RB1-luvr

^^ speaking of Mavic, I heard they are establishing a US office in my state (CT).

DiTBho

Yup, but the mavic derailleur is something that I am not able to build on my own.

At the very least I need the help of a mechanical engineer, an industrial designer for the geometries (because then we don't want to make an ugly aestetically mess, right?) and then there would be to understand how to produce the various aluminum parts, which not only is *very* expensive, but also it's not at all simple if you are a hobbyist.

To all those who think: can't you print it?

Unfortunately ... no, in plastic, with any type of plastic, the anwser is no! Aluminum or steel is needed, and unfortunately metal printers (or pantograph micro welders, experimental, but starting to work) are not yet within everyone's reach, they cost far too much, in addition to the fact that the print must be subjected to various heat treatments, which not only add costs, but also a lot of compromises, both aesthetic and functional.

On the contrary, the twp electro linear actuators that I am building use gears, motors, screws, nuts and parts that you can purchase in stock, there are only a couple of custom cylinders that I had turned to a company.

The nice thing is that in the end I will be able to ***pair*** the linear electro actuator with my NOS Sash rear derailleur, rather than with my old friend Campagnolo Mirage/8sp, without having to hack either of them.

Which *IS* the practical bonus factor of this choice!

choddo

You will post a video when this is complete of course?

Kontact

You will have perfect shifting if you manually adjust the electrical cable pulling device to center under a cog and then make that a set point. After you have saved each position, the only other consideration is the actuator speed.

You don't need to program a geometry, load curve or overshift. Some forces will be absorbed and moderated by the cable/housing and jockey pulley.


I would be shocked if this basic approach yielded anything less than ideal shifting for that cassette.

maddog34

https://www.horizonhobby.com/product...PMSA2030L.html

etc.....


and the spring in the derailleur would become irrelevant using a standard linear actuator, AKA: a Servo.
also available in a Bell Crank configuration.....

Kontact

Hard to control a derailleur using a push rod considering that the cable bends depending on the parallelogram angle.

I suppose if you remove the barrel adjuster you might be able to made the rod stiff enough to control the parallelogram through a pivoting arm, but the side load from the arm is going to make the servo rod bind.

maddog34

i'd tell ya the secret, but you'd just fail to comprehend.
PS.. you are on my ignore list for good reason.

Bye.

Kontact

I address the insulting way you talk to people because I hope you become so embarrassed that you stop.

And I'm smarter than you, so I doubt there is anything you have to say that I could not comprehend.

DiTBho

Houston, we have an ego problem here ...

DiTBho

I started a YouTube channel a long while ago, created a nice icon and a short description of the channel, etc., but I haven't uploaded anything yet.

I have little time due to business travel, and in my free time I also work on other projects, much more GNU/Linux or the wild building of electronic gadgets, rather than mechanical things.

But the idea is precisely to film the things that work and make videos of them.

Apparently bicycles, in particular my old 90s road bike, have nothing to do with that type of content, unless the format is rethought.

In my ideal wishlist, the derailleur project is just a small piece of a larger project, which could act as a bridge and solve the coherence problem.

I don't know, we'll see.

at the moment, the priorities are

• build a pair of new wheels, because I broke my Campagnolo Omega rims
• measure how much force is needed to overcome the rear derailleur spring
• finish sizing the electric actuator
• mount it
• test it
• buy a camera, as my smartphone sucks at filming
• buy a microphone
• then we move on to the control MPU and the firmware
• finally prepare a test bench and see how it derails without load
• then test it on the road with a real load
• buy a Mac/Apple Silicon with a copy of Final Cut installed as I'm pretty ignorant about video editing, but I found Final Cut to be the most intuitive choice I've ever tried

So, we are talking about the far future (mid 2024?)

smd4

You forgot to finish with “neener neener nee-ner!”

Kontact

What's the coherence problem?

smd4

Is this a trick question?

DiTBho

On Youtube, if you open a channel where you talk about "programming and electronics", it makes absolutely no sense to talk in such a crude way about the "mechanics of bicycles" because it has absolutely nothing to do with it; sure you can do it, but then you'll just get a lot of slaps.
You should open two channels, but honestly I don't feel like it. So I need to find a format that can combine the two things, with coherence.
.

maddog34

eliminate the spring, and eliminate the cable.
connect the linear actuator directly to the Derailleur.(ask a mechanically inclined person about that.. it's EASY! And Remote Control Airplane models use it a LOT.)
the control can be programmed to step up or step down, yes?

except, like i've pointed out before... these electric derailleurs already exist, and are now for sale as seasonal parts swap, in good used condition... pro racers swap parts to reduce the chance of failures during races.
i swapped parts at pre-decided periods to prevent failures.. clutch parts, brake parts, driveline u-joints, STEERING BALL JOINTS and TIE ROD ENDS (hint), etc... I miss my Rally car... i do not miss the major expenses and hours upon hours of preventative maintenance.

See: RC Car parts

you are struggling to re-invent the wheel, when there is a Goodyear tire store near you.... to use an American term.

imagine the market demand for a kit that could be retro-fitted to an existing bicycle drivetrain's derailleurs that eliminated the fraying cables and sticking shifters....

Kontact

You aren't the only tech/bike nerd.

But if you have a method to use off the shelf parts, simple programming and an '80s index derailleur to allow folks to homebuild electronic shifting for cheap, I think you'll have a rapt audience for both the technical details and the sheer audacity. So don't sell it short.


As far as servo forces go, consider a pre-loaded counter spring that helps the servo pull against the derailleur spring.

DiTBho

Do you think that an EPS is simply a "RC-servo" meccanically coupled with a derailleur from which they have eliminated the return spring?

Sure, they're more efficient, but EPS is not even close to such a thing. Both the mechanics and the electronics of the EPS is already and specifically designed for the internal linear electroactuaror and its feedback sensor, which is made ad-hoc, and thinking about hacking an EPS, that is, understanding how to drive it from the electronic cable is anything but trivial given how integrated they are, and given that they use electronic parts for which the technical documentation is difficult/impossible to find.

Sure, you can decode at least the protocol they use, but there's no guarantee that you can practically complete the task, both for reasons of time available and for reasons of tools.

From the outside, an RC servo for "RC-Cars" is controlled in PWM, with a period in the order of 1ms, so you have to choose high precision ones, and it then becomes unnecessarily complex on your controller part as that type of servo is fine for making very fluid movements, but it's problematic when you need precision and repeatability of the movement, something that is instead essential in a derailleur.

DiTBho

but then sorry, do I eliminate the cables? What?!?
Are you one of those who would buy wireless cadence and speed sensors?

Like those kits by Sigma?!?

Absolutely not, I want cables! And, call me paranoid, but I want cables in sight, so no one trying to hack them over "Bluetooth air-hacking"!
(like we do with computers and smartphones)

choddo

There are millions of people using wireless groupsets every day including the pros. Admittedly, those big manufacturers will have spent a lot of time on getting the encryption right.

DiTBho

I don't trust it enough, also I don't like to have several batteries(2). You have to periodically check the charge, you have to replace. Too annoying for me.

I khow encryption works and how there is always a way to exploit some software or hardware bug, so I don't want to spend more time on something that needs extra care to only be at best "reliable enough" with always a margin of risk.

Now they sell and advertise on YouTube and Facebook an orange device (I won't tell you the name) that also hacks traffic lights and allows you to pass the subway barrier for free(1), and there are a lot of idiots who buy it and then go around to disturb people.

Imagine if you met one of these idiots who, for a laugh, made you change gear uphill, perhaps in one of those crossed positions to avoid.

No, it's not for me! My philosophy is: don't keep more ports open on your router than you really need, never do something wireless if you can do without it.


(1) it's not hypothetical, they did and filmed it, and uploaded it to Youtube... damn how it depressed me to see this idiotic use of technology. Even worse when they used that device to hack coffee and drink vending machines, with the result that the company that manages them got pissed off and increased the price of the drinks.

(2) batteries:
- one battery for the wireless speed meter?
- one battery for the wireless cadence sensor?
- one battery for the wireless rear deraillerur?
- one battery for the wireless front deraillerur?
- one battery for the wireless front led?
- one battery for the wireless rear position light?
how many bloody batteries do you have? and how many wireless channels do you use?
no way, all wired, only one battery for everything!

79pmooney

A comment and a suggestion: Those Mirage RDs - with 9-speed Campy cassette and SunTour Superbe top-mounted friction DT shifts, wow! Friction shifting doesn't get better. Have one on my best bike with a triple and 12-28 cassette and it shifts perfectly.

For accurately measuring cable pull - tape a piece of wood or whatever with a clean, sharp edge just the thickness of the DT to cable distance. Place a piece of tape on the cable at the edge of the wood with the shifter full forward. Now you can easily measure the cable pull accurately with calipers. (I'd go through the shifting sequence several times, plot all the cable distances on graph paper, run a smooth curve though the points and use that to make my shifter. If spacer thicknesses vary, I'd put elbows in my smooth curve at the cogs of the spacer changes.

mpetry912 

this is true, the upper pulley on the Shimano units has about 1-2 mm of lateral play.

/markp

smd4

Which isn’t necessary.

Kontact

Why do all that when you could just program the shifter by matching it to the gear positions? After you do that once, that map will be always be good for any compatible derailleur and cogset.


As far as friction shifting goes, I've found that even "poor shifting" derailleurs from the '70s shift crisply on ramped cogs. And the old derailleurs require less fine tuning for chatter after you shift because the increased pulley to cog distance acts like a floating pulley.

I think that if hyperglide had come out first, Shimano's index advantage in shifter or derailleur design would have been nullified. All the overshift and geometry design was done to fix the way normal freewheel cogs resisted shifting.

79pmooney

I thought I read in an earlier post that the OP was not a fan of batteries and wired/wireless shifting. I was just trying to help out with accurately measuring cable runs between shifts for someone who might have the machinist skills and consider it fun to make his own index system.

I am not a fan of Shimano's really easy shifting cogs for friction shifting because if I put the derailleur between cogs, the chain has a very disconcerting tendency to wander between the cogs as I increase or decrease pedaling pressure, almost always in the wrong direction. (Modern Shimano freewheels. I've never had their cassettes.) My Campy 9-speed cassettes almost always pick one cog and stay there or maybe, shift once. The shifting is cruder than say the Shimano FWs I've had on other bikes but that super smooth Mirage makes up for that and more. Likewise my older Sachs FWs never wander and do no more than shift once when I screw up. (When I want a very different gear, I expect to be able the just toss the lever forward of back, get out of the saddle and be in a gear that works and is within a cog or so of my intended. I took that for granted when I raced in the 70s and still do.