good brakes in any really important sense. Maybe I'd notice it if I were hauling an extra 100 pounds of gear down a mountain descent, but that's not exactly something most racers do, right?

frantik

I wouldn't either....

um, yes there is a difference. you don't get as much modulation and experience wheel lock up sooner. Cantilevers can be adjusted so that you make contact with the wheel after a short amount of travel but don't lock up until much more travel. In this way you get a lot of range between "barely slowing down" and "OH SHIIIIIIIIIII"

LMAO...

jimmuller 

I spent part of the going through the force diagrams and calculations for canti brakes. Some of this is surprising, some isn't.

"Common wisdom" says you can adjust the straddle cable length to affect the braking power. It turns out you can but it depends on the canti geometry.

First, consider this picture:

In this picture:
alpha is the angle of the straddle cable from vertical
theta is the angle of the caliper's lever arm form horizontal

For a range of angles, the effectiveness of braking is shown in the following table, where 1 indicates a straight pull. Everything is multiplied by the ratio of the arms on the canti mechanism but since they are fixed, this effect is not a factor. Of course all of this discussion applies at the point where the pads are in contact with the rim. Any movement the arms must go through before that contact occurs has no effect on the in-contact result.

Each row shows results for alpha values 0, 10, 20, 30, 40, 50, 60 degrees. I.e. each "column" is a different value of alpha, but I can make them display in columns.
The first column in each row is the theta value, for values 0, 10, 20, 30, 40 degrees.

0, 1, 1, 1, 1, 1, 1, 1
10, .985, 1.015, 1.048, 1.085, 1.131, 1.191, 1.286
20, .940, 1.048, 1.064, 1.137, 1.227, 1.347, 1.532
30, .866, .954, 1.048, 1.155, 1.286, 1.462, 1.732
40, .766, .879, 1.0, 1.137, 1.305, 1.532, 1.879

The key value here is theta, the angle between horizontal and the arm holding the straddle cable. If the arm is horizontal, i.e. alpha is zero, there is no benefit at all in changing the straddle cable angle. Only when the arm is angled upward is there any benefit. This can be accomplished by using worn pads or a narrower rim so that the straddle cable ends sit higher. Otherwise, alpha is fixed by the arm shape.

For those still awake... If T is the tension in the main brake cable (actually half, since it is split between two arms):
The tension in the straddle cable is T / cos alpha. This is equal to T if alpha is 0 (vertical), much higher if the straddle cable is more horizontal.

The torque rotating the arm against the rim is
(T * (R1/R2 ) * sin(90 - alpha + theta)) / cos (alpha).

Sidepull brakes are not so complicated, of course. Their "power", for a given tension in the cable, is simply the ratio of reach to the distance between pivot and cable attachment point. Centerpull brakes are similar to cantis except that the geometry stretches across to the pivot on the opposite side of the fork from the straddle cable attachment.
-----------------------
One final point, valid for any kind of brake except for how it affects the straddle cable angle, is this. The amount of movement the brake lever goes through before pad-rim contact occurs has no effect on the braking power other than the back pressure of the return spring(s). Given that return springs are not at zero compression when the brakes are fully released and given that the return spring pressure is quite small compared to the force of braking, the difference between back pressure of a fully released and fully applied brake is insignificant. Which is merely another way of saying that adjusting the cable so that the lever sits at a lower position makes no change in the braking power. It simply uses up some of the potential displacement of the lever. This does change the hand position, and thus may have an effect on the cyclist ability to squeeze the lever. But if it reduces the available hand strength then it is also giving away potential braking power. In no way is it a good thing.

RFC

I've been following this thread off and on and have wondered -- when is someone going to provide the correct answer?


Modern dual pivot calipers are grossly better than cantis in a number of ways.

Howard

for many but not all applications.

Mud & fenders come to mind.
They've got to be cheaper to install though, so that's a plus. Maybe even cheaper to make. Easier to market too.