Direct Chain Line

[Stradageek]

I'm a firm believer that rear wheel drive recumbents have one Achilles heel, it's called a chain idler.

My otherwise slow Kettwiesel trike races up hills due, in most part, to having a direct chainline. The more direct chainline on my Strada is, I'm sure, responsible for it's much better acceleration and climbing prowess than my (admittedly much heavier) Speedmachine. My flexible aluminium box section BikeE has a direct chainline and climbs, again, remarkably well. However, both the Bike E and Kett are a bit poor against a headwind because of the non-ideal 'laid back' position.

So has anyone built the 'ideal' machine i.e. laid back seat, feet above the hips and no chain idler?

I've seen it done on a Bacchetta by a rider with short enough legs to slide the seat well forward and remove the idler, unfortunately, my legs are very long. What I would need to do on my Strada is drop the rear axle and fit a smaller rear wheel

Anyway, thoughts and idea's welcome or is this mythical machine out there somewhere and I've just missed it?

[nobrakes]

I think the power loss from idlers is vastly overstated. My M5 with 3 idlers outperformed my Cruzbike V20 with no idlers up a local hill for the same power. There were no situations in which it was slower. I've just spent 2 months with a Metabike Daemon that has 4 idlers and I was putting in some of my best climbing times and a few PRs.

Testing with a power meter never lies. I've stopped worrying about idlers and I'm speed obsessed.

I think stiffness is way more important than the number of idlers.

[[XAP]Bob]

My Rapto used to fairly fly, and that had a ~110 degree wrap around the idlers on both lines.

A poor idler will be bad, but a good one shouldn't leave much behind, else they'd get hot, fast.

[StephenW]

A few thoughts:

1. The fact that idlers don't get hot doesn't prove that they are not wasting energy through friction. I'd expect most of the friction to come from the chain being forced to bend whilst under tension, rather than the idler bearings or friction between the chain and idler. The idler itself has ball bearings which are clean and well-oiled, whereas the parts of the chain rub off each other and are liable to be dirty. There is a good flow of air across the chain, and it is a good conductor of heat, so any heat generated within the chain would be dissipated easily. Nevertheless, I don't think all that much power is lost through friction from power idlers (and even less from return idlers).

2. Based on point 1, the thing that matters in terms of friction is the radius of the idler, not the amount that it deflects the chain. If the chain is in full contact with the idler, the amount that each link has to bend is the same, regardless of how much the chain is being deflected.

3. I reckon the key point about power idlers is that they put bending loads into the frame. If you had a stick bike with no idlers, the chain would be loading it in compression, which is very stiff. If you add a power idler, you put in a bending load, and the frame is much less stiff in this direction. The problem is particularly bad if you are climbing in a small chainring, because that makes the chain tension higher.

4. It might be good to remove the power idler to reduce frame flex, but you have to balance this up against more droop in the long unsupported run of chain. This droop can make pedalling a little spongy, and is also not ideal. Perhaps a clutch derailleur or extra-stiff sprung tensioner might help? The Raptobike has the chain forces aligned with the boom and fork, and also has only a short horizontal run of chain, so pretty ideal in both respects.

5. Droop and flex are both less bad if you use larger chainrings and sprockets to achieve a given gear ratio, because chain tension is lower and chain speed is higher.

I tried riding my metabike for a while without any idlers. The chain rubbed on the seat bracket in certain gears, but if you avoided those gears then it was OK. I couldn't say if it was faster, but I really liked how smooth and silent it was! The chain did flap about a bit over the bumps. If you used a hub gear you could avoid the chain rubbing on the seat bracket, but then you would also introduce a new source of losses and possible sponginess. I wrote a thread on here about it ages ago.

[[XAP]Bob]

Even on a turbo the idlers never got warm, which is a good indicator that they don't waste large amounts of heat (even a relatively low power draw will increase the temperature appreciably.

The chain does still get airflow of course, and small bends aren't great, but the smallest bends are in the dedangler on most systems - I know it's not under high tension, but it's still easily the wiggliest bit.
In any reasonably high gear the rear sprocket will also be smaller than a decent idler indeed some idlers *are* sprockets, just with chain guides each side as well.



wrt 4&5 - I had the rapto equipped with a schlumpf HSD, so an effective 75 tooth chainring - minimising chain tension.

[Stradageek]

A few thoughts:

I tried riding my metabike for a while without any idlers. The chain rubbed on the seat bracket in certain gears, but if you avoided those gears then it was OK. I couldn't say if it was faster, but I really liked how smooth and silent it was! The chain did flap about a bit over the bumps.

Thanks Stephen, for your thoughts and in particular the experiment above.

It does seem likely that what I feel I'm losing and what I'm actually losing are probably very different. I'm tempted to repeat your experiment on my Strada as some gears will run without the idler but I guess that some clever power meter tests would be the only way to be sure.

Cheers

Stradageek

[StephenW]

I do feel that the psychological advantage of a smooth and silent drivetrain is not to be sneezed at! It can really help me to pedal more enthusiastically if everything is running smoothly, whereas I really dislike noise coming from the bike.

On some of the longer, lower bikes, without an idler the chain would be sawing through the rider's bottom! But on shorter, higher bikes it seems to just rub on parts of the bike.

I have an idea of how RWD recumbents in general could be better, which would also make it easier to run without a power idler. What I think is needed is a much wider rear hub, with a much longer freehub body. This would allow many sprockets to be fitted (17 perhaps?), without resorting to making them really thin. The rear dropouts would then be offset to the right. This would give a dishless rear wheel, which is nice and strong. Because there are lots of sprockets, you can have a single chainring and still have a really wide range of gears, without big gaps between them. This single chainring could be quite large, which would keep the chain tension down, reducing problems with flex in the frame while climbing.

(This setup would be no use on an upright bike, because the chainline would be terrible, and the rider's right heel might catch on the chainstay. On a RWD recumbent neither of these is a problem).

The relevance of this to having no power idler, is that it would keep the chain further to the right, meaning that it is less likely to rub on the frame, seat bracket etc. Having a single front chainring makes it easier to protect the rider from the chain, without resorting to things that rub on the chain (i.e. chain tubes). (Perhaps also easier to cover the chainring for safety reasons - don't want to impale anyone with a chainring in a crash).

With a single front chaining, you could also have a chain-keeper (maybe consisting of two rollers which the chain passes between), which moves up and down according to what gear the bike is in. I'm imagining that the chain will just avoid touching the rollers whilst it is under tension, but when the rider stops pedalling or goes over a big bump then the chain will touch the rollers.

[[XAP]Bob]

ICE already do an offset rear triangle to get a dishes wheel (unless you use a hub gear in which case the dish is actually inverted).

The alternative, to retain a generic drive train, is to have a two stage drive - with an intermediate drive that is the "front" dedangler.
There are a few ways you could drive that intermediate drive, one relatively obvious one is the newer shaft drives - don't need to change gear, so would be simpler to assemble - but might need the shaft to be simply telescopic to allow for boom adjustment.
Belt drives also have potential in there.

Then you have a fairly normal rear setup, and a front shaft/belt/chain which should be very easy to protect.

[UpWrong]

I've had two bents with a direct chain line: HPV Spirit and Rans Stratus XP. I think it does make a difference when climbing. As you say though, the problems with high racers and even SWBs is the the seat gets in the way of the chain line, and the front wheel gets in the way of the feet. It would be interesting if someone could build and market a bent based on an eliptical drive: https://vimeo.com/32447643

[[XAP]Bob]

You can add a k drive, just need to build it - doesn’t affect chain line (does affect heel strike)

[StephenW]

I believe the Challenge Mistral doesn't have an idler, at least in some versions. I have never ridden one.

[Brucey]

FWIW if you don't have something, you have a long chain run instead, which will tend always to sag and flap about with variations in tension. This may use as much power as an idler.

The 'horrible thing' I built has an IGH mid-drive. Possibly it isn't as efficient as an idler, even when the IGH is in direct drive, but for some reason it is nigh-on silent in operation, so it has stayed. FWIW I built the thing in such a way as I could use a long chain run instead, or an idler system. I have never got around to comparing these arrangements on the same machine, which was the whole idea of the design....

cheers