Jump to content

suicidebike

Members
  • Content Count

    8
  • Joined

  • Last visited

Community Reputation

28 Excellent

About suicidebike

  • Rank
    Newbie

Profile Information

  • Location
    Internet

Recent Profile Visitors

61 profile views
  1. suicidebike

    Strange "Clonck" sound from my V8

    You're right, it'll take some time to get that bad, but you can see from mine already that the key isn't filling the keyway at all, and there's a lot of play in it. The key is right on the edge of the keyway. I don't think you'd need to have much of the key worn away before it slips. Perhaps other wheels have better fitting keys, but based on the consistent videos with the same clonk sound, I suspect they're all the same. The amount of play won't increase slowly - the key/keyway will wear until it slips, and then it'll go from a few degrees of clonking play to a 360ยบ rotation, then the damaged key will drop back into the damaged slot and maybe it'll slam to a stop, or maybe it'll continue rotating. It's worth mentioning one of the weirder behaviours. When your wheel has this defect and you go over a bump or drop off a kerb, sometimes the wheel will end up in a rapid back and forth oscillation as it tries to correct for the motion of the axle vs rotor. The pedals tilt back and forth at maybe 10Hz, along with the loud clunking sound, and the whole thing feels incredibly unstable. I fell off the first time it happened, and thankfully was only doing ~10km/hr so I only had a few scrapes. I found it difficult to predict when this weird oscillation behaviour would happen, but I definitely noticed it becoming more common, presumably as the key/keyway was wearing further. A few days ago, I experienced continuous oscillation for maybe five seconds - I stayed on the wheel at very low speed, but I couldn't stop the oscillation until I got off the wheel, turned it off and back on. Very unsettling, and I didn't ride it after that. That's the event that prompted the full teardown you see above. Nasty stuff.
  2. suicidebike

    Strange "Clonck" sound from my V8

    Finally, the results: After the repair steps above, I rode this wheel for ~10k at full speed, something I hadn't done on it for a few months because I was worried about the worsening clunk problem. It was solid and performed very well. I dropped off a few kerbs, and it felt firm and controlled. Totally fixed the clunk problem, and I'm very happy with it. Full confidence again, and I'm very happy! Now, I have to repeat all this for my other wheel. To anyone experiencing this clunk problem: I wouldn't ride that wheel without extra caution. Go slower, don't drop off kerbs, be more prepared for a crash than you are normally. Be aware that this failure is a progressive thing - it's going to deteriorate and at some unpredictable point, your axle is not going to be connected to the rotor anymore and you're going to faceplant. Hopefully all this will be useful to people here, and I hope it provides the necessary evidence for you to get warranty service from your retailer, and hopefully InMotion will take notice and fix this problem at the source for their future wheels! I saw a video above from someone with a V10 exhibiting the same problem, proving that they haven't fixed this for the V10 model. To address the 'just tighten the screws/nuts' comment that I've seen in this thread and somewhere else: Tightening the axle nut is probably not going to do it. For a loose axle nut to explain this problem, it would have to be *very* loose to the point where the axle mount jaws can release the axle enough to allow it to rotate a little bit between forward/back motions. It's worth tightening those axle nuts (I used a 13/16ths ring spanner, a 21mm wasn't an exact match for it) and demonstrating the problem again before you disassemble your motor, but I wouldn't hold out much hope for that.
  3. suicidebike

    Strange "Clonck" sound from my V8

    Alright, so I've repaired this on one of my wheels. Here are some detailed steps explaining what I did. Before we get into the disassembly and repair, I wanted to do everything possible to verify where the problem was before opening the motor cover. One problem I faced was that I couldn't reproduce the clunking behaviour if the wheel isn't installed in the body with the power on. To work around this, I just connected the three motor power phases together. This effectively acts as a passive electric brake on the motor - you'll have a hard time spinning it by hand. It'll work, but it'll offer a lot of resistance because as you turn it, one coil generates a current that is fed into the other coils, which induces a magnetic field in those coils, and those fields all help to slow the rotation of the wheel. It offers a huge amount of resistance to being manually turned, and the faster you try to spin it, the harder it'll fight back. This is perfect for trying to test this clonk problem on a bench without having everything else connected. No, this won't damage your motor unit - these things deal with far higher power under normal operating conditions than we'll be able to produce by just turning it with our hands. Here's what my janky passive brake cable looked like - just three spade connectors: You probably won't need this if you're certain that your clonk problem is the same as what I've described here. It very likely is, but I figure this brake technique might be useful for someone. Anyway, on with the repair. Before you open the top cover, put a mark on the wheel rim and the cover, to make sure you can reinstall the cover in the same position. Why? Well, the bearing might not be *exactly* centered with respect to the bolts around the edge of the cover, and you *really* want that bearing installed in the same position to avoid either excessive bearing wear and premature failure, or having a misaligned rotor that scrapes the magnets on one side. Pulling the wheel out of the lower housing is pretty awkward. You may need an assistant to hold the wheel housing while you pull it out. You may need to give the axle a few impacts on the ground to try to get the shaft unseated from the bearing, and it'll be very slow going at first. The magnets are also holding the wheel in, but the main problem is the friction fit on the bearing. When you do get it free, try to hold the wheel housing and the rotor securely and separate them gently, so you don't smoosh the rotor or the magnets. The first thing to do is to remove all the wires from inside the axle. If we're going to be welding on the outside of the axle, we're going to be heating the whole area up to something far higher than the wire insulation can handle, so the wires have to be removed for this. Maybe if you're a better welder than I you could get away without this... Flip the rotor over and desolder the wires for the hall effect sensors. Take note of the order! Also - hey, they used bamboo to pack the rotor coils in! Cool! With that done, flip the rotor back over and cut the motor phase wires somewhere in the middle: Why cut the motor phase wires instead of desoldering them? Well, that's because they haven't been soldered at the coil end: This isn't solder, it looks more like they wrapped the wires together and hit it with a torch. I can see why - soldering enamelled wire is a pain in the ass, this is a cheap and effective method of binding them together. Hard to fault them for it, really, and I learned a trick. Start carefully working the wires back out through the axle: Wires removed: With the wires removed, we can start working on the axle/rotor interface. First, I used a flap disc grinding wheel to expose fresh metal surfaces for welding over. The whole area has been painted, and the paint is going to contaminate and weaken any welding, so it has to go. More grinding on the other side. I've decided to repair the original weld, and add two more welds at roughly 120 degree intervals around the axle. Mid-welding. Allowing lots of time for things to cool between welds, because I really want to avoid any warping or excessive heat transfer to the coils at the rotor edges. I scraped off some paint on one of the rotor arms to allow for a good connection on the ground clamp. Not my best welds, but good enough and a lot better than what was in there from the manufacturer: Some of the messier bits in there are actually the circlip that originally held the axle in position. Once it cooled, I painted the welded/ground areas with a black enamel: I fed the wires back through the axle, which was easier than I expected because the wire path inside is actually slanted at the rotor end, so the wires come out the hole cleanly, instead of me having to pick them out of the axle with tweezers as I feared I'd have to. I resoldered the wires for the hall effect sensors, and soldered the wires for the motor phases. Next, I needed something to seal the hole in the axle that the wires pass through. This was sealed with some flexible silicone from the manufacturer, presumably as extra moisture/dust protection. I didn't have any silicone sealant to hand, but I did have some of this hand-formed air-curing stuff: Sealed up the hole nicely: Drop the rotor back into the wheel housing, taking care to avoid damaging the rotor or the magnets. An assistant to hold the wheel housing steady might be helpful. Put the wheel housing cover back on, taking care to reinstall the thin washer on the axle (see above, bottom of the axle beside the "mossy green" goop) before you put the cover on. That washer will come into contact with the bearing in the wheel housing. I put fresh threadlock on the housing bolts. The housing is aluminium and you can't put much torque on those bolts for fear of stripping the threads, so threadlock is a pretty good way of being sure they won't vibrate loose. Finally, I put some threadlock on the main axle nuts:
  4. suicidebike

    Strange "Clonck" sound from my V8

    Yeah - if the weld is decent, you don't need the keyed shaft. Under the weld there's a circlip. I suspect the original design involved the key and the circlip, but there's enough play in that design that I suspect early testing showed it didn't work well and so they added a spot weld. With the circlip still in place, it really looks like an afterthought.
  5. suicidebike

    Strange "Clonck" sound from my V8

    After a bit more thought, I suspect that if this fault isn't corrected with better welding, the oscillation of the key in the keyway will eventually wear the key and keyway enough to allow the axle to rotate freely. That's going to happen suddenly, it's just going to break free without warning. At that point, the wheel is mechanically disconnected from the pedals and a faceplant is the likely result. In the pictures of the key and keyway above, you can already see the keyway deforming, and the key doesn't even look big enough to properly fill the keyway. It'd take a lot of riding for the fault to progress to the point of total failure, but mechanically, it's inevitable and I'm glad I dug as deep as I did before I continued riding it. I have a pile of pictures showing the steps I took to repair this, I'll post those next.
  6. suicidebike

    Post disappeared?

    Great, thank you!
  7. suicidebike

    Post disappeared?

    Ugh, I haven't been able to reply to this for hours because of the restrictions on new user post frequency. I can see Keith's point - I've been involved in moderating and administering big forums before, and contributed to many more. Keeping out spam and disruptive users is frustrating, and I totally understand restrictions on (1) moderating posts by new users, (2) disallowing PM sending for new users, and (3) limits on post frequency for new users. All sensible decisions in the context of an Internet forum. On the other hand, I didn't expect to be partially doxxed by a moderator - that's bad form, and is a great example of exactly why I generally post under a pseudonym no matter what I'm doing on the internet. Being judged based on what part of the world I'm in, or where my email address appears to be, or how many personal details my YouTube account does or doesn't expose, is not something I expected. To Keith's specific points: The email address I used isn't "spoofed", it's a real, active email address. It passed the forum's email verification. Try sending me an email, and I'll reply from it. It's as real as any other email address, and I don't know what you mean by "spoofed". The YouTube account - it's one I keep solely for the purpose of posting videos and it, very carefully, doesn't publish any of my personal information. My behaviour so far is that I've provided one high quality post about the exact cause of an InMotion manufacturing defect that causes a safety issue, and I think it's highly relevant to all InMotion riders. Pretty good behaviour, I would have thought, and the content alone should show that I'm worth some level of trust. Keith's decision was well-intentioned; protecting the forum. However, it had the side effect of hiding this defect and safety information from the forum, which I believe is detrimental to all EUC riders. Keith also didn't PM me about it - he just hid the post, and went on his way. I had to make this public post out of confusion about where my held-for-moderation content went. As others have noted, behaviour trend (post history and post quality) is the only real indicator of trustworthiness in any community. There are a lot of real reasons why people might want to use a pseudonym on the internet, and the EFF have a great article discussing many of them. People like Keith in a privileged position may not have anything to fear from putting their details on the internet, but maybe I do. Having said all that, Keith's the boss here, it's not my show and I'm not in a place to dictate any rules to anyone, so *shrug*
  8. Hello, I added a (lengthy and detailed!) post to a topic on the inmotion forum an hour or two ago, and since I'm a new member it was pending moderation - fair enough! But er, now it's gone, and I can't see it in the awaiting-moderation state anymore. Is that normal for this forum software, or was it just deleted? I'd PM one of the moderators but again, since my account is new I can't send any PMs, so I guess I'll have to post here! Sorry about the noise. Thanks!
  9. suicidebike

    Strange "Clonck" sound from my V8

    Hey everyone, I've had this problem for months, on two different V8 units. Today, I figured out what the problem is. There's a weak weld inside the motor unit, between the rotor and the axle shaft. These two pieces are supposed to be one solid unit because of the weld, but in mine this one weld has broken. Here's a video demonstrating the clunk sound we're hearing, and the broken weld that causes it: The weld isn't the only thing that stops the axle and the rotor spinning freely - the axle shaft is keyed to the rotor, so they can only spin independently a few degrees. The clonk sound we're hearing is caused by the key between the axle and the rotor hitting the edges of the keyway. Here's the disassembled motor unit, showing the rotor and the axle: Here's the cracked weld in two positions: Finally, here's the keyway on the other side of the rotor, showing the two positions that our wheels are banging between, causing the characteristic clunk sound: I believe this is a manufacturing defect. Many of us are having this problem, and in my opinion as an (amateur, to be fair) engineer, one small weld on this part is not enough to stand up to the stress placed on it during rapid acceleration, particularly if you drop off kerbs or otherwise give your wheel some shocks. I don't consider any of the things I've done with either of my wheels abusive - dropping off small kerbs is by far the worst of it. The disassembled wheel has ~1500km on it and one year of commuting use, and the other wheel has ~300km, 95% of which is on very flat ground with no kerbs. Both wheels show the same symptom, and none of my use is abusive. I firmly believe this is a manufacturing defect - that one small spot weld is not enough to stand up to normal use. I'm fortunate enough to be able to TIG weld, so I'm going to attempt to repair this wheel myself and I'll post another update with instructions when I've done that. I own two wheels with the same symptoms, but have only disassembled one of them. I've reached out to the sellers of both, and I'm working with both on getting replacement parts... but will replacement parts have the same manufacturing defect? Hmm. Hope this is helpful for someone.
×