Thread Tools
Sep 10, 2019, 07:18 AM
Registered User
Thread OP
Discussion

RC servo for bycicle


Hello, hope I'm posting in the right forum. I am planning to make an automatic gear shifter for the rear derailleur of my bicycle using an Arduino and (initially) an accelerometer. I've been thinking about a system to pull the metallic cable of the derailleur and came across RC aircraft servos. Now, my question is, would such a servo be strong enough to pull the cable? I searched everywhere, but I didn't find any information about the force required to pull the derailleur, and I also need to mention that I haven't worked with RC servos until now. As for the derailleur, I have a Shimano 105 7500 if it helps anybody. Sorry if I messed the physics terms, but I am not very familiar with their English version. Thank very much!
Sign up now
to remove ads between posts
Sep 10, 2019, 12:57 PM
Registered User
I'd say take a luggage scale and look how much force you need.
Then measure how far you need to pull.

You could then look for a servo arm that gives you enough travel, and see how much force the needed strength is in Ncm.
You could then buy a servo that has at least that strength, although i'd go for a lot more to be safe: most servos don't like pulling all the time, get hot and burn up.
The rest is supply and programming.

I'd say a linear servo would be a selfmade better approach:
- holds without power (depends on threading)
- can pull "infinite" lengths of wire
- high force if the thread is fine enough

It can be driven by stepper with some reference light barriers, or with a linear pot as position sensor and a normal motor.
Most parts of it would be similar to those used in 3D printers.
Sep 10, 2019, 02:01 PM
Registered User
vollrathd's Avatar
Quote:
Originally Posted by testoasarapida
Hello, hope I'm posting in the right forum. I am planning to make an automatic gear shifter for the rear derailleur of my bicycle using an Arduino and (initially) an accelerometer. I've been thinking about a system to pull the metallic cable of the derailleur and came across RC aircraft servos. Now, my question is, would such a servo be strong enough to pull the cable? I searched everywhere, but I didn't find any information about the force required to pull the derailleur, and I also need to mention that I haven't worked with RC servos until now. As for the derailleur, I have a Shimano 105 7500 if it helps anybody. Sorry if I messed the physics terms, but I am not very familiar with their English version. Thank very much!
RC servos would be a bit of a misapplication for the derailleur shifting. But, check out those robot servos that have much higher force capability. In the USA, ServoCity has a variety of them. Here is one with a 44 pound force capability.

https://www.servocity.com/140-mm-str...y-linear-servo

And, geared servos. Servo City can provide these wired up ready to go.
https://www.servocity.com/servos/servo-gearboxes
https://www.servocity.com/cm-785hb-servo-gearbox

PS
I've played with those accelerometer some years back. You may have issues trying to separate the normal bouncing of your bike over the road with the much lower acceleration provided by pedaling the bike itself.
Sep 10, 2019, 03:45 PM
Registered User
Thread OP
Thank you very much for the input! I will for sure check out those robot servos (also, being linear makes much more sense). I'll see what I can do about those vibrations - an idea would be to get a speedometer or, even better, to build a tachometer for the pedal gear.
Sep 11, 2019, 07:30 AM
No bounce, No play.
davidmc36's Avatar
Tach on gear train is easy with optical interrupter. Will only shift by RPM. Up hill with more pedal force will desire shifting sooner. Tilt sensor may help.

How will the shift take place if you don't know when to ease off the pedal force for the Auto device. Maybe an indicator when it desires shift.
Sep 12, 2019, 12:53 AM
Registered User
vollrathd's Avatar

Servo shifting both crank and rear wheel sprockets.


Quote:
Originally Posted by davidmc36
Tach on gear train is easy with optical interrupter. Will only shift by RPM. Up hill with more pedal force will desire shifting sooner. Tilt sensor may help.

How will the shift take place if you don't know when to ease off the pedal force for the Auto device. Maybe an indicator when it desires shift.
Yeah, the bike crank RPM makes a bit of sense. If you're climbing a hill and are slowing down, the Arduino could sense that and move the linear servo to a lower gear. You'd also have to have the software do nothing if you stop pedaling the bike crank. Also, some method of not changing the derailleur while pedaling hard.

The software might be a bit interesting on my bike. It has three gear sprockets on the crank, and 8 gear sprockets on the rear wheel. There are short hills in my neighborhood where I need to go to the lowest gear ratio between the crank and the rear wheel. And with that, the crank is turning 20% faster than the rear wheel.
Sep 12, 2019, 03:27 AM
No bounce, No play.
davidmc36's Avatar
The actual force to pull a cable should be relatively light. The shifters on our 15 yr old bikes take little effort and flick one gear at a time.

The degrees of throw needed is not much. Any "quarter scale" servo with an inch or two arm on it will slam shift that bad boy all day. When it shifts it is a flick and release motion that can for sure be done one direction with no resistance at the shifter. Going to smaller gear drops easy. I think pulling it up to larger gear would do one click at a time with "stretch" on the cable until you back off the pedals a wee bit. I say that because if you are easy on pedals you can move it three gears at a time with full thumb stroke. That's like 40 or 50 degrees. Newer and better gear than mine must be even better.

So arduino flicks the servo and lights a red indicator. You back off the torque. When the derailer moves position a green indicator lights. Much of the time you will get a feel for upcoming shift points. Just a bit of pedal slack and the gears with tapered corners drop reasonably easy.

Make some pipe mounts on the handlebars and sludge up the mechanics to proof of concept. You could shorten the cables and mount it all on the frame to be tidy.

Totally do-able.
Last edited by davidmc36; Sep 12, 2019 at 03:45 AM.
Sep 12, 2019, 08:10 AM
AndyKunz's Avatar
Quote:
Originally Posted by vollrathd
If you're climbing a hill and are slowing down, the Arduino could sense that and move the linear servo to a lower gear. You'd also have to have the software do nothing if you stop pedaling the bike crank. Also, some method of not changing the derailleur while pedaling hard.
I don't understand why a thumb isn't the best activation here.

In drag racing, automatics won out over manual shifting because of the change to bracket racing. It went from being a drivers & mechanics sport to a mechanics sport - the mechanic who could best set up his car for consistency would win, rather than the driver who could best shift with the best mechanic's equipment. So, while I don't like the change, I certainly understand it.

But on a bicycle, things are happening so slowly, what's the reason for an automatic?

Andy
Sep 12, 2019, 10:45 AM
Registered User
Quote:
Originally Posted by AndyKunz
But on a bicycle, things are happening so slowly, what's the reason for an automatic?

Andy
Because it's cool? Because we now can? (just kidding, I have the same reaction to a lo of questions here... but I'm just showing my old age)

I have built a lot of things that really make no sense. Personally I would not like to have a true automatic shifter on my bike, since I prefer to decide what level of effort/gear to use depending on what comes next, and no amount of sensors (short of a look ahead camera with "real" AI) can decide it for me. But I can see how it would be a fun project that will end up being published on sites like Hackaday.com
Sep 12, 2019, 10:51 AM
AndyKunz's Avatar
I enjoy the "because it's cool" just as much as the next guy, but those are for personal mental challenges that rarely anybody but me would see in the end, and definitely fewer still to see before it's working. Asking others for input would totally eliminate the personal satisfaction of it for me, although I might show the final result to others so they could geek out on it.

Something done publicly, for me, is something with a genuine purpose.

Don't know if others think like that, but I know I do.

Andy
Sep 15, 2019, 05:27 AM
I bail out, anywhere, anytime
Taurus Flyer's Avatar
A bicycle has a general accepted model but for sure not the highest efficienty. First disadvantage is aerodynamics. To travel long distances I use my recumbent, for local activities a normal bike and travel with car or train a foldable bike.

No derailleur but 5 and 3 speed gear systems. Maintenance of this system combined with a more heavy, read better, chain is preferable above a derailleur. Cruising speed with 1:1 transmission for efficiency. Also important is , the ability to change gear at zero speed position, waiting for a traffic light for example, thŗtís not possible with a deraileur. Also an important point whatever system you use, when changing gears you have to reduce the 'propulsion force' at the pedals, if not you will ruŪn the system, hub gears or deraileur. Cars do have an hydraulic con/transverter for that. I had a Peugeot 'Familiale' with three speed gear and was able to pull a little truck out an Ūn grass blocked position, the Peugeot standing on road surface. Speed, nearly zero km/h.

Wel, next step als in the past could have been auto gear changing but, thatís not depending of arduino, it would have been possible already long time ago. Arduino is no magic system although many people think.
Automatic? Our human brain can anticipate at much more inputs than we think, switching back in a lower gear when seeing a traffic light, road surface, physical condition etc and Human gear shifting IS human powered, thatís we want anyway and we have it all at our bike. All added xx will make a bicycle heavier, complicated, critical, more drag, etc. Because of the lower aerodynamic drag of a recumbent also a lower amount of 'gear speeds' is needed, combined with the high rpm range of a human propulsion system.
See my recumbent, traction and steering with my legs, gear shifting and braking with my hands, 5 gear Sachs Pentasport (Low friction hub gears) , cruising in the third speed, 1:1 transmission at 15mph. Simpliest system, short chain, low maintenace main drum brake, relaxed position, highers average speed than an 'upricht' position bike.

Taurus Flyer
Last edited by Taurus Flyer; Sep 16, 2019 at 06:50 AM.
Sep 18, 2019, 10:19 AM
Registered User
I have connected a servo directly to the rear mech ( return spring removed) and it works fine - used a PICAXE chip to control it as its very easy to program. Its not even a powerful servo just a small 3kg jobby.
Sep 25, 2019, 02:01 AM
Registered User
Rumcajs's Avatar
Something like that?
https://www.instructables.com/id/DIY...ic-Derailleur/
Sep 25, 2019, 10:23 AM
Registered User
Quote:
Originally Posted by Rumcajs
yes exactly like that, but with picaxe instead of arduino


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Anyone with recumbent bycicle experience? curly56 Life, The Universe, and Politics 42 Feb 25, 2019 01:24 PM
Discussion Small powering a bycicle? satanicoo Power Systems 11 Apr 11, 2009 11:46 AM
Help! which size battery is best for RC bycicle magiclls Batteries and Chargers 7 Dec 29, 2008 04:43 PM