Step 1: Planning
Step 2: Hunting
Step 3: Tinkering
Step 4: Buying
Step 5: Assembling
Build Day 1
Build Day 2
Build Day 3
Step 6: Perfecting (you are here)
Preparation is Everything?
In Planning, I opened by saying Preparation is Everything. With that said…
"Everyone has a plan until they get punched in the mouth." -Iron Mike Tyson
Yeah. Lets talk about the reality check that is coming, once you actually ride the bike you just built. You learn whether what you thought would work actually does. More than likely, something will not work the way you’d like it to.
It won’t be a catastrophic problem, but this is a custom bike and you should expect a do-over or two to make it exactly what you want. This is how I wound up with the materiel and experience to write Musical Chainrings.
On that subject (bicycle gearing), over time that inevitable uncertainty has worked out in my favor. I know I am going to need some time in the saddle to figure out exactly how I want to gear any bike. I may also be surprised when I get a look at actual versus expected chain alignment. Thanks to Tyson’s Law, I have plenty of stock on hand to play around with and get it right.
You have to plan for and budget for this final step. Not necessarily for chainrings. There are a variety of typical culprits.
There is a pretty common hit parade of things most likely to need a tweak. They all have something to do with the human/bike interface: How comfortable the bike is to you when you ride it.
Is their width comfortable? What about the angle? Your wrists feel OK after awhile? Need a rise on the bars? You’ll only be certain you got it right after riding the bike.
For the Apostate I put on a 760mm titanium flat bar. I have tried to use this very snazzy handlebar on a half-dozen bikes over the last few years, and was never happy with it, so it went back to the parts pile. Having had the Apostate on the road for a few months now, it looks like I finally found a permanent home for it.
This is surprisingly important for rider comfort, and is perhaps the part I most often change after a new build hits the road. A longer or shorter stem can make a world of difference in comfort depending on what reach to the bars best fits you and your riding position (seat height relative to bar height also plays a role, so once again you need to be on the actual bike to understand what works best). A stem at a different angle can raise or lower the bars for a different improvement than changing the reach with stem length.
SIDEBAR: No matter what... do not use an adjustable riser stem. The kind that has a hinge you can supposedly bolt down so its safe. I know of two separate instances where they broke loose (thankfully I was not the rider). Both under heavy braking. Want to keep your teeth? Use a fixed stem with a set angle to raise handlebar height.
For the Apostate I tried an 80mm stem with a 6 degree rise. Based on measurements from other bikes, I knew this was likely to work. But once again… you never know until you ride it.
Once I did, my posture naturally gravitated to holding the bars with my thumb and forefinger; not naturally planting my upper body weight on my entire palm. I needed a big change, and so I grabbed the biggest change I had: a much longer 120mm stem with a 45-degree rise. This raised the bars as much as was reasonably possible (about an inch and a half) while not really moving the bars forward much (which would increase my reach and make the problem worse).
After riding it for a week, it felt better, but I still had to think to put my hands down flat on the bar. I hadn’t gotten it quite right yet. I needed to reduce reach a bit while not affecting handlebar height.
Since I was pretty much at my best result on the stem length and handlebar height, my next step in fitment moved from the handlebars to the seatpost, where I knew I had a little room to maneuver, so to speak.
Worth noting: I could have stayed at the handlebars and changed the bar to one with a pullback of some kind. But I wanted to keep the bar flat and straight on this bike so…
If you are having reach or posture issues, one of the tools at your disposal is to change your seatpost. Some have a setback, where the saddle is mounted aft of the seatpost tube itself. Others have no setback and the rail clamps are directly over the tube. The difference moves your body forward or backward depending on what type you use.
I try to solve fitment issues with handlebars and stems. Changing seatpost setback is usually a last resort (and if you have a suspension seatpost, changing that expensive part is usually off the table as an option).
For the Apostate, a vintage 350mm Kalloy Uno came with the frame. This venerable post has been on the market for decades. It is a no frills, sturdy option. It turns out a 350mm post, with a bottom set near to matching the frame’s bottom edge (still well within its safety limits) was perfect for my pedal stroke. Winner winner chicken dinner.
Or not. As noted above, after riding it for a week I felt I still needed a small change, and it seemed like it would have to be a seatpost change.
The alternatives left were scooting the seat forward in the seatpost clamp (minding the limits scribed on the saddle), and changing the post to one with no setback. Since I was already at the forward limit of the saddle, that meant a different post with no setback. I did that in an over-the-top way, which moved this modification out of the ordinary and into the Afterword section below. We’ll discuss details there.
My original Kalloy seatpost had a ‘setback’ that moves the seat’s mount back behind the post’s center axis. The post I replaced it with has no setback.
You won’t know if it works until you sit on it and ride for awhile. But, you don’t have to start from scratch, either. What you like on another bike is liable to work again. I know that for bikes I pedal hard, I like narrower saddles. I knew I liked the WTB Volt (taken off of my Surly Big Fat Dummy) on my GG Smash enduro bike.
So I put on another Volt (I scored the much nicer Chromoly version on a clearance sale) and its fine. No changes necessary. You may not be so lucky as saddles are notorious for not being quite right without some trial and error.
Again… this is about comfort. But budget is a factor as well. I tried going with a more or less period-correct option via some old cage pedals with mtb clips and straps. I had them on a shelf collecting dust, and thought this was a great place to put them back into use.
Wrong answer. Some things are better left to the past. Toe clips are one of them. I only had to fumble getting back into them once (I’m not cleating in here) to remember how annoying that was. Fortunately for my budget I also had a pair of perfectly good, cheap flat pedals on the shelf, which I put on.
And I still wasn’t happy. Again thankfully for my budget, my Smash is stored with its pedals off, and those pedals are Pedaling Innovations Catalysts, which are sort of monsters, but I have several sets. I really like the ability to support my arch, in a mid-foot position that benefits from a stomping pedal stroke.
So on they went and … perfect. I’ll use the cheapie flat pedals on the Smash. For now.
Finally, I built a bike I did not need to play musical chainrings with to get it geared right. Some of that was luck, some of it experience. The 40T Lekkie I used – which requires a special motor cover to be substituted on to fit – was a big ticket item, but its the smallest chainring available that would give me the offset I needed to get excellent chainline on this build.
That chainline was figured out in the Tinkering phase, when I had only the frame, the motor, a wheel and some of my spare chainrings to play with. Chainline is dead straight back to the middle of the cluster, and the gears I am comfortable riding in on this bike are the middle ones as well. One and done. For once.
So… maybe Planning Really is Everything.
The Apostate pictured in my Day 1 ride didn’t stay the same. Most of the changes are documented above. But things don’t always fit into neat little categories. What unique bits did I end up changing or prettying-up?
The battery solution on this bike came out great. The frame fits a certain type of ‘in-triangle’ battery pack, and of those packs, the Wolf Pack from Luna Cycle fits as if the frame was made for it.
However, clearances are tight. Particularly on top where it really matters. It was clear even during test fittings I wanted to keep this battery permanently on the frame and remove it as infrequently as possible. Ideally: Never remove it.
Not just because there isn’t much room to work with in terms of getting the thing off of its (super strong) magnetic mount. That strong magnet, versus the rivnut bottle bosses on this vintage frame… worry me. You have to apply so much force to remove the pack (or move it in any way forwards or backwards), I’m concerned something is going to bend (the mount) or break (one of the bosses tearing loose from the frame). There’s likely no coming back from a failure like that on an aluminum frame 23 years old and counting.
SIDEBAR: Why use the cinch straps if the magnet is so strong? The straps provide additional stability and support. I want to do everything I can to take as much stress off of those two little rivnutted M5 bosses in the frame, which otherwise are holding the entire 9-lb battery on their own through all manner of road and trail shocks.
Initially, I used three velcro cinch straps to nearly cover the pack, and also stabilize the magnetic mount as much as possible. Later on, I decided to take advantage of two of the three slots on the battery side’s mounting tabs. These exist so hose clamps can literally clamp the battery permanently to the frame.
The clamps further reduce the reliance on the bottle bosses to do all the work of holding onto the pack. I had already padded the underside of the mount with a thin pillow of red silicone tape. The hose clamp makes no contact with the actual frame thanks to the mount width on one side, and the wire tunnel for the shift sensor and main motor harness cables on the other.
Those clamps also help reduce the potential of battery theft. Sure, nothing is going to stop a determined thief, but the hose clamps – and I made a point of not hiding them for this reason – make it clear to anyone looking that a few minutes (or an angle grinder) will be needed to get that pack off the bike. There isn’t going to be a grab-and-go theft. That fits in with the very limited likelihood of leaving this bike outside at a shop, locked but unattended.
If someone tries to steal it anyway, once a thief shears off or unscrews the hose clamps, they’ll be confronted with that magnet. I bet it will take some time to realize whats holding the battery so tightly. And once that realization dawns, they will have to figure out how to get it moved just right to angle it out of the frame.
Thats time I can spend setting bear traps, digging pits and buying a baseball bat.
I also used velcro to ‘face’ the cinch straps. This holds them together – really only for cosmetic purposes. The straps don’t move once tightened down. The facing (on the sides and the top) just makes the velcro present a little better; keeping the graphics on the battery from bleeding thru in the gaps between the straps.
The Condor Deployment Bag is something I use on almost all of my bike builds. Its easy to adapt into a secure handlebar mount, its small but still the perfect size for a tool bag that can also hold a wallet, phone and keys. These bags are my go-to for hiding wires – and especially controllers – on my 2wd bikes.
The original brown bag was replaced by a black one I also owned – when I finally found it.
Having several of these on hand, I simply switched from a brown one to a black one. The reason is straightforward: black wires blend in better when they are running along a black bag. Note that in some of the photos you may see a lot of wire stuffed behind that bag. I didn’t cut down the brake hoses to size until the very end of the build and test ride process.
This was a big change, but not for an overtly obvious reason.
The vintage Kalloy Uno seatpost that came with the frame worked great. Except as noted above I had reach issues. I had already moved the seat forward, and I did not want to shorten the stem as that would create other issues. So that limited next steps in terms of fitment.
I didn’t need much reach reduction, so I decided to do a seatpost with no setback. My first thought was a Thomson Elite. Which is a great product but not a trivial purchase at about $115. Since I was in that league in terms of cost, I decided to try a dropper post. They all have no setback. A dropper would be handy for all the reasons droppers are handy.
Also, the frame introduces constraints. The post can’t be super long. 350mm is the right length for a seatpost when fit on the frame to my anatomy; any longer means it protrudes down towards the shock, where the potential for contact is worrisome. Droppers tend to be in the neighborhood of 450mm long, so I wanted to find one with minimal drop. Those posts tend to be closer to 400mm. Also I didn’t want to blow the already blown budget, and a really good dropper costs big money.
I found an interesting option that would be an experiment of sorts, and decided to try it: I bought a PNW Components Coast dropper post, with external cable routing. I could have done internal cabling but a cable coming out the bottom hole in the seat tube could once again be a contact risk with the shock.
Why is the Coast an experiment? Because it is a – unique on the market – suspension post as well as a dropper. Advertised motion is 40mm (it can be more) and its a weird choice because this bike has full suspension already. My reasoning behind doing this – and my results – are enough for a full blog post all by themselves so I’ll just say I did it and it worked well.
As a dropper. Jury is still out on whether it is also an effective suspension feature, but it does seem to work for me in an unusual sort of way. Stay tuned for a separate post on this oddball idea and result.
After all was said and done, I did find a way to test whether there was risk of the seatpost hitting the shock: I removed all but about 20 psi of pressure, which let me easily compress the frame by hand, and observe the result. It turns out, for my frame, there is no risk of contact. Perform this check with yours to learn your result.
This one was pretty straightforward, but boy was it frustrating. I have had occasion to lock the bike up outside a store. The Salsa quick release seatpost clamp that came with the frame carried the usual risk: It makes it easy to steal the saddle and post. Since I am using a US$170 dropper and a US$95 saddle. thats worth taking steps to protect.
With a dropper, there is no longer any need for a QR clamp. So time for a fixed collar. I chose a Bike Yoke Squeezy in 35.6mm size, which turns out to be the wrong size thanks to a mistake on my part. Hint: Take the seatpost clamp off and measure under it. Not below it. This frame has two external seatpost diameters, which is invisible if you leave the clamp in place.
Why the Squeezy for a post clamp? It was a whim. The Squeezy is a bit of a unique animal and I wanted to try it out. Its a neat idea and well-made.
But there was that sizing issue, which I was only able to temporarily overcome with some shimming. I ended up finding a basic 34.9mm Axiom seatpost collar in my parts pile that I made work. Its an unremarkable part not really suited to this build. Still, it was handy to just install so I could move on to the next job … and wait for my annoyance at myself to subside so I could spend another US$35 for the correct 35.0mm clamp. Its on. It works. It holds my weight over time with no shifting. It looks great.
The Squeezy defines low profile. Note the very light torque specification. The T25 socket adds a hair of security without requiring an additional tool in my onboard toolbag.
The Wire Harness Tube
This was a unique need for this build. The usual preferred solution of a battery bag in the triangle didn’t work on this frame. Not so great news, as you use the battery bag to hide wires. My best solution to hiding otherwise bare wires was to enclose them in a pipe that more or less matches the frame.
I originally used cheap red PEX pipe purchased locally for about the price of a candy bar. I ended up not being happy with the red color and did a spiral wrap of red silicone tape to get a better match. In a short time it darkened to be a near perfect match to the frame. But it also had a few problems:
The tape was just not durable. I had rips and breaks in it – more than in the picture above . Also, I had cut the pipe a bit too long. When turning the bars to an extreme, the fork poked the top of the tube and pushed it to one side or the other.
The solution was to replace the pipe. I used a length of furniture grade, red 1/2″ PVC – a better red than the PEX came in, so no tape. I had to wait a couple of weeks for it to arrive. Cost was about US$20.
The shorter pipe didn’t need as much in the way of fastening thanks to two things: Wraps of more red tape around it provided sticky bumpers that hold the pipe to the frame under the pressure of the hose clamps and the velcro straps. The biggest benefit was the shorter length preventing any contact from the forks. So there’s no longer something trying to push the tube out of line all the time.
The wire tunnel is not a perfect solution, but the alternative is bare wires and zip ties.
So… to paraphrase George Lucas, a bike build is never finished. It is abandoned. And so, for now at least, we abandon the tinkering, building and perfecting of the Apostate. I’m sure I’ll do something to it again as time passes, but for now its time to just do to it what is meant to be done to a bicycle…
11 thoughts on “How To Build An Ebike From Scratch: Perfecting”
Just scanned your blog. Interesting read, especially the frame choice and the battery attachment. Having just completed mine 1st build on a Specialized Stump Jumper I know the battery space in the triangle was my biggest concern. Even though I looked at Luna I did not realize they did magnetic mounts. I did end up mounting under the bottom tube and using the Luna No drill kit and cinch straps which seem solid. There is a ‘sweet spot’ in the triangle but it is problematic . I may visit that later.