Equipment/Product Review (1984) Rear Derailleurs for Racing (Berto)
07-30-20, 03:12 PM
#26
Senior Member
My first impression after reading that article carefully is that Berto is very rational and reasonable in his approach. I don't think I've read any of his articles since probably around the time this one came out. I don't take away that he is bashing Campagnolo at all, and he in fact accurately states why Campagnolo was popular with the racing crowd: they are reliable and fully rebuild-able. I also remember most racer types thinking Superbe was great stuff too. However, it cost about the same as Campagnolo, and you could not get spare parts. That made it a hard sell.
I also remember the Dura Ace AX as being a good working derailleur. In fact the whole group worked well, IMHO. It was however, really too weird for the market at that time. Shimano rectified this a few years later when they made the 7400 group, but that's another story.
With the exception of that Huret, which I don't know, by that time IME really all those derailleurs worked adequately well. Ride whatever you like. LIke many, at that time, I preferred Campagnolo. It was insanely reliable. My take at that time was that all these derailleurs shift well enough if properly adjusted, and I don't really care nor did I ever notice a couple degrees of shift lever wiggle required to center. I always did this by ear and moved the lever until the chain was at it's quietest point. It quickly becomes subconscious.
That said, the results clearly show that Suntours slant parallelogram was superior. Note that everyone else copied it when the patent expired - except the post Tullio Campy, but even they copied it eventually.
Regarding the testing set up, Berto I think made a good case for all of the choices he made, and he was clearly trying to be as objective as possible. I do however have some critiques.
1. I think that a European notched tooth freewheel and Sedisport chain would have much more typical of what was used by actual racers at the time. It seems like he went with the HKK chain and Suntour freewheel to maintain consistency with his earlier test. The results IMO would have been quite different. Also, using a Suntour FW is clearly going to bias the results at least slightly towards Suntour.
2. A 13-24T 5 speed freewheel was an odd choice. A 13-21T 6 speed freewheel was pretty much standard at the time. As he notes in the text, the campy RD shifts better with a tighter FW. I'd assume the smaller jumps of a 6 speed would also improve the score.
3. Essentially, he is measuring centering precision, not shifting speed. That's fine, but it's not the same as faster shifting often claimed. Given the available tech at the time, a rig to measure shift speed would have been much more involved.
4. No load on the rear wheel? While everyone knew back then that you needed to back way off on the pedal pressure to shift, in the real world, there often would have been some tension on that chain.
6. Ideally it would have been nice to see each RD used with its own manufacturers preferred chain and freewheel, but then of course you are testing chains and freewheels as much or more than derailleurs.
The reason I did not like the earlier two generations of Dura Ace (Crane through EX), as well as the equivalent 600 rear derailleurs, has to do with Berto's bullet point #4 . They really did not have enough wrap. I'm not sure why they did this other than it probably makes speedier shifting in the stand under no load at all. In the real world, when the freewheel and chain had the slightest amount of wear, they would skip under load. For a racer type, this makes it useless. Imagine you respond to a jump when you're with a group going up a false flat 3% grade, and your chain skips as you take off, and then you have to shift to respond to the new speed, and it skips again as you try to shift. To be fair, Campagnolo RD could do this too, but only after a freewheel had like 10,000+ miles on it, and the stress wouldn't break them. This seems to have long been solved. My modern Ultegra derailleur has the upper jockey wheel very forward, giving a full 180 degrees of wrap for max tooth engagement. Presumably the improved performance of chains and freewheels allowed this position to work without the undershift Berto talks about.
I also remember the Dura Ace AX as being a good working derailleur. In fact the whole group worked well, IMHO. It was however, really too weird for the market at that time. Shimano rectified this a few years later when they made the 7400 group, but that's another story.
With the exception of that Huret, which I don't know, by that time IME really all those derailleurs worked adequately well. Ride whatever you like. LIke many, at that time, I preferred Campagnolo. It was insanely reliable. My take at that time was that all these derailleurs shift well enough if properly adjusted, and I don't really care nor did I ever notice a couple degrees of shift lever wiggle required to center. I always did this by ear and moved the lever until the chain was at it's quietest point. It quickly becomes subconscious.
That said, the results clearly show that Suntours slant parallelogram was superior. Note that everyone else copied it when the patent expired - except the post Tullio Campy, but even they copied it eventually.
Regarding the testing set up, Berto I think made a good case for all of the choices he made, and he was clearly trying to be as objective as possible. I do however have some critiques.
1. I think that a European notched tooth freewheel and Sedisport chain would have much more typical of what was used by actual racers at the time. It seems like he went with the HKK chain and Suntour freewheel to maintain consistency with his earlier test. The results IMO would have been quite different. Also, using a Suntour FW is clearly going to bias the results at least slightly towards Suntour.
2. A 13-24T 5 speed freewheel was an odd choice. A 13-21T 6 speed freewheel was pretty much standard at the time. As he notes in the text, the campy RD shifts better with a tighter FW. I'd assume the smaller jumps of a 6 speed would also improve the score.
3. Essentially, he is measuring centering precision, not shifting speed. That's fine, but it's not the same as faster shifting often claimed. Given the available tech at the time, a rig to measure shift speed would have been much more involved.
4. No load on the rear wheel? While everyone knew back then that you needed to back way off on the pedal pressure to shift, in the real world, there often would have been some tension on that chain.
6. Ideally it would have been nice to see each RD used with its own manufacturers preferred chain and freewheel, but then of course you are testing chains and freewheels as much or more than derailleurs.
The reason I did not like the earlier two generations of Dura Ace (Crane through EX), as well as the equivalent 600 rear derailleurs, has to do with Berto's bullet point #4 . They really did not have enough wrap. I'm not sure why they did this other than it probably makes speedier shifting in the stand under no load at all. In the real world, when the freewheel and chain had the slightest amount of wear, they would skip under load. For a racer type, this makes it useless. Imagine you respond to a jump when you're with a group going up a false flat 3% grade, and your chain skips as you take off, and then you have to shift to respond to the new speed, and it skips again as you try to shift. To be fair, Campagnolo RD could do this too, but only after a freewheel had like 10,000+ miles on it, and the stress wouldn't break them. This seems to have long been solved. My modern Ultegra derailleur has the upper jockey wheel very forward, giving a full 180 degrees of wrap for max tooth engagement. Presumably the improved performance of chains and freewheels allowed this position to work without the undershift Berto talks about.
Likes For Salamandrine:
