Ready the flashing lights, party poppers, dancing girls and brass band...

[Kim]

*giggles* at Kim's latest fettleage.

The really sad thing is that it works surprisingly well.  

I'd like to say how lovely it has been to see people's kind comments, generosity and sharing of ideas.  I'm full of the ill so haven't had much coherent to say other than I hope I'm better enough to make the most of my visit to DTek on Tuesday.

+1

[clarion]

Sporktastic!

[andygates]

That's beautiful, but it needs little tiny corks on the tines.  Health and safety!  Think of the children!

(zipties > epoxy, bah!)

I'd like to say how lovely it has been to see people's kind comments, generosity and sharing of ideas.  I'm full of the ill so haven't had much coherent to say other than I hope I'm better enough to make the most of my visit to DTek on Tuesday.

We just love a challenge.  And bikes.  And probably most of us know at least one rider with a gimped something-or-other and a wacky mod.  

[Arch]

That's beautiful, but it needs little tiny corks on the tines.  Health and safety!  Think of the children!

Corks? Bah. Little cubes of cheese. Or Chocolate or pineapple, or other foodstuff depending on digestive ability.

Then you have emergency rations too.

[redshift]

OK, so we have a prototype.  Now what we need is a dismantled Shimano shifter for the mount pattern, some nice light Ti bar, one of Mr Gates' friend Jarkman's kitchen table Proxxon CNC millers, a gas torch to do the pretty colours, and some ordinary household bleach*, and we have a plan! 


Seriously though, I really have to get the workshop built, don't I?  Custom length machined bar end shifters can't be that hard, can they?





*OK, I lied about the bleach, it's a Young Ones reference.

[TimO]

A modification to a Shimano Di2 changer could be an option, but given the groupset costs >£1500, you probably wouldn't want to go that route for experimenting (and hacking apart!)

I'm not convinced that you could get enough force to reliably make a derailleur change with a servo, and it may not respond as well to slight misalignments as a manual mechanical system (apparently the Di2 self calibrates, which is neat).

As far as I can see, the only currently commercially available electrical gear system is the Shimano one.  There have periodically been others over the years, but they've mostly died a death for one reason or other, which is a shame because this would be the sort of situation that they'd be ideal for.

Have you considered any of the possibilities of hub gears, or even single speed?  I suspect a single speed recumbent may not work all that well, since you can't really get out of the seat and honk like you can with an upwrong, but it might work on relatively flat roads.  Alternatively, you can limit the complications and issues of two derailleurs by only having, or only using one.  For a lot of the time you could possibly get away with the rear changer only, especially if you fitted a wide ratio cassette to it.

An S2C equipped recumbent would be an interesting approach though, since you'd avoid the need for a brake or changer (although you'd really want the brake to disappear from the front wheels, not the rear).  There's possibilities for a trike with front drive and steer, with an S2C, but front drive and steer trikes are pretty rare!

[LittleWheelsandBig]

Perhaps a 3 x 9 hybrid system with the hub controlled by a Sturmey trigger (with extended trigger?) and the derailleur with a twistgrip?

[Kim]

Yay, the mad science suggestions.   

I'm thinking that given the right choice of chainrings and a wide range cassette you can indeed stay on the middle ring the vast majority of the time.  That's pretty much how I've got the gearing on the Streetmachine set up - the granny ring's only needed for really bastard hills or with luggage, and the big ring can be safely ignored entirely below about 20mph (barakta doesn't have any cycling fitness currently, so going fast or caring about being between gears isn't likely to be a problem).  Keep front shifts to a minimum and you can get away with a front shifter that's not fantastic ergonomically, which we're thinking is a reasonable compromise given the cost of wide-range hub gears.

We're currently thinking that twist-grip for the rear and extended bar-end for the infrequently used front.  There's no shortage of trikes with twist-grip shifters as standard, so we'll see how she gets on with that on Tuesday.


The mad science part of my brain wants to know if you could make a self-adjusting electronic shifter by measuring rattle in the mech with a piezo transducer in a manner akin to an engine knock sensor.  The sensible part of my brain says that I've got far too many other half-baked projects on the go which I actually stand a chance of making work.

I am tempted to try a servo-inna-box though.  It'd probably work about as well as the insane cable run on my folding BSO...

[tonycollinet]

Surely a servo wouldn't need to be any more self adjusting than an indexed gear system - as long as the cable pull was reasonably accurately controllable.

Alternativey you could have length of pull defined by twiddly pot, and have it non indexed.

Is gonna need bigger servos than those generally used for R/C modelling though I would think (*thinks about force needed to pull mech cable compared with R/C aircraft elevator*).

[Beardy]

I'm leaning towards a possibly inverted, possibly full-length (indexed) grip shifter for the rear mech and an extended bar-end (friction) for the front, but we'll see how it goes on Tuesday when she actually tries to ride something.

If you are going for the full length grip shifter, tehn there is no need to invert, which keeps the top of the handlebar free for easier griping and releasing. I was looking at this solution for the Grasshopper,, but in the end went for the cheap and messy option of inversion.

[rower40]

There was no way that Spork-n-Shift^(TM) was ever NOT going to make Post-of-the-Day.  Very well done.

[Tigerbiten]

If your going for a twist grip, then a twin bar setup like mine won't work.
There's not enough room between the twin bars to hold the twist grip and twist it at the same time.
The back of your hand will hit the second bar to often.

I tend to hold my main handle-bar high enough so that the back of my hand does not hit the second bar.
This makes the palm of my hand around the pivot level of the bar-end shifter.

What time are you planning to get to Kevin's ??
I'll bring my Trice Q over so you can see/study the setup.

[Kim]

If your going for a twist grip, then a twin bar setup like mine won't work.
There's not enough room between the twin bars to hold the twist grip and twist it at the same time.
The back of your hand will hit the second bar to often.

I'm thinking that twist grip + bar end means that a twin bar won't be necessary (at least, not on the right, where the controls are.  It might be a sensible way to give more reach on the left).

Thinking about it, there are going to be cable-routing issues with doing that.  I know that drilling holes in handlebars to route cables is considered harmful on upright bikes, but what's the verdict on a recumbent trike, where they're only going to be subjected to steering forces?  My gut instinct says it'll be fine, and in any case the consequences of failure would be less nasty, but my gut is the wrong kind of engineer.

What time are you planning to get to Kevin's ??
I'll bring my Trice Q over so you can see/study the setup.

We're aiming to get there between about 12 and 1pm, depending on how much the car hire people decide to faff me about (they seem to be deliberately sloppy about having the cheaper cars ready on time).

Do I sense an excuse to come and play with the shiny toys? 

[LittleWheelsandBig]

Drilling the bars of a recumbent trike would be fine. Just ovalise the holes (1 near barend shifter, 1 past the grip) and avoid stressraising notches.

[Quint]

I don't understand the need to drill the bars, thought cable from bar end went down outside of bar (as does of course the twistgrip cable

[Kim]

I don't understand the need to drill the bars, thought cable from bar end went down outside of bar (as does of course the twistgrip cable

Exactly.  The cable from the bar-end goes down the bar, where it meets the twist-grip, and

[LittleWheelsandBig]

How do you get the bar end lever's cable past the gripshifter without getting in the way of your fingers? The answer is to run it inside the bar but the lever starts the cable outside the bar, hence two holes needed, one in the grip region.

[Cunobelin]

This may sound silly but bear with me.....

On recumbents the standard twist shifter tends to be operated by the heel an palm ofvthe hand. it is possible to fit them the other way up so the thumb grip is used tomtwist.

This may need some revision of cable lines, but it may be worth looking at which way up provides the better and more comfortable change

[Cunobelin]

OK, so we have a prototype.  Now what we need is a dismantled Shimano shifter for the mount pattern, some nice light Ti bar, one of Mr Gates' friend Jarkman's kitchen table Proxxon CNC millers, a gas torch to do the pretty colours, and some ordinary household bleach*, and we have a plan!  


Seriously though, I really have to get the workshop built, don't I?  Custom length machined bar end shifters can't be that hard, can they?


*OK, I lied about the bleach, it's a Young Ones reference.

Call yourself engineers -where is the TEA

[Quint]

Duh, must remember engage brain before typing    

[Kim]

This may sound silly but bear with me.....

On recumbents the standard twist shifter tends to be operated by the heel an palm ofvthe hand. it is possible to fit them the other way up so the thumb grip is used tomtwist.

Yup, we've already considered this:

https://yacf.co.uk/forum/index.php?topic=45696.msg902191#msg902191

And I think it's pretty clear that nothing's too silly a suggestion for this thread


I suspect that an inverted twist-grip shifter is going to make a bar-end shifter on the same bar rather tricky, though (well, it's going to need an inch or so of bar for its cable to disappear into a hole above the shifter).  The cable issue can probably be handled by a nice wide loop - which will have to clear the brake lever - secured with cable ties lower down the bar.  Nasty-looking, but might work.

[tonycollinet]

Is gonna need bigger servos than those generally used for R/C modelling though I would think (*thinks about force needed to pull mech cable compared with R/C aircraft elevator*).

OK, I am wrong
Servo Shop (19Kg.cm)

And that is a comparative baby.


EDIT: however, curent draw for such servos would seem to rule that sort of system out. (2A)

[Cunobelin]

It is so nice spending other people's money!


Nokon cables are not cheap, but can allow some sharp bends unachievable with standard systems

[Cunobelin]

Depending on the trike then it is possible to mount a twist grip laterally on a stub, and the bar end above this.

I have seen it done with the stub below the grip, although it does mean moving the hand to change gears

[Kim]

OK, I am wrong
Servo Shop (19Kg.cm)

And that is a comparative baby.


EDIT: however, curent draw for such servos would seem to rule that sort of system out. (2A)

2A's no problem for a vaguely decent battery, and a gear change only lasts a second or so.  You'd need a reasonably high-capacity battery and/or something to keep it topped up, but that's doable.  Hmm.