Twitchy steering fix for recumbents, velos etc

Logic11

100 W
Joined
May 2, 2022
Messages
278
Thinking about the twitchy steering on recumbents mentioned by neptropix

NB that at speed the steering angles through which you want to turn the front wheel are small.
It's only at low (feet down type) speeds that you want very large steering angles on the front wheel.

Look at my sketchy diagram depicting a recumbent from (almost) above:
  • For near staight travel, as used at high speed: For a large handlebar angle change; you get a small front wheel angle change. (= no twitchyness)
  • But at the low (feet down type) speeds at which you want large steering/wheel angles to maneuver or turn a sharp corner; the elliptical steering drive changes that to the opposite and you get a large wheel angle for a small steering angle.

Bike twitchy steer fix.jpg

NB: No change in belt or chain etc length at any angle, even if one of the pulleys/sprockets is circular...
= no chain/belt/etc going slack issues.

NB that as you don't need those ellipses to rotate more than 180 degrees; you don't need teeth on the ellipses:
You could use anchored wires of the type used in brake cables, or similar.

You also get your handlebars closer to where you can reach them without extending them backwards and ending up with that clumsy 'handlebars move off to the right when turning left' issue.

That solves a lot of issues for recumbents, space constricted velos, trikes etc.

(The world would be a better place if people didn't 'own' other people's ideas: What are the chances of further idea sharing when that shit happens?)
 
Early in my recumbent development I tried 14 different steering linkage experiments.
One of the main takeaways was any ratio change between handlebar and front wheel ruined the feedback to the rider.
It wasn't a good idea.
The smoothest system with the most steering angle movement was with aircraft cables and one to one ratio pulleys.

bare frame 7-600dpi.jpg
 
I've successfully improved or totally solved twitchy steering with a wider handlebar than stock on 100% of the recumbents i've ridden. If reach is a problem, there's also handlebars for that.

The wobbliness at low speed is easily fixed with the addition of an electric motor :)

On the one bike ( Maxarya ) i don't have dead confident steering on, i've created the problem by replacing a stock non-suspension fork with a lot of offset.. for a suspension fork with an average amount of offset.

I think it is possible to build a recumbent with excellent handling properties that doesn't require the user to hack on it it to get it to handle well. The Cannondale Bent is a great example of that. But this category of bike isn't purchased by performance enthusiasts often, so makers don't focus strongly on that.
 
I've successfully improved or totally solved twitchy steering with a wider handlebar than stock on 100% of the recumbents i've ridden.

Makes me wonder how much closer to the aero drag of a regular bike with historically average width bars that gets you.
 
With a lightest.bike mid drive that's currently capped to 900w, on the same battery and tire pressure, with max effort pedaling, the top speeds are:
Upright 26er bike ( tuck position ): 28mph
Maxarya Ray 2 ( leaned back ~2 degrees more than stock ): 35mph

25% difference in top speed on the same watts.

Full recumbents are expected to produce a 33% top speed advantage versus an upright in a tuck. But i like my semi recumbent anyway because it has good visibility on the road.
 
Thinking about the twitchy steering on recumbents mentioned by neptropix

NB that at speed the steering angles through which you want to turn the front wheel are small.
It's only at low (feet down type) speeds that you want very large steering angles on the front wheel.

Look at my sketchy diagram depicting a recumbent from (almost) above:
  • For near staight travel, as used at high speed: For a large handlebar angle change; you get a small front wheel angle change. (= no twitchyness)
  • But at the low (feet down type) speeds at which you want large steering/wheel angles to maneuver or turn a sharp corner; the elliptical steering drive changes that to the opposite and you get a large wheel angle for a small steering angle.

View attachment 363367

NB: No change in belt or chain etc length at any angle, even if one of the pulleys/sprockets is circular...
= no chain/belt/etc going slack issues.

NB that as you don't need those ellipses to rotate more than 180 degrees; you don't need teeth on the ellipses:
You could use anchored wires of the type used in brake cables, or similar.

You also get your handlebars closer to where you can reach them without extending them backwards and ending up with that clumsy 'handlebars move off to the right when turning left' issue.

That solves a lot of issues for recumbents, space constricted velos, trikes etc.

(The world would be a better place if people didn't 'own' other people's ideas: What are the chances of further idea sharing when that shit happens?)

Thx PaPaSteve.

But do NB that the handlebar site pulley could be replaced with a std circular pulley of the same diameter as the large diameter of the ellipse on the wheel side.
That would give you 1 to 1 steering at all the angles used at speed, while allowing sharper steering angles at walking speeds.

Did you try such an arrangement with the aircraft cables?
 
I've successfully improved or totally solved twitchy steering with a wider handlebar than stock on 100% of the recumbents i've ridden. If reach is a problem, there's also handlebars for that.

The wobbliness at low speed is easily fixed with the addition of an electric motor :)

On the one bike ( Maxarya ) i don't have dead confident steering on, i've created the problem by replacing a stock non-suspension fork with a lot of offset.. for a suspension fork with an average amount of offset.

I think it is possible to build a recumbent with excellent handling properties that doesn't require the user to hack on it it to get it to handle well. The Cannondale Bent is a great example of that. But this category of bike isn't purchased by performance enthusiasts often, so makers don't focus strongly on that.

NB that wider handlebars has the same effect as gearing with std circular gears would neptropix:
More deflection distance (circumference) of the handlebar ends for the same deflection (angle) of the steered wheel...

I NB that the Goosneck on a std bicycle shifts the bars, as a whole, in the direction in which one leans slightly.
Turn that Goosneck around on a std bike and your steering isn't going to feel as good or be as stable.

So there you are on your recumbent turning, say, left:
Now with bars that reach backwards from the goosneck a lot; your arms are reaching off to the right, while you (your body) are leaning left more than usual to compensate, so you need to reach off to the right more while leaning left even more in a kind of negative feedback loop.
That seems clumsy to me..?
(Sadly I've never had the opportunity to ride a recumbent)
 
Hmmm.. really?
You actually don't lean much on a semi recumbent to turn.
 
NB that wider handlebars has the same effect as gearing with std circular gears would neptropix:
More deflection distance (circumference) of the handlebar ends for the same deflection (angle) of the steered wheel...

I NB that the Goosneck on a std bicycle shifts the bars, as a whole, in the direction in which one leans slightly.
Turn that Goosneck around on a std bike and your steering isn't going to feel as good or be as stable.

So there you are on your recumbent turning, say, left:
Now with bars that reach backwards from the goosneck a lot; your arms are reaching off to the right, while you (your body) are leaning left more than usual to compensate, so you need to reach off to the right more while leaning left even more in a kind of negative feedback loop.
That seems clumsy to me..?
(Sadly I've never had the opportunity to ride a recumbent)
Observing first time test rides on different vehicle styles has taught me :

Taking preconceived notions into a test ride often results in disastrous results.

i.e. " If it doesn't feel exactly like my bike it must be wrong ".

How does a person advance their education if they aren't open to exploring other concepts ?
 
Observing first time test rides on different vehicle styles has taught me :

Taking preconceived notions into a test ride often results in disastrous results.

i.e. " If it doesn't feel exactly like my bike it must be wrong ".

How does a person advance their education if they aren't open to exploring other concepts ?

I'm not sure what you mean here?
I'd love to test ride a recumbent but haven't ever seen one here and importing into The New South Africa = buying stuff for some "customs official". All tracking ends the moment stuff lands on the dock here and your only inkling of where your, yours no longer, import might be is some official trying to get you to pay import fees, taxes, levies and fines etc into their personal bank accounts.

Now combine that with the difficulty of finding income as a white person in The New South Africa
and the fact that there is no point in opening a case of theft with the New Police when your tools are stolen..!

ie: I'm stuck with 'The Science' only, an affinity for new ideas, and a desire to help...
The Science hints at untried ideas that may or may not improve things.
So: 'Preconceiving' new, untested ideas vs already experienced solutions as not worth trying: Right back at ya! :)
 
I'm not sure what you mean here?
I'd love to test ride a recumbent but haven't ever seen one here and importing into The New South Africa = buying stuff for some "customs official". All tracking ends the moment stuff lands on the dock here and your only inkling of where your, yours no longer, import might be is some official trying to get you to pay import fees, taxes, levies and fines etc into their personal bank accounts.

Now combine that with the difficulty of finding income as a white person in The New South Africa
and the fact that there is no point in opening a case of theft with the New Police when your tools are stolen..!

ie: I'm stuck with 'The Science' only, an affinity for new ideas, and a desire to help...
The Science hints at untried ideas that may or may not improve things.
So: 'Preconceiving' new, untested ideas vs already experienced solutions as not worth trying: Right back at ya! :)
Science theory is proven with testing.
Sorry your not able to try other concepts or have the tools to build them.

Here's a South African who builds his own.
His avatar is "panpan" on recumbents.com and Bentrideronline

His first paragraph on a build thread :
There are no recumbents for sale in Cape Town (and our exchange rate makes importing a bike out of the question for me ) so I am slowly building a collection of DIY recumbents.
Despite my previously posted reservations regarding SWB I think that I should try to build and ride one before becoming an armchair "expert " at matters I have no experience with.



Hope that helps ;)
 
The first recumbent I saw maybe 45 years back was built by a 13 year old in his older brothers appliance repair shop.
Fashioned after the Ground Hugger being advertised in magazines.
Using wood for the main frame, seat and bolting recycled bicycle frame bits to hold the steering and wheels.
Cable steering instead of a "U" joint.
Kid was quick.

For another low cost wood build look up "Stan Bennett Buck Board Recumbent"
 
@Logic11 :
My original steering design for the first bent I was going to build (and discarded due to being too low to be safe in traffic conditions here) would have been sprockets with a bike chain. I don't recall the ratio, but it would have been non-1:1, to make a tiny input from the bars allow a larger response by the wheel, so I would nto have to move my arms much while riding. I maintained this idea in the next build, though a completely different steering method was used, and the ratio was much reduced:

One of the main takeaways was any ratio change between handlebar and front wheel ruined the feedback to the rider.
It wasn't a good idea.
My CrazyBike2 has a slight non-1:1 ratio which took me some time to get used to, but once I did it worked fine.****

Other people that have tried to ride it vary in their responses. If they are only used to cars, they didn't really have a problem***. If they were used to bikes or motorcycles, they couldn't ride it, or found it scary.

***once they got up to speed, which for that bike is probably around 10mph (been a long time since it was decommisionned, so I don't recall exactly; it's probably in the thread for it somewhere). Like most similar bikes with low-rider-seated-positions it's wobbly until that speed is reached and exceeded, and stable after that point.


****the way I originally constructed the remote steering allowed a serious problem to occur, where the front wheel could be overpushed beyond the point at which the rod could pull it back, and attempting to steer back would then actually flip the front wheel around. :( Fixed that, but it wasnt' even something I'd thought of, as it didn't happen just sitting there, it would only happen when riding....
 
Always good to have feedback from another craftsman.
Seems as though your experiments were about the same as mine.
 
It's as if you don't think all these problems were solved 100 years ago. It's what resulted in regular normal bikes.
 
Innovation, progress, creativity are normal human traits.
Streamlined self propelled vehicles are becoming more common in some countries where the cycling infrastructure is in place.
Some mindsets are 100 years behind the curve.
 
It's as if you don't think all these problems were solved 100 years ago. It's what resulted in regular normal bikes.
If there were only a single solution to every problem, the world would be very limiting and boring.

There would only be one bicycle design, in one size, with one type of tire and tread, and one set of gearing, one type of brakes, one total solution for bicycling. No one would be able to to customize anything for the varied conditions and usages that they have, because no other solutions would exist for them.

The real world doesn't work that way; while there are "optimum" solutions for various problems, which work for most situations, not every one is the same, and so the optimum solution for those situations can be different.


Sometimes a problem isn't just one problem--it's a series set of them, defined by the usage and needs of an individual. The set of solutions to the problem set will then vary depending on the desired results, which may be very different between individuals.

As an example, peek at my Snuggles The Wolf Robotics Project thread--I have a specific set of problems, usage, needs, that apparently don't apply to any other person. So the solutions everyone else has don't apply to my project...but most people don't seem able to grasp that there *are* different sets of problems and requirements, and that a different solution set is needed for each one, and insist that I must use their solutions that only work for their problems, and don't apply to mine.
 
Last edited:
If there were only a single solution to every problem, the world would be very limiting and boring.

There would only be one bicycle design, in one size, with one type of tire and tread, and one set of gearing, one type of brakes, one total solution for bicycling. No one would be able to to customize anything for the varied conditions and usages that they have, because no other solutions would exist for them.

The real world doesn't work that way; while there are "optimum" solutions for various problems, which work for most situations, not every one is the same, and so the optimum solution for those situations can be different.


Sometimes a problem isn't just one problem--it's a series set of them, defined by the usage and needs of an individual. The set of solutions to the problem set will then vary depending on the desired results, which may be very different between individuals.

As an example, peek at my Snuggles The Wolf Robotics Project thread--I have a specific set of problems, usage, needs, that apparently don't apply to any other person. So the solutions everyone else has don't apply to my project...but most people don't seem able to grasp that there *are* different sets of problems and requirements, and that a different solution set is needed for each one, and insist that I must use their solutions that only work for their problems, and don't apply to mine.
You make an argument for DIY, but not for commercial recumbents. Those failed generations ago, and continue to do so.
 
3 recumbent companies have been in business since 1980 ~ 45 years
2 have changed ownership
1 has had one owner

Mass consumption of any product should not be the hallmark of success.
Look at the mess fossil fuels, soft drinks, tobacco has wrought.
 
3 recumbent companies have been in business since 1980 ~ 45 years
2 have changed ownership
1 has had one owner

Mass consumption of any product should not be the hallmark of success.
Look at the mess fossil fuels, soft drinks, tobacco has wrought.
Rans, greenspeed, hase, ice, hp velotechnik?
 
Science theory is proven with testing.
Sorry your not able to try other concepts or have the tools to build them.

Here's a South African who builds his own.
His avatar is "panpan" on recumbents.com and Bentrideronline

His first paragraph on a build thread :
There are no recumbents for sale in Cape Town (and our exchange rate makes importing a bike out of the question for me ) so I am slowly building a collection of DIY recumbents.
Despite my previously posted reservations regarding SWB I think that I should try to build and ride one before becoming an armchair "expert " at matters I have no experience with.



Hope that helps ;)
I'm a 1000km away, but will try to contact/visit when next in CT.
Thx! :)

Also you got me thinking... :)
 
@Logic11 :
My original steering design for the first bent I was going to build (and discarded due to being too low to be safe in traffic conditions here) would have been sprockets with a bike chain. I don't recall the ratio, but it would have been non-1:1, to make a tiny input from the bars allow a larger response by the wheel, so I would nto have to move my arms much while riding. I maintained this idea in the next build, though a completely different steering method was used, and the ratio was much reduced:


My CrazyBike2 has a slight non-1:1 ratio which took me some time to get used to, but once I did it worked fine.****

Other people that have tried to ride it vary in their responses. If they are only used to cars, they didn't really have a problem***. If they were used to bikes or motorcycles, they couldn't ride it, or found it scary.

***once they got up to speed, which for that bike is probably around 10mph (been a long time since it was decommisionned, so I don't recall exactly; it's probably in the thread for it somewhere). Like most similar bikes with low-rider-seated-positions it's wobbly until that speed is reached and exceeded, and stable after that point.


****the way I originally constructed the remote steering allowed a serious problem to occur, where the front wheel could be overpushed beyond the point at which the rod could pull it back, and attempting to steer back would then actually flip the front wheel around. :( Fixed that, but it wasnt' even something I'd thought of, as it didn't happen just sitting there, it would only happen when riding....

So if we were looking at my... simple pic above but with circular sprockets, from bars to front wheel you (lever?) tried:
bike-twitchy-steer-fix-jpg.363367

  • Big to small sproket: Too twitchy?
  • 1 to 1: fine
  • Small to big sprocket is what you currently use: Less twitchy? but issues with low speed high wheel angles. (where lever geometry doesn't help)
This solves all that... and does the 'less twitchy' (or not) you currently use.
IMHO :)

It'll take a lot of playing with different elliptical cable pulleys for at least one side in the above pic, but it'd be interesting!
 
Last edited:
Before discovering the recumbent bike concept I had close to 20 years of fabrication experience.
Creating or wrenching on go-karts, motorcycles, dune buggies, racecars, light aircraft, NASA missile launch technician etc..
The 14 different steering linkage experiments covered many variations, multipliers, dividers, cables types, pulleys, levers and rod linkages.
1 to 1 ratio had best feedback.
For remote steering aircraft cables and pulleys were the smoothest and provided the most steering movement.
Rod linkage has less total movement and when using quality rod ends has a little "stiction".
With "tiller" the hand movement is opposite / different from an upright bike so some riders don't adapt as quickly.
"Twitchy" . . . the common fix is more time on the bike.
A light touch is all that's needed not a tight grip and heavy feel that and upright often promotes.
During my racing years I spent many hours doing circles and figure 8's in small parking lots and outdoor basketball courts.
This helped to gain confidence in the bike's handling qualities and also the level of tire traction capability.
 
3 recumbent companies have been in business since 1980 ~ 45 years
2 have changed ownership
1 has had one owner

That's actually sad, but understandable.

I finished a restoration on an Easy Racers Gold Rush today. It was way too short for me, so I had two of my co-workers test ride it for me. They both succeeded, and they both decided it sucked. I wasn't surprised, even though I had been open to the idea that they'd like it.

"Oh but they'd get used to it in time". Yeah, I've built enough choppers and mutant bikes that I know you can get used to whatever abomination. That doesn't make it a non-abomination.
 
Back
Top