My Puzzling Ebike Problem: An Interim Report

By Stan Purdum

In his RBR column a few weeks ago, Jim Langley, RBR’s Technical Editor, published an email conversation he and I had regarding a puzzling ebike problem I was experiencing. He did so with my full consent and our shared hope that you — our readers — might be able to help us figure out the cause of this problem. 

You can read Jim’s column here, but in summary, when riding my Specialized Turbo Creo SL Comp E5 Road Ebike with assist turned on, the bike performs perfectly. But when I ride without the assist engaged, it feels like the drivetrain is completing the rotation before the crank itself does, so that during the last 10 or 20% of the rotation, the crank “falls” to catch up with the spin of the drivetrain. And when that happens, there is a noticeable clunk sound. 

In the comments section below Jim’s column, several of you made thoughtful suggestions, and I’ve spent some time investigating those possibilities. None of them have led directly to a resolution of the problem, but they have helped us understand it. Since you were generous with your time and reasoning, I believe I owe you an interim report on what Jim and I now think is going on. (I’ve called it “interim” because there’s still room for further investigation.)

Here’s what I did in response to your suggestions and to related thinking your comments prompted:

1. I had already gone over the bike carefully to ensure that nothing was loose or overly worn.
   
2. I tried riding with not only the assist turned off, but with the motor itself turned off (there’s no convenient way on the Creo to disconnect the battery from the motor, as they are connected internally). It made no difference: The crank fall and clunk were the same whether I turned off the motor or only the assist.
   
3. I removed the range extender auxiliary battery, just in case that was somehow influencing the problem. It made no difference either way, however.
   
4. I tested the Creo’s DT Swiss hub to see how direct the cassette’s freehub drive system is. I did this by lifting the chain off the cassette, so it was not touching any of the cogs. With my free hand, I turned the cassette backward until it clicked, and then I rotated it forward to see how quickly it engaged. It engaged immediately. I repeated the process several times, and it engaged at once every time.
   
5. I checked to ensure that my firmware updates had all, in fact, taken place. They had.
   
6. I spent some time studying my pedaling stroke, pedaling one-legged on first one side and then the other. I could find no hesitation in my stroke.
   
7. On a windy day, I rode the bike several miles with the assist turned off. I discovered that when I rode into a stiff headwind or pedaled up an incline, the crank-fall problem mostly went away, but as soon as I turned out of the wind or topped the hill, the problem returned. 

Test #7 got me thinking about my early days of cycling, when my (non-electric) bike would get going fast (such as rolling down hill) and I’d start pedaling, but because I was in too small a gear, the crank couldn’t keep up and would “fall” until I shifted to a higher gear. So on this windy day, I started riding in higher gears when not using the assist, and the problem was somewhat reduced, with the fall happening only when I slowed my pedaling a bit.

So I’m wondering if my natural cadence at my current age (79) is somehow out of sync with the gearing on the ebike when the assist is off. (Again, there’s no problem or mismatch at all when the assist is engaged.) Could “old-man cadence” be the issue? I don’t have the problem on my non-electric bike, so I’m not too convinced about that, but the Creo is as close to a high-end road bike as I’ve ever had.

I talked to Jim about this, and he thinks I’m onto something. But he suggested describing it differently: “Instead of a cadence problem, I think it’s a gear choice issue, in that you are now choosing easier gears more of the time and that means that if the bike picks up speed quickly, it’ll take more of the pedal revolution to ‘catch up’ to the rear wheel’s rate of rotation.” 

Jim continued, “So I think trying to shift earlier into larger gears as much as possible will reduce the issue further. Or maybe just trying to remember to ride in gears slightly harder than what you’re naturally choosing now. It might even be worth it to make a gearing chart and track when this happens. Maybe a different cassette would result in you being in the ‘right’ gear more of the time and not experiencing the falling pedal issue so much.” 

In line with Jim’s thinking on this is what happened when I took the Creo to the bike shop where I purchased it. The shop owner inspected it thoroughly and could not identify a problem. Also, when test riding the bike, he could not duplicate the problem, leading me to surmise that the issue might be related to my stroke or cadence. The shop owner is younger than I am, and he was likely pedaling more vigorously.

So that’s where I am for the moment. There are still some things from your responses to Jim’s column for me to follow up, including possible speed sensor issues, and the whole list of Creo-specific possibilities that reader Harvey Miller provided.

Fortunately, as already mentioned, the problem is not present at all when I have assist turned on, and I use assist for most of my riding these days, so the pedal-fall issue, while annoying, isn’t keeping me off the road. And exploring the problem is helping me to learn more about the workings of my ebike, which is a good thing. 

Thanks again for your help.

---

Stan Purdum has ridden several long-distance bike trips, including an across-America ride recounted in his book Roll Around Heaven All Day, and a trek on U.S. 62, from Niagara Falls, New York, to El Paso, Texas, the subject of his book Playing in Traffic. Stan, a freelance writer and editor, lives in Ohio. See more at www.StanPurdum.com.