I can't find anything that discusses the correct way to set up the DiaCompe 980 cantilevers. I'm aware of the need to secure the main cable so the straddle cable can't slip into the tire. But beside that, what do you know about adjusting this brake from a mechanical advantage point of view.

https://www.circleacycles.com/cantilevers/

Interactive calculator for determining and adjusting mechanical advantage. Very nice interface.

A couple of immediate observations from your picture:

- Changing straddle height in your setup may not improve mechanical advantage based on the angle of your brake arms.
- Moving the pads as close to the brake arms as possible may help a little, but not much to improve the brake arm angle.

Great reference, thanks! My picture doesn't show it very well, but the brake is at rest. When applied, the brake arms are horizontal, perpendicular to the arm going down to the pivot. My uneducated thought was that this was the most efficient transfer of power to the brake pad. As for the interactive calculator, it appears that in all cases with these old brakes, lengthening the straddle cable considerably lessens mechanical advantage...which makes me wonder why so many of the old mountain bikes were set up with the straddle cable looking like a freakin' teepee!

https://www.circleacycles.com/cantilevers/

Useful!

If you noticed by playing around with the calculator slide adjusters, having the brake arms at or near horizontal largely negates any effects of straddle cable height. Contrary to my own empirical understanding (based largely on what I've read online) about cantis, the smaller the angle (closer to vertical) the brake arms are, the more of an effect you will get on mechanical advantage by lowering the straddle cable height. On those 80s mtn bikes, most of the cantis were high-profile cantis, with the arms straight out (~90 degrees); therefore it didn't matter if the cable height was high or low... there was little mechanical advantage, period.

It's my understanding that those non-low profile cantis have more mechanical advantage, but are impractical because of heel strike in the rear. I set them up with the cable as close to perpendicular to the arms as possible and that means a long straddle cable. It works for me.

No... the only situation where the brakes mechanical advantate is entirely insensitive to cable saddle hangar height is if you are using wide profile brake (Like old MAFAC's) and have the alfa angle set at 90 degrees (so the cable anchors are setup exactly on a horsontal line through the brake pivot bolts. The OP brakes are not wide profile, the saddle cable anchors are well above the cantilever pivots, probably more like an alfa angle of 60 degrees. The fact that the arms are constructed to be horizontal at this position does nothing to effect the actual geometry, it is all about the pivot locations, not the metal that connects them. Lowering the saddle cable of these brakes will result in a higher mechanical advantage.
Note that the adjustable mounting position of the brake pad studs can effect the resulting alfa angle. If you were to set the studs so that the pads are closer together, alfa will decrease or if you slide the pads so they are wide (closer to the arms) then alfa will be increased. Just as southpaw noted, decreasing the alfa angle will give the brakes a higher mechanical advantage (make them more like narrow angle brakes) however downside is that narrow angle brakes are regressive, in that they loose mechanical advantage the more you pull them. This makes for brakes that move very little during the inital cable pull and so requires that the resting pads be set close to the rims. A cantilever setup with a wide alfa angle can be setup so that the pads are further from the rim (more clearance for mud, more tolerant of bent rims) and the mechanical advantage dynamically changes less as the brakes are actuated. Tradeoff of wide angle is that there is less mechanical advantage to start with. With alfa angle of exactly 90 degrees, you have the original linear pull brake. If you were to pair alfa 90 wide angle brakes with brake levers that have more than typical mechanical advantage, it would be a sweet combination of linear response and sufficient mechanical advantage. Too bad that there were not brake levers with variable cable connection positions to alter the lever mechanical advantage to suit the brakes.

Oops, yes, you are correct. I missed that. Yes, the alpha angle is more like 60 degrees.

Thanks for the discussion, I'm starting to look at brakes in whole new light. I didn't have a protractor handy, but I made some "best I could" measurements with a little geometry and came out with a 51 degree alpha. When I entered all my measurements the mechanical advantage fell on the curve about a third of the way up the graph. by the way, where do you guys learn this stuff?

Same place you learned about bike forums... those interwebs!