2nd December 2015 Shimergo double pully, cable pull ratio converter.

A double pulley style, pull-per-click, ratio changer could be mounted to pivot around the smoothly cylindrical, stainless steel, Trykit bottom bracket. Such a device uses the JTek Shiftmate principle but without the weakness of putting too sharp a bend on the cable. The pulleys could easily be turned as one unit in the lathe from Tufnol. This material is used as a bearing material in some applications so may well turn smoothly on the bottom bracket. It just needs a suitable lubricant and a retaining ring. I can but hope that friction is not high enough to inhibit indexing.

Perhaps I should look at pulleys with smaller radii which would allow a single length of cable to be wrapped over both via an internal guide? Thus avoiding the need for cable clamps.

Originally I had planned to run each cable half way around the pulleys before making the cable turn inwards to meet a normal brake pinch bolt sitting in a cross drilled hole. The friction applied by the cable to the pulley rim would reduce stress on the cable at the pinch bolt. Where a sharper inward bend would be required just to reach the pinch bolt.

The image shows how the unbroken cable could be made to wrap around both pulleys much as it does in the JTek Shiftmate but without the inevitable local stressing. The bare, downtube cable inner joins the smaller pulley and the run to the rear changer continues on from the larger pulley. 

With a split cable system the independent halves of the rear changer cable would each be anchored with a pinch screw after taking a half turn around their own pulley. Since the pulleys are fixed together, or turned from the same stock, each circumference rotates simultaneously at a fixed ratio of 1:1.36. This converts the Ergo lever's meager 2.6mm pull-per-click to the required 3.6mm to suit the Shimano MTB rear changer and the 39mm wide 11 speed Sram MTB cassette. The XT IRDM8000 rear changer is really designed for 40 and 42 teeth largest sprockets. I don't see why it shouldn't easily cope with a 36tooth sprocket with suitable adjustment of the B-screw. The 36 tooth cassette profile will be very similar to the larger ones. The MTB changer has quite a soft pull over its 1.1mm lateral movement per click so should place no extra loading on the Ergo lever mechanism.

I shall just have to wait for the cassette to arrive from the UK before I can have a play with my new toys. Meanwhile I can turn the double pulley to judge friction levels while riding on the BB. Absolute excentricity  is unimportant so a tight fit is not necessary. I have some thin PTFE [Teflon] sheet which could be wrapped as a low friction bearing material if the Tufnol alone should prove too sticky in use. The BB area is very prone to the gritty spray off the roads from the front tyre.  The cable[s] will be pulling at a considerable radius compared with the BB shell so there should be no real problems. The mechanical advantage is quite considerable at that radius but is gradually reduced with smaller pulley diameters.

My pull-per-click conversion idea might have appeal to other cyclists hoping for an MTB/Brifter system but I have relied on the smooth BB of the Trykit as both a pulley bearing and support. The pulley mechanism would have to be capable of being fitted to a universal downtube clamp to make any sense on most bikes. Which rather defeats the huge advantage of an almost unchanged cable runs with my BB-concentric pulley arrangement. Few cyclist would be willing to fit such a pulley system onto their turning BB axle, even if it were possible, because of the increased friction and wear. That would be the only other concentric pulley option I can think of.

So it looks as if my double pulley idea is unique to an unpainted SS Trykit trike bottom bracket shell. I will now look more carefully at suitable pulley sizes and thicknesses to suit the diameter of my ~38mm BB shell. Fortunately it will be hidden behind the chain wheels so will be almost invisible. A smaller diameter pulley system has considerable advantages as to cosmetic invisibility, more closely matching original cable runs, allowing a normal, full length, single rear cable to be used and of course, reduced weight.

This image shows a mock-up with a 68mm large pulley and 50mm smaller pulley inboard fixed onto a 38mm bottom bracket. Not too cosmetically shocking. Clear plastic would become invisible but require some 6mm polycarbonate and a lot of polishing to remove the lathe tool marks.

Let's assume a nice round figure of 50mm for the smaller pulley diameter. This equates to the larger pulley having a diameter of 68mm to achieve the 1:1.36 change in cable pull-per-click. The maximum allowable pulley thickness is about 4mm at the hub [BB shell] though the pulley rims could/should be made wider if necessary. I shall use 4mm Tufnol sheet for my first trial to confirm the basic idea will work in practice. Fortunately I am using a sealed BB unit which has a slightly larger flange at the bearing housing than the BB shell. So I have a natural shoulder against which the double pulley will be safely retained. There is very little sideways pressure and the rotation is intermittent and only moves over a small angle during actual gear changes. The cable only moves over a 26mm or 1" arc at the rim of the smaller pulley. Which amounts to 1/3 of a turn.

Pi.D/arc. 3.142 x 50mm /26mm = 157/26 = 6.
360/6 = 60 degrees of movement.

The images show a first mock up, turned from a scrap piece of 3.5mm clear poly-carbonate. I wanted to get some idea of clearances before forging ahead with the real thing. It proved that the rim wants to be about 6mm thick to safely accept the cable grooves. With a general cup shape incorporating a couple of mm offset outwards from a 4mm center thickness. Apologies for the awful flash images and the filth on the trike. It hasn't really stopped raining all morning so I was taking images one-handed in the dark. The trike was cleaned prior to the last ride. Which shows just how much grit is thrown up and sticks in only a few short miles.

The smaller pulley must be fitted on the inside to avoid contact with the nearby chain stay.  The larger pulley then has slightly more room to clear the splayed chainstay. I may need an o-ring to space the pulleys away from the brazing. A very small parting tool can put the cable profiles on the pulley rims. A shaped hacksaw blade in a tool holder would allow the cutting edge to reach the inner pulley rim. Cable "derailment" is highly unlikely with the usual cable tension. Though I had better sink the cable below the edges of the rim by at least half its own diameter for security.

Fortunately the closeness of the two pulley grooves helps to avoid any twisting effect due to cable tension. I can modify the bends at the drilled inter-pulley, guide hole with needle files and burrs to reduce local cable stressing. I just hope the inner gear cable is long enough to accept the considerable extra pulley wrap. They usually are, but the last one I fitted was only just long enough with a standard cable run! Further reducing the size of the pulleys is possible if the chain stay obstructs them at the present size. 

I made the ratio adapter out of [very stinky] 6mm Tufnol. There was so little room on the BB shell that I had to use the fixed BB bearing housing as a bearing surface. I managed to drill the tangential hole for the cable to rise from the small to the large pulley. It needs more work to smooth the entry and exit holes to allow the cable to lie flat in the pulley grooves. Which I made deep enough to completely hide the cable. Though it doesn't really look like it in the posed image where only sections are hidden from view. It really needs a stronger and thicker material than the Tufnol I used. Probably requiring that it overhangs the outboard BB bearing. The BB shell proved to be oval which made life rather difficult when trying to obtain a close, but not loose, fit. The ovality was probably the result of the silver brazing causing distortion. Had I used aluminium alloy I could have made both pulleys smaller though still in the correct size ratio of course. The main problem is feeding the cable through the inter-pulley guide hole against considerable friction. Perhaps the whole idea is crazy and I should have stayed with the L-shaped bellcrank system of the earlier post.

Click on any image for an enlargement

*