Kontact 

8 speed Campy and 8 speed Sachs were compatible, but different designs. 7 speed Sachs was compatible with Shimano. And then there was Mavic index

All four used a drop style slanted parallelogram "servo pantograph" design, the last bit means that the attachment point to the frame was spring loaded. This was an older Shimano design that they tried to get shifting as good as Suntour's slant parallelogram, and they kept using it when they adopted that feature as well. It is mostly unused today.

The specifics of where the cable is anchored to the parallelogram, the start angle of the parallelogram arms, etc are what determine the way the cable pulls (until SRAM added a cable guide cam to the outer plate).


If you want to play with indexing, buy a cheap used Shimano or Suntour RD and a matching 7 speed downtube shifter. All on the bar systems are essentially just fancy mechanisms to turn that round bottom of the downtube shifter, pulling the cable.

choddo

I think you misread it

As maddog pointed out, that part of the article was describing how a 10S Campy RD has to move different distances as it goes across the cassette. The bit you highlighted in bold is the distance as it goes through the 7th & 8th out of 10 cogs, not how far an appropriate RD moves on every cog of a 7 or 8 speed cassette which is, I think, what you said you measured? And you didn’t mention non-linear movement.
You were talking about spacer width of 3.1mm and the article was talking about cable pull of 3mm. I think? Just coincidence.

Regarding your question on that - I think it’s pretty obvious the lateral movement would need to be the cog-cog distance, not just the spacer width, no?

As an aside; I’m assuming your unconventional use of “doubt” is just because you’re not a native English speaker?

maddog34

one cog width plus one spacer, per shift..
changes in the cable angle changes the pull length, per shift.
Trig will give the exact distances of each cable pull needed, per shift, as the angles involved change across the shifting range.
i've torn apart a few Shimano trigger shifters.. the detents are spaced differently from one gear position to the next. gear position.
the distance from one detent to the next gets longer as the derailleur moves inward on the cassette, ergo Longer Pulls...
the derailleur's movement, per pull unit, changes, as the angles involved change.
the derailleur's movement can be represented by a series of triangle alterations.

i moved from engineering to Machine tech. when i was barely 14. i found the endless math review boring... i chose to Make Things instead of push a slide rule all day and night. Engineering Calculators were an expensive toy of the rich then... the very first TI Eng.Calc. models hit the market the next year... we could rent one for $10 a month.

see: Sine, Cosine functions
https://en.wikipedia.org/wiki/Trigonometric_functions

when represented in "an arc" as Kontac alluded to, as the Included angle increases, the sine length change decreases.
it's been 50 years since i first played with trig... i have, fortunately, forgotten almost all of it. 50 years wipes away most bad choices.

some of the actual metal pieces i've created are featured between two custom formed teak railings on the Pope's restored yacht..."the Candlesticks".. others may be in an antique computer's hard drive someone has saved from the scrapper, and others may have helped a dentist bore holes in your teeth, but they weren't the toolbits. And SOME of the things i've made may have gone over to Iraq for use during Desert Storm.

Ever driven a Sheridan Tank? I have.. they WHEELIE if you pin the throttle in first, then slam second gear. YEE HAW! Tracks up 4 feet in front... 21 tons of Aluminum Fun... the landing hurt, but it's worth it!
My Rally car was way more fun... i miss that Blue Mag00. it always did the right thing at the last second... or was that the driver... hmmmmm....
the ski jump in my former front yard was a hoot too... 60 feet elevation drop... i bet trig could be used to calculate the actual airtime distance... meh.. take another lap, screw the math.
See: "Washougal MX Park" i wonder how Mark is doing.. and if his son still has his old phone number.. let's see.. Jeremy McGra.... nope, it changed two decades ago. Jeremy move on into desert racing trucks......
i treated Travis Pastrana to his first ride in a Rally Car.. We caught air on Borin Rd...I then went to dinner with his mom... a hole-in-the-wall mexican place in Camas, with great food..."two small pitchers of cherry-lime margaritas, please!"... ahh. memories.......Mom made sure Trav was wearing his helmet during his Rally Car introduction.....
I saved Kevin Windham from death once... he was riding a crate bike for a brochure shoot.. Ally hadn't checked the Spokes.. they were DEAD LOOSE in the Back wheel.... the story board had them flying off the Uphill From The TOP... a trip to my shop and an old spoke wrench saved the day... i still have that wrench...it handles up to 8mm nipples, IIRC.

we miss what has been, and crave what will never be...

i've been instructed to ignore trolls... so i will.

nap time... night, all.
30 years ago,i'd have finished stowing 25 mics,50 or so cords, and be thinking about whether to have dinner or breakfast, this time of night... the up-and-coming band would be heading out the door.to their van or bus.. Watch out for whipettes on the dance floor!... i taught Dave Grohl about stage monitors one evening... and i got the distinct pleasure of physically tossing Courtney "love" out into the street, in a pouring rain. She cut her cheek when she landed. Blood.. Mascara.. Rain. Torn fishnets, and a skinned Knee.... It was the SECOND time that we'd tossed her out that morning.... SE 47th and Hawthorne
Ahh...... memories.

did you know that Rented '97 Cadillacs shut down and lock the tranny in Park on Really, REALLY steep uphills? and Rented Pontiacs shed their front valances when you try to clear the Chuck Sun tabletop jump in them....




.

choddo

Yeah I understand trig. I did a physics degree. Understand the principle (you’d need some calculus to get that airtime calculation right though ) - so the pull ratios described by that blog are fixed for a Campy 8S but the physical cable pulled changes, thanks to the ratchet spacing, as you go through the range to maintain a fixed cog-cog movement distance laterally at the business end of the mech. Makes sense and yeah the OP needs to account for it.

RB1-luvr

when shifting up the cassette, do the shifters pull the derailleur a little farther than the next cog and then release it back to the exact distance to that cog once the chain is there? If I'm correct, then that would need to be considered, no?

Kontact 

You can account for it by doing it directly. You don't have to calculate it in advance. And you'll find that the actual index positions are influenced by cable stiffness, do you'll find the math model inaccurate.

No, over shift is not necessary.

Andrew R Stewart 

Measuring 9 Speed Movements - Bike Forums

I made a cable and ser movement measuring set up here. Andy

Buzzkill53120

I was thinking the same thing. You're measuring 1st order effects: the final resting position of the RD, amount of cable pull, and final cable tension. I think taking measurements during the shift process will catch certain dynamics in the process of shifting. There may be some amount of momentary overshifting before it settles into its final position. This overshifting may be designed in the the indexing system and may differ for high/low gear combinations.

Of course this would a higher level of instrumentation as time intervals are sub-second with cable and RD sub mm. With type of measurements needed doing a GR&R would be required.
Having said all that, I'm not sure what the OP trying to do with his data collection, I may be looking at his question using a scalpel where a hacksaw is the proper tool.

DiTBho

this is precisely another doubt I had. I made approximate measurements in static conditions, that is, I put a marker on the cable, a mark point on the frame, a mark point on the cable near, removed the wheel, and shifted forward one position, from cog1 up, and measured. This is for all 8 positions.

I am specifically thinking of creating a dynamic test bench, i.e. with a pulley in place of the left crank, so as to be able to simulate the gear change during a pedal stroke, even if in ideal conditions, i.e. without considering the weight force, and possibly its component along an inclined plane during climbs.

(although you should NEVER change uphill)

The software and hardware part is not a problem, it's my job, the real problem is finding sensors that don't cost a fortune, and then pair everything up, the mechanical part isn't my thing.

Kontact 

Again, no overshift needed.

DiTBho

Intrinsically do I introduce it because the linear electroactuator that I am designing is a feedback system, and as such it adds an overshoot to the response to the step; that is, when you go from cog_n to cog_n+1, there are brief moments of time where it dances around the final position.

Obviously I have to reduce the overshoot as much as possible, but the price to pay is a slower shifting system.

Dave Mayer

I've ignored all of the previous posts in this dysfunctional thread due to the excesses of ego and drama.

Answers: Campy 8-speed cog spacing is a uniform 5.0mm cog center to center. The actuation ratio of this generation of Campy derailleurs is 1.43 : 1. That is, for every 1mm of cable pull, the rear derailleur moves on average 1.43mm. Which by division, it means that the average cable pull per 'click' is 3.5mm.

Yes, I know these are averages, but if you use these numbers to test for derailleur/shifter/cogset compatibility, and the math works, then the system will work flawlessly every time. I've matched up all kinds of official and unofficial indexed system combinations, and if this simple math works, then the indexing will work.

Cable tension: I haven't measured. If you are running 8-speed Ultrashift shifters with compatible derailleurs, then you don't have to worry. The Ultrashift derailleurs have stiffer springs, and work better with the matching shifters. As far as the Powershift and Escape models: you don't want those.

DiTBho

dysfunctional thread?
excesses of ego and drama?
here everything is fine, on my side.

thanks for your info, anyway!

DiTBho

OK, thanks guys, I think I have enough information to go ahead =)

Dave Mayer

You weren't the problem.

maddog34

yep.. yet another thing extensive product development brings to us end users.

maddog34

He wants to build his own shifter, Dave. (It will never be built.) And you've just provided a few generalities that will feed his bad choice enough to keep it alive.

Kontact 

Oovershoot is slower, because you engage the ramps on the next set of cogs.

DiTBho

lol!

AeroGut

Just a side comment for those confused by the usage of “doubt” - in Indian/south Asian English it is common to use “doubt” as equivalent to American/British English “question”.

maddog34

"the linear electroactuator that I am designing.."

the Goalposts just moved.

so it's a hybrid system using a linear servo?
why bother with cable at all? will this simply add a servo and electrical link in-line?

build a linear derailleur instead.
you'll be needing that trig calc. for the proper movement since the cage needs to follow the cassette angle.

here's an idea... make your actuator adjustable to allow near-perfect following of ANY SIZE CASSETTE. The actuation distance will change with input of the Angle required... the controller would change the shift distance according to.. what a trig calc. tells us online.

that will be a true advancement of technology... that other, well funded, R&D folks in white lab coats are already hard at work on, i'd bet.

i think back to Honda Tech school in Torrance, California... every day a team of white lab coat guys would be out mowing numbered patches of the huge Lawn at the facility... they were testing new mower deck and cutter designs.. they'd bring out a chair and a small desk to sit at as they made notes on clipboards we used at Car races, equipped with a stop watch... Magnifying glasses hung from all their necks. one day they were measuring mulch size... the next they'd be examining cut quality... etc.
Shimano will have similar team of folks doing their R&D work.
a Lack of R&D folks and funding is believed to be why Suntour failed.

sorry for jumping ahead so far in this process of futility.
just being real with no candy coating

your "idea" has already been built.... by folks that know how far it is between the "shifts" on a cassette drive system.

choddo

The mobile phone had been built by some pretty sharp individuals before Steve Jobs had a go at it

maddog34

ahh, the Brick....wasn't that a Motorola thing ?
has Sharp ever made a cell phone? not sure how "pretty" their engineers are, tho.....

and then, there was that "mobile phone" that was an INSTALL into the cars.. it even had a classic rotary dial and handset.

i hear Flip Phones are regaining market share.. some of us like to have a phone but don't need all the bells and whistles.

See: DI2 Shimano.
Also see: E-tap SRAM

any bets on if they're moving beyond the parallelograms soon?
i bet the development is nearing pre-production testing by now....

smd4

And it probably isn’t necessary on most rear derailleur pulleys. At least if my 7700 Dura Ace RD with Omni Racer aftermarket pulleys can be believed.

Kontact