It’s been the best part of a fortnight since I last posted and there’s a reason for that – it’s got to the time of year when I suddenly realise there won’t be too many more days when the weather is good enough to do all the outdoor tasks that need doing before the winter. I’ve been busy painting woodwork at the front of the house, including the porch doors, door frame and an excessive number of decorative spindles. It’s about 2/3 completed so far, but when that’s done there are a couple of roof-level jobs too, including fitting a “birdcage” vent terminal that I’ve 3D printed onto our soil pipe.
I couldn’t find a white one locally (we replaced the grey PVC soil pipe and other external waste pipework from the bathroom with white not long ago, to match the colour of the render), so I thought I’d try and make one. It’s the longest 3D print I’ve ever done at just over 7 hours, and the shape was just on the edge of what is feasible without support, because I wanted to use PETG filament for its waterproof properties. Unfortunately, supports don’t break off cleanly from PETG.
I had to cheat a little by printing the cap at the top separately and gluing it onto the birdcage part. The print isn’t perfect – the ribs are a little uneven towards the top and would definitely have been improved by supports – but it should suffice. All it has to do is keep leaves out in the autumn and prevent birds from nesting on it in the spring.
When I haven’t been outside painting I’ve found time to start a couple of textiles projects. First, I’ve cast on a lacy scarf to use up the fine linen yarn remaining from my recent shawl.
It’s called Branching Out and is available for free on Knitty.com. I’m adding beads again, because well, I have beads left over from the shawl too and why not?
The second new project is a blouse I’ve cut out from a long cotton skirt I haven’t worn in years. The pattern I used is even older, it’s a classic shirt style that I first made in 1979. More about that next time, and also another patterned plywood project that I’m planning.
OK, where I left things before was that I’d abandoned the idea of bending copper pipe into a ring by running a profiled roller around a former and reverted to Plan A, a 3-roll pipe bender. And I’d decided to try using a vice and Dexion to make my new bender.
Removing the replaceable vice jaws revealed little ledges that they rest on. The Dexion is wider than the vice jaws are tall, which meant I needed to cut pieces of plywood to sit on the ledges behind the Dexion.
That done, I could screw the Dexion onto the vice, through the plywood, using the same machine screws as held the removable jaws. And when I tried bolting my one existing roller onto the horizontal flange of the Dexion, I was pleased to find it was high enough above the vice for there to be enough clearance for the bolt head. This is a top-down view.
I’ve 3D-printed a second roller of the same size to attach next to the first one on the Dexion strip that’s screwed into the vice’s moving jaw, and also a larger roller to attach to another Dexion strip on the fixed jaw.
The larger roller is the driven one. I bent a strip of aluminium to form a crank and screwed a handle on the end of it. The whole caboodle is bolted together with a bunch of M8 bolts, washers and nuts.
But disaster has struck again. Before I’d achieved much of a bend at all on the offcut of copper pipe I had left after the previous attempt, it proved too hard to turn the handle to wind the pipe back and forth between the rollers.
When I looked at them from one end I realised why – the Dexion was flexing, allowing the rollers to tilt out of the horizontal plane.
Clearly the force I’m applying by winding the vice jaws ever closer is enough to take the 2mm thick steel angle strips into the elastic deformation zone.
So it’s back to the drawing board again. I’ll have to find a way of stiffening the thing up.
My attempt at bending a 15mm diameter copper pipe into a ring with a 190mm inside diameter hasn’t gone entirely to plan. It proved perfectly feasible to rotate the roller on its arm around the edge of the circular former I made, forcing the pipe into the groove around the edge as it went. I was afraid that it would take more force than I could provide, especially given that I had filled the pipe with sand to help to support its walls, but I’d made a long lever arm for the rotation handle which made it relatively easy.
So what went wrong? Basically, the problems only started when I reached the straight part of the former disc’s edge towards the end of the ring. My former has this flat part for two reasons: firstly the MDF sheet I used wasn’t wide enough for a complete ring, and secondly, I thought it would be necessary to have a straight section in order to be able to get the completed ring out of the groove and off the former. But in practice what happened was I needed to release the starting end of the pipe from its clamp once I’d gone round all the curved part of the edge, so I could slide the whole pipe around to get a curved section of former in front of the roller again. It proved impossible to get the bent pipe back into position in the groove, because of springback, and as a result I got a kink in the pipe.
I found lots of YouTube videos showing versions that people had made, mostly starting from scratch with steel strip, angle and bar, but some shortened the process by using a vice to provide the slider mechanism that moves the driven roller towards the other two. But every video I saw required a lot of cutting, bending and drilling steel components – and I don’t really have the equipment for that, nor a local source of steel stock. The real killer was a requirement for welding equipment, which is something else I don’t have.
Then it occurred to me that I might be able to eliminate the need for steel stock and all the cutting, drilling and welding by using bolted-together Dexion strips with a metalworking vice.
I love Dexion. If you haven’t come across it, it’s basically Meccano for grown-ups. The Dexion company makes a number of products, but the one it’s most famous for is “slotted angle”, steel angle strips with numerous bolt holes. You can make all sorts of temporary or semi-permanent structures from it. It’s strong, stiff and reusable. I already used two of the 3 strips we possess to make the arm that carries the roller around the edge of the circular former. I will dismantle that apparatus and see if the Dexion can be bolted onto a vice. If so, then I’ll need to 3D print another two profiled rollers, one the same size as the existing one and the other a larger one that I can rotate with a handle.
A week after electrifying my old hybrid bike, how is the e-bike experience? I love it. Now I can climb hills at anything up to the UK pedal-assisted speed limit of 15.5mph, which is a darn sight faster than my usual uphill speed. And I no longer need to get off and push up the steeper hills, or the less steep ones when I’m carrying a few extra kilos of shopping or whatever. It means I can choose safer cycling routes on quiet backroads or off-road tracks, which tend to be hillier than the main roads that follow the rivers in this part of the world. Around town I feel safer too, because I can accelerate from junctions, traffic lights and zebra crossings as fast as the traffic, and keep up with it for more of the time, or at least travel at a speed closer to that of cars and lorries so those approaching from behind have longer to see me before they overtake.
I haven’t been very far yet as I’m still getting familiar with the controls. I don’t want to do anything too adventurous until I’ve developed the muscle memory to be sure I’ll cope with any emergency situation quickly and effectively, without having to pause and think what’s best to do.
I wasn’t sure I’d use the handlebar-mounted throttle, but I fitted it anyway. It powers the bike even when the pedals aren’t turning, but only up to 4mph. That means it can be used for “walk assist”, ie for pushing the bike up a steep hill when riding isn’t possible for some reason, but also for giving a little extra oomph when setting off from a standing start. I’m finding I use it quite a bit for the latter, as the pedal assist mode only works when the pedals have started turning (obviously). The extra weight of motor and battery on the bike, plus the fact that there’s now only a single 46 tooth chainring instead of the previous 28T/38T/48T triple set, means it can be hard to get going on a hill without a quick flick of the throttle. There are times when it would feel quite unsafe turning right across the traffic from a T-junction, or even moving off from traffic lights, without it.
I wish I’d done this conversion before our recent Dutch holiday. We could have covered greater distances and/or avoided me arriving at our accommodation every evening in a state of semi exhaustion. Talking of foreign cycling tours, the main drawback I can see from going electric is that it will no longer be possible to take this bike on a plane. But it’s been getting harder and harder, and more expensive, to fly with a bike in recent years, which is why we hired when we toured in Latvia in 2019 and in Denmark in 2015. Fortunately, there’s no problem taking e-bikes on ferries, although P&O did charge us for our bikes on the Hull-Rotterdam route this time whereas they used to travel for free.
The only thing that worries me a little is how I’d deal with a puncture away from home, or any other issue that necessitates turning the bike upside down. The throttle, control switch and display are all mounted on the top of the handlebars and would be easily damaged. The display in particular is vulnerable and too large to be able to rotate out of the way, and in any case, the wiring isn’t loose enough to permit it. (And if it were any looser, it would just get in the way while the display is in its normal position.) But punctures are a rarity these days, with Kevlar-reinforced tyres – I literally can’t remember when I last had anything more serious than a slow leak that could still be ridden on with occasional top-ups of air. I guess I’ll just have to push the bike to the nearest railway station if it ever becomes unrideable while I’m out and about.
More knitted lace
I have a small amount of the fine linen yarn left after making my Frosty Apples shawl, as well as plenty of beads. I’m hoping there’ll be enough yarn for a small scarf, using the Branching Out pattern from Knitty. In an effort to stretch it out I cast on with 6mm needles, but I’ve decided it looks too open. I’m going to have to rip it out and start again on smaller needles.
I am now the proud possessor of an electrically-assisted bicycle. There was nothing technically difficult about fitting the Bafang conversion kit I bought, but I did run into a few problems. It’s a mid-drive kit, with a motor drive that fits through the bike’s bottom bracket shell, the part of the frame between the pedal cranks. To get it in, you need to clear everything out of the shell by removing the cranks, the spindle between them, the two bearings that allow the spindle to rotate freely, and the cups that hold the bearings in place on each side of the frame.
I knew from the homework I’d done before deciding on this kit that removing the bottom bracket bearings and cups can be a struggle, especially if (like my mid 90s bike) they’ve been in there for years. But I thought I would be OK because I’d already removed the right hand side bearing – with the aid of husband-provided brute force – to check the internal diameter of the bottom bracket was the standard 33.3mm diameter, as it needs to be for this particular mid-drive kit. Surely the left hand bearing wouldn’t be any harder to get out? Well, it was. And the fixed cup on that side was even more of a brute to get off – I guess it’s called fixed for a reason.
If you are considering doing a similar e-bike conversion, then unless you do a lot of maintenance on your bike and as a result are fully familiar with all its quirks and have the requisite tools, you’ll find that you spend a lot of time trying to figure out exactly what type of component you’re dealing with, and hence how to get it off and/or replace and re-adjust it. This applies particularly to bottom brackets, some of which have threaded shells like mine (ie the bearing on each side screws into the frame), while others are cup and cone type, press fit, or even carry the bearings on the cranks instead of within the bottom bracket. And those are just the more common variations. Plus there are sub-variations of each type, all potentially requiring different tools. When you’re removing the left hand side crank, in most cases it will be screwed on with a left hand thread so that it doesn’t come undone while pedalling, but you can’t even rely on that on all bikes.
I spent a fair bit of time consulting my trusty bike repair manual (I swear by Richard Ballantine’s books) and watching YouTube videos, before ringing round cycling friends in an effort to borrow the right tools. Many of the difficulties I encountered were as a result of not having the ideal tool, but I had no wish to spend time and money buying specialist items – like a fixed cup spanner – that I’ll never need again now the bottom bracket contains a motor shaft. In hindsight, it would have been quicker and less frustrating to get a bike shop to remove the cranks and everything in the bottom bracket shell, and I could still have wheeled the bike home afterwards. As it was, I did need to make the odd trip (on foot or by car) during the process to borrow tools and buy such things as cable ties, duct tape and heat-shrink tubing to tidy up the cabling.
Actually, the cabling part took more time than expected, much of it in just trying to work out the best way to route cables and how best to bundle up the excess so they could be tucked out of the way and made to look as neat as possible. Something else that was unexpected was having to search around online for the right instruction manual for the display I’d chosen, in order to be able to access the set-up instructions and put in the right wheel size and so on. All in all, and allowing for weather-related delays and other demands on my time, the whole installation process took about a week. Now I know what’s involved I reckon I could complete another e-bike conversion on a similar bike within an afternoon, even without all the ideal tools, as long as I didn’t run into any major bottom bracket issues.
Having done everything, I cautiously pushed the bike along the road a few hundred metres to an off-road path for a test ride. I came back 20 minutes later, on the road, with a big grin on my face. I found the controls and the changes I needed to make to my riding style fairly intuitive, and the extra power available on hills made the experience a lot of fun. I’m going to enjoy this!
First, I used a 90° router bit to cut a V in the edge of the MDF former.
Then I designed and 3D-printed a roller with a profile to match the copper pipe, and attached it to a pivot bolt in the centre of the former using Dexion-type angled steel strips. The roller can rotate around this pivot, leaving space for the pipe running around the edge of the former in the V groove.
I need something to hold the end of the pipe in a fixed position against the former while the roller is rotated to press the pipe into the V. I’ve used a wooden block with a hole drilled down the middle and then cut into two pieces. The larger piece, with a half-pipe void running down it, is screwed onto the edge of the former with triangular plates I found in the garage – I think we used them to make a long-dismantled futon base in the 80s. I then made the smaller piece of wood wedge-shaped, so it can be pushed into place as far as is necessary to hold the end of the pipe tangentially against the edge of the former.
The final step has been whittling a couple of wine corks down to fit the ends of the copper pipe, and drying out a few litres of builders’ sand in the sun before sieving it to get the lumps out. I’ll fill a generous length of pipe with sand and cork the ends to stop it escaping. This may not be necessary – when a pipe is bent into a fairly large radius with its walls well supported externally, it shouldn’t buckle, but I don’t want to take any chances. Supporting the walls internally with sand seems like an easy, cheap precaution.
I have no idea whether this will work, but it’ll be fun finding out.