Link to the cranks and axle in description. Copped both for about 200 quid. Let me know if you understood the bit about the taper geometry. I kept hammering it home but don't know how obvious or not it is to people. Edit: reading some of the comments, it seems this design isn't the most user friendly or clear to install. To help those with issues here are some vital points: 1. You MUST back off fully the preload collar before assembly otherwise you can deadlock the BB bearings. 2. If your BB shell (thus bearing stance) is slightly too wide, doing up the tapers to the full torque may compress the BB bearings and over preload (deadlock) them. In this case, remove the plastics 0.5mm shims and oring. This will give you up to 2mm more shaft to play with. 3. If the 35Nm wont fully lock the chainring spider or power meter, tell Rotor its out of tol. If they mess you around and chat bs, then gently heat the crank to 60c and reinstall with grease on the spline. This will slightly enlarge the female taper.
As a machinist who has in the past needed to nail axial positions of shafts relative to their tapered sections I can vouch for every statement you made about the production of a part like this. Very tricky! In fact I might wager that they may cut the tapered section in one operation then locate that part on that section to cut the corresponding shoulder. Even then, most lathes would be at or beyond their ability to position a cutter accurately enough to put that shoulder where it needs to be. If it’s a steel axle, I would be looking at cylindrically grinding that to size. I too have swapped out my Shimano crank for a modular design but I went with Easton. Their solution to this is quite different in that the chainrings get splined and bolted to the crank arm before that whole sub assembly gets mated to the spindle. Very clever and it works great despite its aluminum axle. (One piece Hambini bb ensures there isn’t undue wear from misalignment.)
Nice to get a machinists POV! I don't envy those who have to make tapers to axial length specs. What you said about mating them and then cutting the shoulder would work, but then they'd be matched pairs (crank and axles). The fact that they're not matched serial numbers kind of makes it impressive i think.
Except he is mistaken about how the rotor is produced. This will be machined on a Muti-axis CNC lathe with in one setup with additional milling axis. It is commonplace technique in precision machining.
I never thought of putting a pen mark on my male taper and then testing insertion depth, but this might be a good way of testing the dimensional tolerances.
I can’t help but be weary of Rotor cranks after the experience I had with the one that came on my Cervelo. It creaked like crazy and once I swapped to Shimano the creaking went away. 5 years later, still no creaks, BUT I’m on my second Ultegra crank. Pick your poison?
@@shamuslamont100 the BB was Rotor too, yet I haven’t had issues with Shimano or SRAM Dub BBs afterwards. I am currently running a SRAM GXP crank on that frame (long story, lots of crank swappin this past year) and the BB is from a brand I’d never heard of… yet it doesn’t creak. I guess the Cervélo reps I met at a race some years ago were telling the truth about that Rotor crankset from 2014-2016
Square tapers went out because indexed front derailleurs demanded chainrings remain in the same position relative to the frame after every bb overhaul, according to Shimano. How often do you need to overhaul the bb? couldn't you readjust the derailleur if something changed? isn't that exactly what we did over decades of shift-by-muscle memory and square taper bb's? OTOH external bb cups are cool, but they were a patch for the poor durability of small sealed bearings inside a BSC bb shell with the 22mm ISIS spindle. Then again a Phil Wood sealed bb lasts pretty much forever especially if you own more than one bike.
@@slowerandolder Front derailleur argument seems ridiculous. Good square tapers last 50k km.... smoothly. Surely adjusting a front derailleur isnt a big deal at that stage. The main problem seems to be that this solution is somewhat heavier and indeed like the ISIS standard shows very difficult to oversize to get to a lower weight. Not sure if its less stiff too, but track sprinters still use dura ace track cranks, so probably its a non issue too.
@@ohne_speed that's interesting - in the 1970's, as alloy cranksets became common on cheaper bikes, shops occasionally saw wallowed out left arms; customers would blowoff the free checkup offered on new bikes, the left arm would loosen a bit, the oblivious owner wouldn't notice until the arm's tapered hole was deformed = new crankarm. Shops had been torquing that left arm's retaining bolt properly, but properly also required re-torquing after 50 miles or so. Once this retail reality was recognized, the shops started doing up the bolts as hard as possible - which shortened the number of times the crank could be reassembled (after bb overhauls) but no one ever complained about that. This was never a problem for driveside crankarms, since that pedalling load went straight onto the chain.
I have two ROTOR Cranks, one in 30 mm, one in 24 mm, both DM. They perform really well and I do not have any issues with them. They are not cheap, because they are high grade products. What I like about ROTOR cranksets is, the seem to be more quiet than Shimano ones I had previously (R8000). As I am not an expert, I think, less noise, less friction, better quality.
It could be that the tolerances don't need to be quite as tight as you think. There is a phenomena known as elastic averaging for coupling designs like this. Your analysis is correct if the material is assumed to be perfectly rigid. The shoulder sets the axial position, and the rest of the material will elastically deform to allow this to occur (which is why you need to use the force of the screw to draw the assembly together). To extend your comparison of machine tool tapers a little further - dual contact and HSK interfaces have a shoulder perpendicular to the axis of the spindle such that an accurate z-location is set regardless of differences in the tapers. The requirement for this is important as rpm of the spindle increases as the hoop strength of the outer "taper" (spindle) can actually open up and the draw bar can pull the tool holder further in to the spindle. Heat is also something of concern, which you note. I haven't gotten my hands on one of those Rotor shafts, but I would imagine that they are machined with a swarf style 5-axis strategy in a mill turn machine. In cycle probing and cutter compensation would make this a somewhat straight forward task with a straight fluted endmill (no custom tooling required),
Nice comment. Someone out there knows their stuff! With relatively thin wall tapers yes they can elastically deform, but the tol will effect the amount of force required to hit the axial end stop. It was my belief that tool tapers being virtually solid there was barely any elastic give in them, which would mean an taking an accurate z cal every operation. But my machine tool knowledge is limited!
@@PeakTorque The tool holder itself might not "crush" inwards much, but the machine spindle is a hollow cylinder at some level of analysis. Another way to look at it: make the assumption that all the torque applied to your bolt is being converted to linear force at the pitch of the thread (no friction losses) if that force is translated to the 1 degree taper (on the coupling) that is a hell of a lot of expanding force! Wedges are really good at expanding things. The HSK spindle interface actually clamps the inside of the tool holder for this reason - as RPM increases the clamping force expands outwards. For your cranks: you could check this relatively easily on the crank arms by mic-ing the "width" across the interface area in a relaxed state and compare when it is assembled and torqued. Assuming you can repeat your measurement location accurately, I would imagine you see a difference (cross section growing when it is assembled). My point is simply that the torque spec on the bolt may be the "dominating" variable so long as the components are machined with relatively good tolerances. IE the tolerance window on the machining is a little wider than pure geometry / trig allows for, because nailing those tolerances would be next to impossible - not to mention the varying thermal expansions of the discrepant materials. Black anodized alum might get hot enough in the sun to fall off ;)
This business with the tapers is quite bad on SRAM GXP and others. It messes up the chainlines and you end up having to mess around with spacers. The "average" bike mechanic can't get their head around it.
Since this design locates the spider on the machined shoulder (which sits on the drive side bearing) wouldn't the chainline be exactly determined? I mean to the extent that the frame is manufactured precisely anyway.
I don't think that applies to this design. Both of the arms bolt down to the axle, and then the preload collar is expanded to tension the bearings. The only need for spacers is for different BB types.
@@galenkehler ON the GXP, the non drive side bearing is fixed so it determines the axial position. If the taper nips up more or less then it moves the effective chainline. SRAM obviously know about this because their solution is to add more grease and clamp it up again.
@@Hambini I'm just a bike mechanic so apologies for my lack of engineering intellect but are you referring to DUB, not GXP? I only say this as with GXP on the non drive side the axle butts up to the non drive side bearing. The NDS crank arm then compresses to the opposing side of the axel on NDS bearing. The chain-line is determined by the location of the NDS bearing and the only way to manipulate the chain-line is by moving the NDS on the BB shell left or right. With dub you have a series of spacers which you can replace etc to manipulate your chain line should wish to stray from Srams instructions.
Back in the Isis bottom bracket spindle days, Race Face cranks specced a particular torque but also said it had to bottom out. Their instructions said if the arm doesn't bottom out at that torque, to remove the crank and reinstall however many times until you did. I've installed several Aldhu cranks. Sometimes I need to remove and reinstall once to ensure it's bottomed out.
I had an Isis BB and truvativ cranks. The tapers never mated, and always came loose and ruined themselves. They never worked well. These spline tapers are risky!
That's interesting. Basically they were factoring a bedding in it seems. Im curious to see if a wave washer could be utilised to add some preload to allow for any tolerance variables or fluctuations. The design should allow for and bedding in of the taper.
I have 2 of these both linked to powermeters- one on a power2max and one on an xcadey xpower-s. I haven't seen any problems so far and both read consistent with a quarq riken spider, and all 3 powermeters are within 5% of a elite direto trainer. Happy with the performance.
It looks like this design has even more in common with plastic cups then with a lathe rear tool post mount: cranks must be undersized and considerably soft to press onto the axle with plastic deformation
I 100% agree. There is no way to get a proper contact and fit on both interfaces without one of them deforming. If it works it is still quite impressive.
I love the power meter setup on these, and they have used the same interface for quite a long time so there are actually plenty good used power2max meters available online. You can also play with crank lengths quite easily. After spending years riding BMX where this kind of design is the norm, I have always found two piece cranks quite annoying in comparison considering how the whole thing needs to change if you change bike or damage something.
Historically road and MTB cranks were 3 piece setups back when most of BMX was running one piece cranks and oversize BB's with press fit cones. Then BMX moved to 1.37" threaded BB's which 1 piece cranks couldn't be inserted through, so they moved to 3 piece cranks. /In the meantime road and MTB moved to 2 piece cranks and external BB bearings. I expect it's more the case that BMX hasn't moved to 2 piece yet rather than that 3 piece is the end of their evolution. Sadly the bike industry has never cared whether their designs mean dramatically increased expense when part of a one-piece assembly gets damaged and you're forced to scrap the whole lot. It is possible to buy individual left cranks from Shimano, which came in very handy a few years ago when the pedal threads failed. The UK importer of Shimano doesn't do any of this though, nor do they even admit it's an option as it's more work for them for less profit. Just go to Shimano's own website, look up the exploded diagram of the crankset to get the part number (Usually the PDF you download from Shimano with the diagram is actually an order form because Shimano expect you to be able to order individual parts), and google the part number to find sellers - There are some decent shops in Germany that support these orders, bicikli.de/shop/ is the best one I've used for having Shimano parts in stock - they're excellent for things like freehub bodies and ball bearings for hubs too.
It is worth pointing out that if you had any play on the powermeter spider it would not all have beeb Rotors fault. Its a 3rd party power meter so if that would be out of spec then Rotor could have got is 100% right but not get it clamped down because of it.
The dimensional tolerance of the power meter in this application is simply the width of the splined portion. Getting this right is trival compared with the matched taper.
@@galenkehler Yes you are totaly right about that. They still have to make it to those specifications. I am not saying its right or wrong it simply not gets mentioned.
Hey, first of all, thanks for the video. I recently bought a Xcadey powermeter (PM from now on) for Rotor aldhu cranks. As you said, the interface was not perfect between the powermeter, the cranks and the axle, so there was an unnaceptable amount of play. I took it to my local bike shop, thinking that maybe I did not know how to assemble it correctly, and they returned it to me assembled, and said that more force was needed in order to get a snug fit. Weeks go by and the PM starts returning some wonky readings until it eventually dies completely. After dismounting all the pieces I discover that the PM outer cover is cracked, undoubtedly because my LBS applied too much torque. I have to add that Xcadey sent a new one, almost no questions asked. This second unit I mounted myself (not going back to that shop) and had no troubles, in fact it´s still going on trouble free after 6 months. I guess that the problem lied within the first PM mounting interface, since that´s the only thing that changed, but that´s my expierence.
I had more than one problem with these: 1. Two sets produced vastly different chainlines and Crank q on the same bike, also cranks were not centered in reference to the frame. 2. The crank bolts loosen over time, and there is no fail safe that prevents the cranks from coming off (did not happen to me, but it happens acording to the forums) 3. less space between the crank arms and pedal than with Shimano, Quarq (relevant for me during winter when wearing overshoes) 4. Preload ring is an absolute fail! The ring gets compressed when turned. It's really tricky to dial in... need to place a flathead where the ring opens (which does chip the material and introduces splinters into the fine threads. 2 cents of thinking would have at least yielded to put some knurling on the opposing side of the preload ring pinch bolt. 5. Rotor support just stopped responding when asking for support with this. So in summary: I can't recommend those. Really wanted to like them, because they were the only option to stay with a 24mm spindle and using a power2max. Really wish there was a better option for Shimano BBs. Imho they should keep the modular design, but a) change the spider interface b) use a similar design to Shimanos for attaching the cranks
I`ve always had issues with preload on these. 35nm to attach the non drive side arm seems way too much, they clamp so tight there`s no preload to take out with the ring and I`ve tried the same chainset on 3 bikes all with the same outcome. At 20nm it`s usable but I`m worried it`ll come undone. I`m now convinced mine is faulty in some way,
I've had all these issues too over c. 10,000 miles of use. On mine the power meter spider (rotor PM) sits dead square but over time and usually after a few big power efforts the spider twists on the spline. I have to un tighten the bolt and reset and tighten up again. It's currently so twisted Di2 can't handle it with auto trimming on the front mech. It's particularly bad when the bike is in a direct drive turbo.
@@ClayDavies how does the spider twist on the axle? That's sort of physically impossible unless your spline teeth are gone. I have one on a P2m spider and one on an Xcadey spider, no such issues so far.
Thanks so much for doing this and confirming that Rotor makes quality stuff! I'm seriously considering Rotor cranks with Power2Max power meters and don't wish to risk having a damaged crank with Ultegra or Dura Ace!
Great video. My big 1x11 Vegast chain ring came lose on a bike trip. I never fastened it with a torque wrench. So now I understand why it came lose. I found a guy with a big enough Allen wrench and tightened it very well. Back home it came lose with a lot of force and squeeking sound. Now I know that the tapered side caused this. I will use a torque wrench next time, this is precision work!
A fantastic modular system I run myself on the TT bike, with interchangeable spiders/PMs/chain rings. You can get the Vegast crank arms for about half the cost of the Aldhu's too.
Yup. I'm running Vegast cranks on the 24mm steel axle with a 30/46 direct mount one piece chainring (with a Hambini PF30 bottom bracket). Works perfectly on my otherwise Ultegra di2 setup.
I am on my third Rotor crankset. On the first one the left arm came loose and destroyed itself. The second one was DOA. The axle had terrible burrs and so the surface on both arms was destroyed. The third one is working so far.
@@PeakTorque Most of these issues are user error. Insufficient torque or not grease on the mating surface, poor adjustment of the preload. The issue with the alloy axles destroying themselves is usually just bad bottom bracket alignment or poor preload. But sure to check the bolt after the first few rides which is good for peace of mind.
Aldhu 24 + Power2Max NGeco + Praxis Buzz 10/11/12-speed chainrings + AXS road 2x was a nightmare. 1) The cranks hit the derailleur cage when set up to SRAM’s spec. 2) The 5mm wide flat-top chain is can get stuck in between the chainrings because the NGeco mounting tabs are >3.5mm thick. 3) That >3.5mm thickness causes shifting problems because the FD overshifts to a lesser extent while the rear is in the smaller sprockets.
Long-term review would be interesting, i.e. if you’d have any play eventually e.g. from wear. It’s of course normal to torque down chainring bolts time to time, but here it’s quite an exquisit setup.
My daughter has been using these cranks in 30mm version for XC training and racing (high level I'd say, currently 2nd in Poland), including doing some tricks and jumps. After half a year zero problems and it's been offroad riding in technically difficult terrain. In my opinion these cranks are on par with shimano DA or XTR as reliability goes. And the support from Rotor is fantastic. I chose road cranks for mtb because of the need for low q factor. Combined arms with Rotor's XC direct mount XC chainring. Asked Rotor if it would work and they even calculated chainline for me.
Speaking of tapers, what is your experience with square taper bottom brackets, and do you have a clear favourite among them? Always love these engineering analyses.
: I, for one, think that for the time they were good and they lasted a long time(speaking for Dura-Ace but I think a lot of them did) but Octa-Link was the superior way.
24mm axle works only with specific 24mm cranks and 30mm with 30mm cranks and afaik carbon one is only 30mm last time i checked (have both of them 24 and 30mm axles)
Whenever I've looked at Rotor cranksets, I've been shocked at the prices, so that has ruled them out for me. I have wondered about the machining of the tapered splines in cranksets that I own (FSA and SRAM). There's quite a few standards for straight cylindrical involute splines, but I don't know about for tapered splines. The exposed spindle length a.k.a. distance between crank arms (as Hambini calls it) can vary quite a bit due to manufacturing variations.
Yeah, it’s impressive tolerance but, but also kind of demonstrates how the sram or race face style chainring attachment are better. With repeated assembly and wear, the rotor style is likely to reach a point where the chainring isn’t clamped tight and a shim would be needed.
which cranks can i use with sram red etap chainring dub system? would like to go for shorter than 170mm cranks... cant find sram cranks anywhere shorter than that
Very interesting and I always learn something from your reviews (as a total layman engineering wise). Personally I have two bikes with Rotor InPower 1x cranks and they are bulletproof (aside from the single A battery which is relatively short lived) and easy to swap chainring sizes. . Quality machining and really not that expensive especially when they have them on sale. By far my favorite cranks.
I don't deny that that is a very nice piece if machining but modern CNC machine tools should easily be able to hold 10micron tolerance on those size parts so the 50micron you use as an example is a bit excessive. Also - there is always an also - the simple trigonometric analysis does not account for the elasticity of the crank spline hole (and I suspect a lesser extent the axle) which will allow axial movement when tightening the bolt. I have not run the analysis but on my crank-set (see the first link below) once the taper is 'nipped-up', tightening the retaining bolt can push the crank about 1mm onto the taper on the axle. Also, you should have considered the thickness tolerance on the power meter spline area which is also important for the fit to be good enough - how well so Sigeyi control this? I suspect the tolerances on the tapers are not as critical as your analysis might suggest. I bought this Sigeyi power meter about one year ago (th-cam.com/video/wXIaALuMPCg/w-d-xo.html), I have not been able to test its accuracy (its not that important to me) but so far it has worked reliably and the battery lasts a long time. The only real issue I have had is getting a firmware update transferred to the power meter. I'm looking forward to your full review of the power meter. OK, I've been a bit critical. If you want to get your own back, I'm look to get engineers to critique is video (th-cam.com/video/Q4HGhrBXDIM/w-d-xo.html). In particular comment on what I found in the seat-pin would be welcomed.
Great comment. Yeh i didn't go on deformation of the parts, thought the video was nerdy and long enough for most. Always a balance between detail and retention to satisfy you lot! Didn't mention the spider thickness control, that should be very easy to control, but yes its important. I just chose 50 microns as an extreme figure. But even with 10 microns because of the tangent relationship its still quite a large effect on axial position on a fine taper. Cheers
there was a reason we abandoned tapered axle and cranks long ago. I used them through the 70's and 80s. . Shimano engineered the new axle design - and we have no problems.
Hey PT! Many thanks for the video. I had Vegast version if the crank (cheaper version) and I had this very problem. Crank arm wasn’t pressing all the way to the spider and had this terrible creak going on. I contacted Rotor and they eventually replaced it-they tried to convince me there was some bb/bearing misalignment but I did try other 30mm crank which worked fine. In the end, I sold it as I wasn’t keen anymore.
Funny that you say this. Just got an Aldhu 24mm and no matter what I do, the spider is loose. I also tried googling this but didn't find anything. I guess I'm the only one with this issue... I called Rotor, they told me to go to 40Nm and it still wouldn't compress. I can't seem to win with any brand.
Thanks very much for this comment. This was exactly the problem I had envisaged! So there are problems with QC then. Try this: grease the spline, heat the female taper (the crank) in an oven to about 80c, then install. Should get you a bit further up the spline.
Did you check the inside of the crankarm where it mates with the spindle? You may have squashed the washer that's inside there. This happened to me when I was installing one, supposedly it somehow ran off-centre at some point.
A similar review of SRAM’s cranks would be a good video idea. They have a much higher torque spec of 54Nm. I think they may have more issues with cranks coming loose despite the higher torque spec.
@@yonglingng5640 I’ve had one come loose. They also have an FAQ on their website stating if it comes loose and you continue to pedal it can deform the splines and you’ll never get it to not come loose again.
Always loved Rotor stuff. The 3DF made the BB30 on my old bike actually usable. SRAM DUB looks like a not-so-cheap knockoff of 30mm Rotor down to the preload adjuster. Difference is the Rotor one is beautifully machined whereas the SRAM one is a piece of plastic tat.
To add to this, doing a build for a customer have convinced them to go for a Rotor Vegast rather than Ultegra crankset (in part due to availability it must be said).
Not to disagree on the general importance of getting tolerances right in over-constrained interfaces like this one. But even if the male part was just ever-so-slightly too big (or the female part too tight), you could probably still torque it down flush against the lateral stop; the crank-arm portion of the interface would likely just stretch a little more. With my set (30 mm), I can kind of tell as the preload pinch nut, while being very (like, very!) loose on the pinch bolt before torquing down the left* crank arm, will seize considerable after doing so. Also, the 3 plugs in each arm will usually become clearly visible after torquing down the cranks. And stretching holes (of cranks, way beefier around the hole than the rotors) has been a thing for decades ... well, ok, this might be easier with a tiny square taper compared to a huge spline taper, but still(?) * I was told by some Rotor person that the left arm, pretty much the same as the right one, bottoms out against a lateral surface when mounted as intended. Side note: Sram has been mass producing that kind of over-constrained interface (tapered spline + lateral stop) since the introduction of gxp. Now, by no means, am I trying to say they nailed it or that there weren't people having issues with it in the field, just that, by and large, it kind of can work. Another side note: I still find it somewhat funny how we went from 3-piece cranks, some designs with detachable spider, to integrated/2-piece to, now, fully modular (kind of 3 piece again). (Sorry for shitty English. Also, not really an engineer. I like your videos much more than the ones from the French aerospace guy :D)
Good review and well explained. I have the same combination on my bike and will check it out later. My question: If the chainring is lose like you described at 8:47 and can angle would that not result in a better more straight chain line?
The Aluminum Aldhu turned out to be to flexible for me, the pedals end up at a jaunty angle under the force of standing pedaling. The new carbon ones are supposed to be better but haven't tried those yet. Rotor are bad at q-factor. On my carbon road bike there is >1cm on EACH side between the arm and the frame. They could easily have narrowed the q factor with straighter arms and increased stiffness at the same time.
I had the center bolt that affixes the NDS crank shear off inside the axle on a ride. Rotor attributed it to me not installing it properly (I had put a few 1000mi on it since the last time I had taken anything apart) but ended up sending me a full set of new arms and axle. I had only needed the center bolt, but oh well. Otherwise I've been happy with these married to a power2max spider and praxis chainrings. The old rotor setup I had prior to this (the 3d+) had a catastrophic failure at the pedal spindle interface and I broke the whole end off the crank arm.
They use this fine spline for spider mount interface just to allow 1 degree clock of direct mount oval chain ring. Want to dial oval chainring orientation a bit? No problem, take it out and reinstall it rotated one degree clockwise or counter clockwise at a time until you find the perfect angle. As we use round rings, this design complexity give us no benefit.
I had to torque more than 35nm to remove the play and still have creaking with 24mm and Hambini BB. Doing my head in. Think need to speak to Rotor as Power2Max were responsive but not helpful.
Hello, I have this wobble too. I screwed the crank, the shaft and the Power2mx together, but there is 1-2mm play and the crank wobbles, what could that be?? In your video you can see it at 8:45, thanks for help
Nice explanation! I recently bought a power2max NCeco with the aldhu 24mm for my 18 scott foil (PF86 BB) . My drivetrain is a r8050 ultegra di2 so I kept the shimano chainrings. However, when I first installed the aldhu crankset, I noticed the crankset is sitting slightly further than my ultegra crankset. I noticed that because the chain is rubbing the front mech when I am on the big chainring. So I thought no big deal, I will just adjust my FD. So I got it to work, but since then, my drivetrain is super noisy when in the big chainring (worst when cross chaining). It's been a few months and still the same. Any ideas? Appreciate for your input!
I think Rotor headquarters are in a little village outside Madrid(Ajalvir). They build their stuff at home. I had a few Rotor products in my hand,very good looking but never try one.
@Peak Torque I have just purchase at 24mm Rotor crank set and 1-piece chain ring. The machining on the chain ring is outstanding! When you torqued up the arm, did you apply force to the pedal with a chain attached to mate the splines in the best position as there's a fair bit of lash between the ring and spindle. Thanks!
My Rotors click (30mm DM version) and haven't been able to resolve it until this video. I've had them for over two years now but it was already clicking under each (driveside) pedal stroke after a few months from new so I suspect the DM and crank interface developed some give resulting in the click. Oddly my older Rotor 3D+ cranks (24mm) on another bike have been silent. Both of my Rotors have Q Rings if that makes a difference?
I have noticed that you went back to quarq on gxp I suppose, any reason why you chose gxp over 24mm aldhu. I m currently choosing my crank set up and can't get passed by disadvantage of gxp
@peak torque just seeing how your going with these cranks?? I'm building up a new bike and leaning towards these cranks. Do you have any feedback please after 6months. Cheers
@Peak Torque how much clearance have you got between the drive side crank arm and the outer derailleur cage? I've got P2M +3d24 and there isn't enough clearance. If this works well I might go down this route instead. Are you using Mechanical or Di2? Thank you. GP
i have a Aldhu 24mm and even when torqued to way above the torque setting the chainring still wobbles. tried the chainring on my aldhu 30mm crankset and it fits perfectly. i might need to get a spacer to make the crank arm compress the chainring enough to keep it tight.
I would imagine that the elasticity of the crank material at the hole plays an important role in this fit too. the relative fine taper can be an advantage then
Is the left crank axially restrained on the face (similar to the drive side) using the bearings (I'm thinking with a spacer keeping the inner races apart) or is the tapered spline the only thing holding it in place? EDIT: It doesn't seem to be this way looking at it. But does the end of the spindle bottom out against a face on the inside of the splined crank? I'm worried about the splined fit ''squirming'' under load if you will, and fretting wear occurirng between the crank and spindle due to elastic backlash under load.
I wonder, is the taper issue described the reason for the failure rate I've seen (2 sets of my own) in the rotor flow cranks. Cracks in the crank arms at the spindle interface.
I've got a guess here, if the dub spline's got a taper of 2° and preload of 54Nm, and the rotor one has a 3° taper, it'll only need about 35-36Nm to achieve the same amount of radial preload, which is what is recommended. I've also got another question, why none of these manufactures make the spline out of tangent arcs, instead they use this "small round cutouts in a big circle"? (hope I made myself clear)
@@endercrafts9056 it doesnt work like that. In theory the shallower the taper, the less clamping force you need. Larger angle tapers actually need more force! As you have less mechanical advantage in a broader wedge. Second question: probably because they cant reliably rely on the locked taper to transfer pedalling torque so they use a very simple spline shape (easy to program) on cnc.
Great style of explanation of the important engineering details. Subscribed. I saw this vid as I was looking at these crank sets for a power meter. I really don't like the idea of taper and spider clamping combined as you describe... Even if they do manage to get it to work in this case. I'll probably look for other brands. Is this the only option for these style power meters for shimano rings?
Ive not long but the cheaper version for my 2007 cannondale Caad 9 and what a great idea, rap going forward as can fit most bbs and if you buy the spider any chainrings!
My challenge in clanging my chainset to rotor was in the quality or compatibility of the chainrings. I will not do that on my new bike. I will opt for a 4iiii powermeter rather than rotor
@Peak Torque, I went with a 24mm 3-bolt setup, but full Chinese - Sigeyi AXO with Zrace Hardrock cranks. I'm preparing to install and have found there is a bit more backlash than your splined Rotor setup. How concerning is backlash with these power monitors' accuracy? Is the issue that some level of accuracy is lost for a moment in each pedal stroke?
It won't be good for the zero offset and repeatability. You'll also get a lot of noise and lack of stiffness from the chainrings. Do the cranks up to reccomended torque and then measure the gap. Use thin shims to fill the gap.
Hi!! Did you used all the washers and o-rings that are shown in the video when installing the crankset in the Giant Tcr BB86?? I’m asking because I bought a set of Aldhu’s second hand and thet came without any washer and o-rings, to know what to buy to install them properlly in my Giant Tcr.
I actually think i didn't use any orings or the washers, because the tapers were squeezing these too tight when everything was clamped up. So just left them off, which is counter to what the instructions say
If I'd ever tried to put something like this design into production it wouldn't have got past the first design review. It's just dumb engineering design requiring stupid levels of machining accuracy to dig themselves out of a hole created by a fundamentally stupid concept.
SRAM solved it with 3 screws, race face/shimano with a threaded lock ring, neither of which require special tolerances. The main reason I can think to do it this way is if its a patent issue and they don't want to or can't pay the licensing fees.
Yep I agree. Its overcomplicated but Rotor seem to like doing that. For me the best way is clearance spline plus pinch bolts but that's patented, somewhat ridiculous, but that's how it is.
Has there ever been a crankset with a threaded interface to secure the crankarm to the spindle? (As in you turn the whole crank arm to affix the crank arm.)
Have an ALDHU and it isn't a straight swap with my Ultegra... chain-line is vastly different so I need to adjust the front derailleur always by a lot when switching
Would like to hear your thoughts on the Cannondale si series cranks. Same concept, although the spider is screwed down onto the arm, which is then bolted on with the aforementioned taper
In that case, you always know the chainrings are held on nice and stiff independently of the crank which is actually in my mind a better solution. Cannondale cranks don't seem to fail, even though theyre hollow (like shimano)
I did exactly this. I got the power2max track powermeter on the rotor aldhu, and it gives a perfectly centered chainline (42mm)on my disc brake road bike (142mm rear axle)
What is the order of the o-rings and spacers? Should the spacer touch the bottom bracket or the crank on both sides? Trying to find the same with the o-ring on where it should be located. I'm trying to figure that out for the install and can't find a good guide. Thanks!
if i remember correctly i did not use the oring or spacers as it made the whole lot too pinched when clamped up! And added too much preload to the shimano bb. That is the risk with tapered splines grrrr!
One thing that bothers me is the use of o-rings on the axle. O-rings eventually dry-rot and fall off/out of things like this. Usually when you still have sixty miles to go before reaching civilization. Also, they crush during assembly and use, then when you go to replace the bearings or swap the crank to a different bike, they don't fit properly anymore.
I'm curious, is that taper absolutely necessary? A non tapered spline interface would have worked too, unless you are concerned that too much clamping force is applied to the chainring spider at 35Nm make up torque.
For an untapered spline crank to fit an untapered BB axle there has to be a small gap between them, so there will always be a small amount of movement between the parts. As a result, if you stand with the cranks level, then turn the cranks 180 degrees and stand again, you're guaranteed to get a small amount of movement between the untapered crank spline and the BB axle spline. This movement will chew material away over time (And feel terrible to the rider), leading to cranks that can never be fully tight, as well as premature loosening of the crank retaining bolts. Best case it'll be a similar level of wear to that caused by the constant small movements due to thread precession in pedal threads where they interface with cranks, but in my experience on badly fitting ISIS style splined BB's, it'll get a whole lot worse a whole lot quicker.
@@peglor Thanks for taking the time to reply. I reckon the small gap is not necessarily required. A small amount of interference fit would have solved this problem. This, plus the bolts should be sufficient, without over engineering it with a taper.
@@kev7355 When you fit the crank if there's interference in the joint, the axle spline will shave the interfering material off the crank to allow the parts to fit together. An assembly like this might work acceptably the first time it's put together, but will deteriorate fast once it's been rebuilt a few times. Also keep in mind that these are mass produced parts, so normal production tolerances will mean some will have clearance and some won't fit together at all unless there is an extreme level of quality control.
@@peglor You make very good points there. I understand about production tolerances, and ISO has interference fit standards that cover the tolerances. Machining of interference fitting machine parts is not something new, and crank is not something that is removed and reinstalled constantly. With all due respect, I still don't think its required, but the engineers who designed it and thought about the whole system ( manufacturing, costs, reliability, patents, etc ) thought differently. I'm sure they know something I don't, and that is fine.
I think the best/simplest design is clearance spline + pinch bolts, but shimanos patent covers that. Rotor did something similar a while ago with a single pinch bolt but maybe it infringed so we don't see it now. Personally I don't like the spline and taper. Its risky and hard to get right.
Is it possible to comment on the fit of the taper tolerance in mass production with a sample size of 1? I guess you could make some type of statistical assumption?
What should you do if you have some play when mounting the spider on the spindle. I have a aldhu24 crank for my BB90 Trek. And when mounting the spider it is not totally play free :(. I can remove the play when tighten the crank into frame but that can not be good, is it? Any one this issue? @PeakTorque?
Rotor Aldhu/Vegast 24 is a nightmare. I have two spindles with bad threads and during a casual ride a brand new Vegast crank arm fell off. It was one week old and I used a torque wrench. The quality is just horrible. Same for their web shop that sends mails with template placeholder and customer supper that promises replacement parts but never sends a return label, wants me to cut the crank in half. Do I own a machinery shop?! Ridiculous.
It's a good idea if yours is about to delaminate (if you have this problem, check the adhesive seams) or the spindle has shown considerable wear & tear. Also, if you want to play around with crankarm lengths, just buy the crankarms, no new crankset needed. This is what's good about a modular crankset module. A more affordable alternative is Croder's Spirit modular crankset.
Excellent video! The question is, as always…why is almost nobody making a 24mm steel spindle crankset? Especially with carbon cranks they all go for aluminum…
Aside from Shimano, FSA still offer exactly that. It's just hidden under their "MegaExo" moniker and it's not as apparent as it should be IMO. The advantage is that they'll work as a straight swap into Shimano bottom brackets or those sized for the same spindle.
@@TypeVertigo Not all of them though, some measure 24.07 mm and my FSA Gossamer's spindle is exactly at that diameter. I can't fit it inside an Enduro BB made for 24 mm spindles.
Update: I got the carbon arms a few days ago and the increase in stiffness is pretty dramatic. The carbon arms feel like a crankset should, without the noodly flex of the aluminum arms.
@@lastfm4477 I realize your trying to make a joke, but it's actually in high torque situations like starts and steep climbs, especially for a 200lb rider like myself
Link to the cranks and axle in description. Copped both for about 200 quid. Let me know if you understood the bit about the taper geometry. I kept hammering it home but don't know how obvious or not it is to people.
Edit: reading some of the comments, it seems this design isn't the most user friendly or clear to install. To help those with issues here are some vital points:
1. You MUST back off fully the preload collar before assembly otherwise you can deadlock the BB bearings.
2. If your BB shell (thus bearing stance) is slightly too wide, doing up the tapers to the full torque may compress the BB bearings and over preload (deadlock) them. In this case, remove the plastics 0.5mm shims and oring. This will give you up to 2mm more shaft to play with.
3. If the 35Nm wont fully lock the chainring spider or power meter, tell Rotor its out of tol. If they mess you around and chat bs, then gently heat the crank to 60c and reinstall with grease on the spline. This will slightly enlarge the female taper.
seems like the links arent working
@@danielhostetler2896 cheers should be working now
Clear explanation, even for a non engineer. Thanks for clearing these things up. Cheers
The illustration with the cups was great!
i am using rotor aldhus on 2 of my bikes nad lovem also with oval rings! nom :)
As a machinist who has in the past needed to nail axial positions of shafts relative to their tapered sections I can vouch for every statement you made about the production of a part like this.
Very tricky!
In fact I might wager that they may cut the tapered section in one operation then locate that part on that section to cut the corresponding shoulder. Even then, most lathes would be at or beyond their ability to position a cutter accurately enough to put that shoulder where it needs to be. If it’s a steel axle, I would be looking at cylindrically grinding that to size.
I too have swapped out my Shimano crank for a modular design but I went with Easton. Their solution to this is quite different in that the chainrings get splined and bolted to the crank arm before that whole sub assembly gets mated to the spindle.
Very clever and it works great despite its aluminum axle. (One piece Hambini bb ensures there isn’t undue wear from misalignment.)
Nice to get a machinists POV! I don't envy those who have to make tapers to axial length specs. What you said about mating them and then cutting the shoulder would work, but then they'd be matched pairs (crank and axles). The fact that they're not matched serial numbers kind of makes it impressive i think.
Except he is mistaken about how the rotor is produced. This will be machined on a Muti-axis CNC lathe with in one setup with additional milling axis. It is commonplace technique in precision machining.
You writing a book or want a metal? jeesh looooooooong response,
@@waterdeptworker manufacturing is complicated and nuanced. If reading isn’t your forte perhaps leave that to the big kids and fuck off.
@@waterdeptworker calm down jeff
I never thought of putting a pen mark on my male taper and then testing insertion depth, but this might be a good way of testing the dimensional tolerances.
There's a joke in there somewhere....
I did and the better half smacked me around the head. The one on my shoulders.🤪
@Luv23c With the difference heat-driven expansion makes, you'll want to gather a large n to find the true variance. Best get cracking.
Just goes to show, if people in the bike industry want to put in the effort to make a product with high tolerances and QC it is quite achievable.
I can’t help but be weary of Rotor cranks after the experience I had with the one that came on my Cervelo. It creaked like crazy and once I swapped to Shimano the creaking went away. 5 years later, still no creaks, BUT I’m on my second Ultegra crank. Pick your poison?
They will sell what is profitable. If high tolerances and qc components will make them rich then that's what they'll sell.
@@RicardoRocha-lg1xo that's the shit cervelo BB quality control, not the rotor crank
@@shamuslamont100 the BB was Rotor too, yet I haven’t had issues with Shimano or SRAM Dub BBs afterwards. I am currently running a SRAM GXP crank on that frame (long story, lots of crank swappin this past year) and the BB is from a brand I’d never heard of… yet it doesn’t creak. I guess the Cervélo reps I met at a race some years ago were telling the truth about that Rotor crankset from 2014-2016
0:31 the subtle pause is absolutely amazing
The more I learn about modern crank and BB standards, the more admiration I have for square taper.
Square tapers went out because indexed front derailleurs demanded chainrings remain in the same position relative to the frame after every bb overhaul, according to Shimano. How often do you need to overhaul the bb? couldn't you readjust the derailleur if something changed? isn't that exactly what we did over decades of shift-by-muscle memory and square taper bb's? OTOH external bb cups are cool, but they were a patch for the poor durability of small sealed bearings inside a BSC bb shell with the 22mm ISIS spindle. Then again a Phil Wood sealed bb lasts pretty much forever especially if you own more than one bike.
@@slowerandolder Front derailleur argument seems ridiculous. Good square tapers last 50k km.... smoothly. Surely adjusting a front derailleur isnt a big deal at that stage. The main problem seems to be that this solution is somewhat heavier and indeed like the ISIS standard shows very difficult to oversize to get to a lower weight. Not sure if its less stiff too, but track sprinters still use dura ace track cranks, so probably its a non issue too.
@@ohne_speed that's interesting - in the 1970's, as alloy cranksets became common on cheaper bikes, shops occasionally saw wallowed out left arms; customers would blowoff the free checkup offered on new bikes, the left arm would loosen a bit, the oblivious owner wouldn't notice until the arm's tapered hole was deformed = new crankarm. Shops had been torquing that left arm's retaining bolt properly, but properly also required re-torquing after 50 miles or so. Once this retail reality was recognized, the shops started doing up the bolts as hard as possible - which shortened the number of times the crank could be reassembled (after bb overhauls) but no one ever complained about that. This was never a problem for driveside crankarms, since that pedalling load went straight onto the chain.
I destroyed 3 square axles riding MTB. I'm glad that standard is long gone.
@@ohne_speed Square tapers are still used on Olympic level track bikes, so I doubt it was your 1400W that rounded the axle.
I have two ROTOR Cranks, one in 30 mm, one in 24 mm, both DM. They perform really well and I do not have any issues with them. They are not cheap, because they are high grade products. What I like about ROTOR cranksets is, the seem to be more quiet than Shimano ones I had previously (R8000). As I am not an expert, I think, less noise, less friction, better quality.
It could be that the tolerances don't need to be quite as tight as you think. There is a phenomena known as elastic averaging for coupling designs like this. Your analysis is correct if the material is assumed to be perfectly rigid. The shoulder sets the axial position, and the rest of the material will elastically deform to allow this to occur (which is why you need to use the force of the screw to draw the assembly together). To extend your comparison of machine tool tapers a little further - dual contact and HSK interfaces have a shoulder perpendicular to the axis of the spindle such that an accurate z-location is set regardless of differences in the tapers. The requirement for this is important as rpm of the spindle increases as the hoop strength of the outer "taper" (spindle) can actually open up and the draw bar can pull the tool holder further in to the spindle. Heat is also something of concern, which you note. I haven't gotten my hands on one of those Rotor shafts, but I would imagine that they are machined with a swarf style 5-axis strategy in a mill turn machine. In cycle probing and cutter compensation would make this a somewhat straight forward task with a straight fluted endmill (no custom tooling required),
Nice comment. Someone out there knows their stuff! With relatively thin wall tapers yes they can elastically deform, but the tol will effect the amount of force required to hit the axial end stop. It was my belief that tool tapers being virtually solid there was barely any elastic give in them, which would mean an taking an accurate z cal every operation. But my machine tool knowledge is limited!
@@PeakTorque The tool holder itself might not "crush" inwards much, but the machine spindle is a hollow cylinder at some level of analysis. Another way to look at it: make the assumption that all the torque applied to your bolt is being converted to linear force at the pitch of the thread (no friction losses) if that force is translated to the 1 degree taper (on the coupling) that is a hell of a lot of expanding force! Wedges are really good at expanding things. The HSK spindle interface actually clamps the inside of the tool holder for this reason - as RPM increases the clamping force expands outwards. For your cranks: you could check this relatively easily on the crank arms by mic-ing the "width" across the interface area in a relaxed state and compare when it is assembled and torqued. Assuming you can repeat your measurement location accurately, I would imagine you see a difference (cross section growing when it is assembled). My point is simply that the torque spec on the bolt may be the "dominating" variable so long as the components are machined with relatively good tolerances. IE the tolerance window on the machining is a little wider than pure geometry / trig allows for, because nailing those tolerances would be next to impossible - not to mention the varying thermal expansions of the discrepant materials. Black anodized alum might get hot enough in the sun to fall off ;)
Yes, good comment. After reading this and a few other comments saying similar (deformation, elestic averaging etc., Maybe this design isn't TOO bad.)
This business with the tapers is quite bad on SRAM GXP and others. It messes up the chainlines and you end up having to mess around with spacers. The "average" bike mechanic can't get their head around it.
Since this design locates the spider on the machined shoulder (which sits on the drive side bearing) wouldn't the chainline be exactly determined?
I mean to the extent that the frame is manufactured precisely anyway.
I don't think that applies to this design. Both of the arms bolt down to the axle, and then the preload collar is expanded to tension the bearings. The only need for spacers is for different BB types.
@@galenkehler ON the GXP, the non drive side bearing is fixed so it determines the axial position. If the taper nips up more or less then it moves the effective chainline. SRAM obviously know about this because their solution is to add more grease and clamp it up again.
@@Hambini for sure, I'll never defend GXP 😁
@@Hambini I'm just a bike mechanic so apologies for my lack of engineering intellect but are you referring to DUB, not GXP? I only say this as with GXP on the non drive side the axle butts up to the non drive side bearing. The NDS crank arm then compresses to the opposing side of the axel on NDS bearing. The chain-line is determined by the location of the NDS bearing and the only way to manipulate the chain-line is by moving the NDS on the BB shell left or right. With dub you have a series of spacers which you can replace etc to manipulate your chain line should wish to stray from Srams instructions.
Back in the Isis bottom bracket spindle days, Race Face cranks specced a particular torque but also said it had to bottom out. Their instructions said if the arm doesn't bottom out at that torque, to remove the crank and reinstall however many times until you did.
I've installed several Aldhu cranks. Sometimes I need to remove and reinstall once to ensure it's bottomed out.
I had an Isis BB and truvativ cranks. The tapers never mated, and always came loose and ruined themselves. They never worked well. These spline tapers are risky!
That's interesting. Basically they were factoring a bedding in it seems. Im curious to see if a wave washer could be utilised to add some preload to allow for any tolerance variables or fluctuations. The design should allow for and bedding in of the taper.
I have 2 of these both linked to powermeters- one on a power2max and one on an xcadey xpower-s.
I haven't seen any problems so far and both read consistent with a quarq riken spider, and all 3 powermeters are within 5% of a elite direto trainer. Happy with the performance.
It looks like this design has even more in common with plastic cups then with a lathe rear tool post mount: cranks must be undersized and considerably soft to press onto the axle with plastic deformation
I 100% agree. There is no way to get a proper contact and fit on both interfaces without one of them deforming. If it works it is still quite impressive.
I love the power meter setup on these, and they have used the same interface for quite a long time so there are actually plenty good used power2max meters available online. You can also play with crank lengths quite easily. After spending years riding BMX where this kind of design is the norm, I have always found two piece cranks quite annoying in comparison considering how the whole thing needs to change if you change bike or damage something.
Historically road and MTB cranks were 3 piece setups back when most of BMX was running one piece cranks and oversize BB's with press fit cones. Then BMX moved to 1.37" threaded BB's which 1 piece cranks couldn't be inserted through, so they moved to 3 piece cranks. /In the meantime road and MTB moved to 2 piece cranks and external BB bearings. I expect it's more the case that BMX hasn't moved to 2 piece yet rather than that 3 piece is the end of their evolution.
Sadly the bike industry has never cared whether their designs mean dramatically increased expense when part of a one-piece assembly gets damaged and you're forced to scrap the whole lot. It is possible to buy individual left cranks from Shimano, which came in very handy a few years ago when the pedal threads failed. The UK importer of Shimano doesn't do any of this though, nor do they even admit it's an option as it's more work for them for less profit.
Just go to Shimano's own website, look up the exploded diagram of the crankset to get the part number (Usually the PDF you download from Shimano with the diagram is actually an order form because Shimano expect you to be able to order individual parts), and google the part number to find sellers - There are some decent shops in Germany that support these orders, bicikli.de/shop/ is the best one I've used for having Shimano parts in stock - they're excellent for things like freehub bodies and ball bearings for hubs too.
It is worth pointing out that if you had any play on the powermeter spider it would not all have beeb Rotors fault. Its a 3rd party power meter so if that would be out of spec then Rotor could have got is 100% right but not get it clamped down because of it.
The dimensional tolerance of the power meter in this application is simply the width of the splined portion. Getting this right is trival compared with the matched taper.
@@galenkehler Yes you are totaly right about that. They still have to make it to those specifications. I am not saying its right or wrong it simply not gets mentioned.
I have the previous version of ALDHU labeled 3D24 I think. I like them. Durable and nothing really to go wrong. They replaced shimano 105 cranks.
Hey, first of all, thanks for the video.
I recently bought a Xcadey powermeter (PM from now on) for Rotor aldhu cranks. As you said, the interface was not perfect between the powermeter, the cranks and the axle, so there was an unnaceptable amount of play. I took it to my local bike shop, thinking that maybe I did not know how to assemble it correctly, and they returned it to me assembled, and said that more force was needed in order to get a snug fit.
Weeks go by and the PM starts returning some wonky readings until it eventually dies completely. After dismounting all the pieces I discover that the PM outer cover is cracked, undoubtedly because my LBS applied too much torque.
I have to add that Xcadey sent a new one, almost no questions asked. This second unit I mounted myself (not going back to that shop) and had no troubles, in fact it´s still going on trouble free after 6 months.
I guess that the problem lied within the first PM mounting interface, since that´s the only thing that changed, but that´s my expierence.
I had more than one problem with these:
1. Two sets produced vastly different chainlines and Crank q on the same bike, also cranks were not centered in reference to the frame.
2. The crank bolts loosen over time, and there is no fail safe that prevents the cranks from coming off (did not happen to me, but it happens acording to the forums)
3. less space between the crank arms and pedal than with Shimano, Quarq (relevant for me during winter when wearing overshoes)
4. Preload ring is an absolute fail! The ring gets compressed when turned. It's really tricky to dial in... need to place a flathead where the ring opens (which does chip the material and introduces splinters into the fine threads. 2 cents of thinking would have at least yielded to put some knurling on the opposing side of the preload ring pinch bolt.
5. Rotor support just stopped responding when asking for support with this.
So in summary: I can't recommend those. Really wanted to like them, because they were the only option to stay with a 24mm spindle and using a power2max.
Really wish there was a better option for Shimano BBs. Imho they should keep the modular design, but
a) change the spider interface
b) use a similar design to Shimanos for attaching the cranks
I have two of these and don't see the problem with the preload ring, so long as you aren't ham fisted!
I have issues with my rotor chainring not sitting parallel with the axle. Very annoying considering the ring cost €90.
I`ve always had issues with preload on these. 35nm to attach the non drive side arm seems way too much, they clamp so tight there`s no preload to take out with the ring and I`ve tried the same chainset on 3 bikes all with the same outcome. At 20nm it`s usable but I`m worried it`ll come undone. I`m now convinced mine is faulty in some way,
I've had all these issues too over c. 10,000 miles of use. On mine the power meter spider (rotor PM) sits dead square but over time and usually after a few big power efforts the spider twists on the spline. I have to un tighten the bolt and reset and tighten up again. It's currently so twisted Di2 can't handle it with auto trimming on the front mech. It's particularly bad when the bike is in a direct drive turbo.
@@ClayDavies how does the spider twist on the axle? That's sort of physically impossible unless your spline teeth are gone.
I have one on a P2m spider and one on an Xcadey spider, no such issues so far.
Thanks so much for doing this and confirming that Rotor makes quality stuff! I'm seriously considering Rotor cranks with Power2Max power meters and don't wish to risk having a damaged crank with Ultegra or Dura Ace!
Great video. My big 1x11 Vegast chain ring came lose on a bike trip. I never fastened it with a torque wrench. So now I understand why it came lose. I found a guy with a big enough Allen wrench and tightened it very well. Back home it came lose with a lot of force and squeeking sound. Now I know that the tapered side caused this. I will use a torque wrench next time, this is precision work!
A fantastic modular system I run myself on the TT bike, with interchangeable spiders/PMs/chain rings. You can get the Vegast crank arms for about half the cost of the Aldhu's too.
Yup. I'm running Vegast cranks on the 24mm steel axle with a 30/46 direct mount one piece chainring (with a Hambini PF30 bottom bracket). Works perfectly on my otherwise Ultegra di2 setup.
I've been seriously considering the Aldhu for some time. Thank you very much for making this video. Cheers!
I am on my third Rotor crankset. On the first one the left arm came loose and destroyed itself. The second one was DOA. The axle had terrible burrs and so the surface on both arms was destroyed. The third one is working so far.
That does not sound good!
@@PeakTorque Most of these issues are user error. Insufficient torque or not grease on the mating surface, poor adjustment of the preload. The issue with the alloy axles destroying themselves is usually just bad bottom bracket alignment or poor preload. But sure to check the bolt after the first few rides which is good for peace of mind.
Aldhu 24 + Power2Max NGeco + Praxis Buzz 10/11/12-speed chainrings + AXS road 2x was a nightmare.
1) The cranks hit the derailleur cage when set up to SRAM’s spec.
2) The 5mm wide flat-top chain is can get stuck in between the chainrings because the NGeco mounting tabs are >3.5mm thick.
3) That >3.5mm thickness causes shifting problems because the FD overshifts to a lesser extent while the rear is in the smaller sprockets.
I'd like to see you analyse the carbon version (unfortunately only 30mm axle).
Long-term review would be interesting, i.e. if you’d have any play eventually e.g. from wear. It’s of course normal to torque down chainring bolts time to time, but here it’s quite an exquisit setup.
My daughter has been using these cranks in 30mm version for XC training and racing (high level I'd say, currently 2nd in Poland), including doing some tricks and jumps. After half a year zero problems and it's been offroad riding in technically difficult terrain. In my opinion these cranks are on par with shimano DA or XTR as reliability goes. And the support from Rotor is fantastic. I chose road cranks for mtb because of the need for low q factor. Combined arms with Rotor's XC direct mount XC chainring. Asked Rotor if it would work and they even calculated chainline for me.
I have the 24mm model with Power2Max power meter and also the 30mm track model. Both are very stiff and superbly made.
Speaking of tapers, what is your experience with square taper bottom brackets, and do you have a clear favourite among them? Always love these engineering analyses.
: I, for one, think that for the time they were good and they lasted a long time(speaking for Dura-Ace but I think a lot of them did) but Octa-Link was the superior way.
@@philippbeckonert1678 still using my octalink xtr cranks with dura ace BB
24mm axle works only with specific 24mm cranks and 30mm with 30mm cranks and afaik carbon one is only 30mm last time i checked (have both of them 24 and 30mm axles)
Whenever I've looked at Rotor cranksets, I've been shocked at the prices, so that has ruled them out for me. I have wondered about the machining of the tapered splines in cranksets that I own (FSA and SRAM). There's quite a few standards for straight cylindrical involute splines, but I don't know about for tapered splines. The exposed spindle length a.k.a. distance between crank arms (as Hambini calls it) can vary quite a bit due to manufacturing variations.
Yeah, it’s impressive tolerance but, but also kind of demonstrates how the sram or race face style chainring attachment are better. With repeated assembly and wear, the rotor style is likely to reach a point where the chainring isn’t clamped tight and a shim would be needed.
which cranks can i use with sram red etap chainring dub system? would like to go for shorter than 170mm cranks... cant find sram cranks anywhere shorter than that
Very interesting and I always learn something from your reviews (as a total layman engineering wise). Personally I have two bikes with Rotor InPower 1x cranks and they are bulletproof (aside from the single A battery which is relatively short lived) and easy to swap chainring sizes. . Quality machining and really not that expensive especially when they have them on sale. By far my favorite cranks.
My takeaway - apply boiling water to male parts for a more impressive size.
I don't deny that that is a very nice piece if machining but modern CNC machine tools should easily be able to hold 10micron tolerance on those size parts so the 50micron you use as an example is a bit excessive. Also - there is always an also - the simple trigonometric analysis does not account for the elasticity of the crank spline hole (and I suspect a lesser extent the axle) which will allow axial movement when tightening the bolt. I have not run the analysis but on my crank-set (see the first link below) once the taper is 'nipped-up', tightening the retaining bolt can push the crank about 1mm onto the taper on the axle. Also, you should have considered the thickness tolerance on the power meter spline area which is also important for the fit to be good enough - how well so Sigeyi control this? I suspect the tolerances on the tapers are not as critical as your analysis might suggest.
I bought this Sigeyi power meter about one year ago (th-cam.com/video/wXIaALuMPCg/w-d-xo.html), I have not been able to test its accuracy (its not that important to me) but so far it has worked reliably and the battery lasts a long time. The only real issue I have had is getting a firmware update transferred to the power meter. I'm looking forward to your full review of the power meter.
OK, I've been a bit critical. If you want to get your own back, I'm look to get engineers to critique is video (th-cam.com/video/Q4HGhrBXDIM/w-d-xo.html). In particular comment on what I found in the seat-pin would be welcomed.
Great comment. Yeh i didn't go on deformation of the parts, thought the video was nerdy and long enough for most. Always a balance between detail and retention to satisfy you lot!
Didn't mention the spider thickness control, that should be very easy to control, but yes its important.
I just chose 50 microns as an extreme figure. But even with 10 microns because of the tangent relationship its still quite a large effect on axial position on a fine taper.
Cheers
@@PeakTorque I don't know about anybody else but I come to Peak Torque because it's nurdy. The more nerdiness the better in my opinion.
Terry
Been using Power2Max with the 24mm Aldhu for 2 years with no issue. Just about to fit 30mm aldhu+p2m combo on a new bike.
there was a reason we abandoned tapered axle and cranks long ago. I used them through the 70's and 80s. . Shimano engineered the new axle design - and we have no problems.
Hey PT! Many thanks for the video. I had Vegast version if the crank (cheaper version) and I had this very problem. Crank arm wasn’t pressing all the way to the spider and had this terrible creak going on. I contacted Rotor and they eventually replaced it-they tried to convince me there was some bb/bearing misalignment but I did try other 30mm crank which worked fine. In the end, I sold it as I wasn’t keen anymore.
It seems making this video and the comments has brought into light the issue i was wondering about before i purchased them!
Funny that you say this. Just got an Aldhu 24mm and no matter what I do, the spider is loose. I also tried googling this but didn't find anything. I guess I'm the only one with this issue... I called Rotor, they told me to go to 40Nm and it still wouldn't compress. I can't seem to win with any brand.
Thanks very much for this comment. This was exactly the problem I had envisaged! So there are problems with QC then. Try this: grease the spline, heat the female taper (the crank) in an oven to about 80c, then install. Should get you a bit further up the spline.
Did you check the inside of the crankarm where it mates with the spindle? You may have squashed the washer that's inside there.
This happened to me when I was installing one, supposedly it somehow ran off-centre at some point.
A similar review of SRAM’s cranks would be a good video idea. They have a much higher torque spec of 54Nm. I think they may have more issues with cranks coming loose despite the higher torque spec.
So far, I've never seen one come loose, but the ones that require a breaker bar to remove, now that I've seen a handful.
@@yonglingng5640 I’ve had one come loose. They also have an FAQ on their website stating if it comes loose and you continue to pedal it can deform the splines and you’ll never get it to not come loose again.
@@ProffessorSeen1 Even when torqued that high?
second
Always loved Rotor stuff. The 3DF made the BB30 on my old bike actually usable. SRAM DUB looks like a not-so-cheap knockoff of 30mm Rotor down to the preload adjuster. Difference is the Rotor one is beautifully machined whereas the SRAM one is a piece of plastic tat.
To add to this, doing a build for a customer have convinced them to go for a Rotor Vegast rather than Ultegra crankset (in part due to availability it must be said).
Not to disagree on the general importance of getting tolerances right in over-constrained interfaces like this one. But even if the male part was just ever-so-slightly too big (or the female part too tight), you could probably still torque it down flush against the lateral stop; the crank-arm portion of the interface would likely just stretch a little more. With my set (30 mm), I can kind of tell as the preload pinch nut, while being very (like, very!) loose on the pinch bolt before torquing down the left* crank arm, will seize considerable after doing so. Also, the 3 plugs in each arm will usually become clearly visible after torquing down the cranks. And stretching holes (of cranks, way beefier around the hole than the rotors) has been a thing for decades ... well, ok, this might be easier with a tiny square taper compared to a huge spline taper, but still(?)
* I was told by some Rotor person that the left arm, pretty much the same as the right one, bottoms out against a lateral surface when mounted as intended.
Side note: Sram has been mass producing that kind of over-constrained interface (tapered spline + lateral stop) since the introduction of gxp. Now, by no means, am I trying to say they nailed it or that there weren't people having issues with it in the field, just that, by and large, it kind of can work.
Another side note: I still find it somewhat funny how we went from 3-piece cranks, some designs with detachable spider, to integrated/2-piece to, now, fully modular (kind of 3 piece again).
(Sorry for shitty English. Also, not really an engineer. I like your videos much more than the ones from the French aerospace guy :D)
Good review and well explained. I have the same combination on my bike and will check it out later. My question: If the chainring is lose like you described at 8:47 and can angle would that not result in a better more straight chain line?
Maybe, but wouldn't that cause power drift?
Can't wait for the full review... Been looking at switching!
The Aluminum Aldhu turned out to be to flexible for me, the pedals end up at a jaunty angle under the force of standing pedaling. The new carbon ones are supposed to be better but haven't tried those yet.
Rotor are bad at q-factor. On my carbon road bike there is >1cm on EACH side between the arm and the frame. They could easily have narrowed the q factor with straighter arms and increased stiffness at the same time.
Rotor chainrings are really noodley, so not enormously surprised their cranks lack stiffness.
I had the center bolt that affixes the NDS crank shear off inside the axle on a ride. Rotor attributed it to me not installing it properly (I had put a few 1000mi on it since the last time I had taken anything apart) but ended up sending me a full set of new arms and axle. I had only needed the center bolt, but oh well. Otherwise I've been happy with these married to a power2max spider and praxis chainrings.
The old rotor setup I had prior to this (the 3d+) had a catastrophic failure at the pedal spindle interface and I broke the whole end off the crank arm.
They use this fine spline for spider mount interface just to allow 1 degree clock of direct mount oval chain ring. Want to dial oval chainring orientation a bit? No problem, take it out and reinstall it rotated one degree clockwise or counter clockwise at a time until you find the perfect angle.
As we use round rings, this design complexity give us no benefit.
Exactly the reason why they did the design this way.
I had to torque more than 35nm to remove the play and still have creaking with 24mm and Hambini BB. Doing my head in. Think need to speak to Rotor as Power2Max were responsive but not helpful.
Hello, I have this wobble too. I screwed the crank, the shaft and the Power2mx together, but there is 1-2mm play and the crank wobbles, what could that be?? In your video you can see it at 8:45, thanks for help
Nice explanation! I recently bought a power2max NCeco with the aldhu 24mm for my 18 scott foil (PF86 BB) . My drivetrain is a r8050 ultegra di2 so I kept the shimano chainrings. However, when I first installed the aldhu crankset, I noticed the crankset is sitting slightly further than my ultegra crankset. I noticed that because the chain is rubbing the front mech when I am on the big chainring. So I thought no big deal, I will just adjust my FD. So I got it to work, but since then, my drivetrain is super noisy when in the big chainring (worst when cross chaining). It's been a few months and still the same. Any ideas? Appreciate for your input!
I think Rotor headquarters are in a little village outside Madrid(Ajalvir). They build their stuff at home. I had a few Rotor products in my hand,very good looking but never try one.
@Peak Torque I have just purchase at 24mm Rotor crank set and 1-piece chain ring. The machining on the chain ring is outstanding!
When you torqued up the arm, did you apply force to the pedal with a chain attached to mate the splines in the best position as there's a fair bit of lash between the ring and spindle.
Thanks!
My Rotors click (30mm DM version) and haven't been able to resolve it until this video. I've had them for over two years now but it was already clicking under each (driveside) pedal stroke after a few months from new so I suspect the DM and crank interface developed some give resulting in the click.
Oddly my older Rotor 3D+ cranks (24mm) on another bike have been silent. Both of my Rotors have Q Rings if that makes a difference?
Thanks for the video. I was recently looking at getting a power2max power meter with the aldhu cranks, so I imagine this video will be useful.
I have noticed that you went back to quarq on gxp I suppose, any reason why you chose gxp over 24mm aldhu. I m currently choosing my crank set up and can't get passed by disadvantage of gxp
Personally had trouble with the taper, vegast crank, had to go above the 40nm to get it the lock correctly the chainring
At first look, it looks the same as XCadey, but they're different aren't they? I think the XCadey power meter is mounted with 8 screws.
@peak torque just seeing how your going with these cranks?? I'm building up a new bike and leaning towards these cranks. Do you have any feedback please after 6months. Cheers
@Peak Torque how much clearance have you got between the drive side crank arm and the outer derailleur cage?
I've got P2M +3d24 and there isn't enough clearance. If this works well I might go down this route instead.
Are you using Mechanical or Di2?
Thank you. GP
i have a Aldhu 24mm and even when torqued to way above the torque setting the chainring still wobbles. tried the chainring on my aldhu 30mm crankset and it fits perfectly. i might need to get a spacer to make the crank arm compress the chainring enough to keep it tight.
I would imagine that the elasticity of the crank material at the hole plays an important role in this fit too. the relative fine taper can be an advantage then
Very true
I have those on my TCR and 8-Bar fixed gear. They are lovely cranks.
Sugino offers Shimano compatible 24mm road cranks, though I have no idea about availability where you live(or where I live for that matter)
hey, nice video. Any update on it after year of use? thinking of swapping to it
Is the left crank axially restrained on the face (similar to the drive side) using the bearings (I'm thinking with a spacer keeping the inner races apart) or is the tapered spline the only thing holding it in place? EDIT: It doesn't seem to be this way looking at it. But does the end of the spindle bottom out against a face on the inside of the splined crank? I'm worried about the splined fit ''squirming'' under load if you will, and fretting wear occurirng between the crank and spindle due to elastic backlash under load.
I wonder, is the taper issue described the reason for the failure rate I've seen (2 sets of my own) in the rotor flow cranks. Cracks in the crank arms at the spindle interface.
I've got a guess here, if the dub spline's got a taper of 2° and preload of 54Nm, and the rotor one has a 3° taper, it'll only need about 35-36Nm to achieve the same amount of radial preload, which is what is recommended.
I've also got another question, why none of these manufactures make the spline out of tangent arcs, instead they use this "small round cutouts in a big circle"? (hope I made myself clear)
@@endercrafts9056 it doesnt work like that. In theory the shallower the taper, the less clamping force you need. Larger angle tapers actually need more force! As you have less mechanical advantage in a broader wedge. Second question: probably because they cant reliably rely on the locked taper to transfer pedalling torque so they use a very simple spline shape (easy to program) on cnc.
what kind of big torque wrench do you use? Looking for a reasonably priced one. Thank you!
I'm hearing a lot of good things about Praxis cranks. I'm hoping you get a chance to review that.
if they made 24mm spindle~
Not a good spindle design.
30mm drive side, 28mm NDS. Basically, oversized GXP design.
Great style of explanation of the important engineering details. Subscribed. I saw this vid as I was looking at these crank sets for a power meter. I really don't like the idea of taper and spider clamping combined as you describe... Even if they do manage to get it to work in this case. I'll probably look for other brands. Is this the only option for these style power meters for shimano rings?
Ive not long but the cheaper version for my 2007 cannondale Caad 9 and what a great idea, rap going forward as can fit most bbs and if you buy the spider any chainrings!
My challenge in clanging my chainset to rotor was in the quality or compatibility of the chainrings. I will not do that on my new bike. I will opt for a 4iiii powermeter rather than rotor
@Peak Torque, I went with a 24mm 3-bolt setup, but full Chinese - Sigeyi AXO with Zrace Hardrock cranks. I'm preparing to install and have found there is a bit more backlash than your splined Rotor setup. How concerning is backlash with these power monitors' accuracy? Is the issue that some level of accuracy is lost for a moment in each pedal stroke?
It won't be good for the zero offset and repeatability. You'll also get a lot of noise and lack of stiffness from the chainrings. Do the cranks up to reccomended torque and then measure the gap. Use thin shims to fill the gap.
Hi!! Did you used all the washers and o-rings that are shown in the video when installing the crankset in the Giant Tcr BB86?? I’m asking because I bought a set of Aldhu’s second hand and thet came without any washer and o-rings, to know what to buy to install them properlly in my Giant Tcr.
I actually think i didn't use any orings or the washers, because the tapers were squeezing these too tight when everything was clamped up. So just left them off, which is counter to what the instructions say
If I'd ever tried to put something like this design into production it wouldn't have got past the first design review. It's just dumb engineering design requiring stupid levels of machining accuracy to dig themselves out of a hole created by a fundamentally stupid concept.
SRAM solved it with 3 screws, race face/shimano with a threaded lock ring, neither of which require special tolerances. The main reason I can think to do it this way is if its a patent issue and they don't want to or can't pay the licensing fees.
Yep I agree. Its overcomplicated but Rotor seem to like doing that. For me the best way is clearance spline plus pinch bolts but that's patented, somewhat ridiculous, but that's how it is.
@@PeakTorque I guess it could have been worse, they could have gone for a hirth coupling.
They don't make carbon with 24mm spindle?
Has there ever been a crankset with a threaded interface to secure the crankarm to the spindle? (As in you turn the whole crank arm to affix the crank arm.)
No. Primarily because you don't want the crankarm to unthread as you pedal.
Pretty hard to get that right too. I mean your crankarms have to end up at 180 degrees at specified torque.
Does rotor crank arm compatible with the sram red direct mount chainring?
Have an ALDHU and it isn't a straight swap with my Ultegra... chain-line is vastly different so I need to adjust the front derailleur always by a lot when switching
Did you try it without the plastic shims they suggested? If you do that, the chainline for me was the same.
@@PeakTorque Yup no plastic shims, that's interesting to hear then!
Great video, good explanation about the taper!
Got it 35Nm but no mention of how you got the cut on your thumb? 😁
Would like to hear your thoughts on the Cannondale si series cranks. Same concept, although the spider is screwed down onto the arm, which is then bolted on with the aforementioned taper
In that case, you always know the chainrings are held on nice and stiff independently of the crank which is actually in my mind a better solution. Cannondale cranks don't seem to fail, even though theyre hollow (like shimano)
@@PeakTorque the lightest ones sisl and sisl2 fail too . Quite often for a small production. Debonding issues as well .
That’s good. I’ve been considering these cranks for more 1x chainring options
I did exactly this. I got the power2max track powermeter on the rotor aldhu, and it gives a perfectly centered chainline (42mm)on my disc brake road bike (142mm rear axle)
What is the order of the o-rings and spacers? Should the spacer touch the bottom bracket or the crank on both sides? Trying to find the same with the o-ring on where it should be located. I'm trying to figure that out for the install and can't find a good guide. Thanks!
if i remember correctly i did not use the oring or spacers as it made the whole lot too pinched when clamped up! And added too much preload to the shimano bb. That is the risk with tapered splines grrrr!
@@PeakTorque Great. Thanks for the tip. I'll be aware of that issue and test it out.
Hey, i just watched this and this product is awesome looking
One thing that bothers me is the use of o-rings on the axle. O-rings eventually dry-rot and fall off/out of things like this. Usually when you still have sixty miles to go before reaching civilization. Also, they crush during assembly and use, then when you go to replace the bearings or swap the crank to a different bike, they don't fit properly anymore.
I'm curious, is that taper absolutely necessary? A non tapered spline interface would have worked too, unless you are concerned that too much clamping force is applied to the chainring spider at 35Nm make up torque.
For an untapered spline crank to fit an untapered BB axle there has to be a small gap between them, so there will always be a small amount of movement between the parts. As a result, if you stand with the cranks level, then turn the cranks 180 degrees and stand again, you're guaranteed to get a small amount of movement between the untapered crank spline and the BB axle spline.
This movement will chew material away over time (And feel terrible to the rider), leading to cranks that can never be fully tight, as well as premature loosening of the crank retaining bolts. Best case it'll be a similar level of wear to that caused by the constant small movements due to thread precession in pedal threads where they interface with cranks, but in my experience on badly fitting ISIS style splined BB's, it'll get a whole lot worse a whole lot quicker.
@@peglor Thanks for taking the time to reply. I reckon the small gap is not necessarily required. A small amount of interference fit would have solved this problem. This, plus the bolts should be sufficient, without over engineering it with a taper.
@@kev7355 When you fit the crank if there's interference in the joint, the axle spline will shave the interfering material off the crank to allow the parts to fit together. An assembly like this might work acceptably the first time it's put together, but will deteriorate fast once it's been rebuilt a few times.
Also keep in mind that these are mass produced parts, so normal production tolerances will mean some will have clearance and some won't fit together at all unless there is an extreme level of quality control.
@@peglor You make very good points there. I understand about production tolerances, and ISO has interference fit standards that cover the tolerances. Machining of interference fitting machine parts is not something new, and crank is not something that is removed and reinstalled constantly.
With all due respect, I still don't think its required, but the engineers who designed it and thought about the whole system ( manufacturing, costs, reliability, patents, etc ) thought differently. I'm sure they know something I don't, and that is fine.
I think the best/simplest design is clearance spline + pinch bolts, but shimanos patent covers that. Rotor did something similar a while ago with a single pinch bolt but maybe it infringed so we don't see it now. Personally I don't like the spline and taper. Its risky and hard to get right.
Is it possible to comment on the fit of the taper tolerance in mass production with a sample size of 1? I guess you could make some type of statistical assumption?
No, i don't know how it is. Time will tell. It does seem like a risky design and a bit over complex.
Thanks. I learnt a lot from that one.
What should you do if you have some play when mounting the spider on the spindle. I have a aldhu24 crank for my BB90 Trek. And when mounting the spider it is not totally play free :(. I can remove the play when tighten the crank into frame but that can not be good, is it? Any one this issue? @PeakTorque?
If I ever have a Shimano crank fail, I'll get these. #prawnwatts
Rotor Aldhu/Vegast 24 is a nightmare. I have two spindles with bad threads and during a casual ride a brand new Vegast crank arm fell off. It was one week old and I used a torque wrench. The quality is just horrible. Same for their web shop that sends mails with template placeholder and customer supper that promises replacement parts but never sends a return label, wants me to cut the crank in half. Do I own a machinery shop?! Ridiculous.
What is the answer to the question? Should I replace Shimano cranks and, if so, why?
It's a good idea if yours is about to delaminate (if you have this problem, check the adhesive seams) or the spindle has shown considerable wear & tear.
Also, if you want to play around with crankarm lengths, just buy the crankarms, no new crankset needed. This is what's good about a modular crankset module.
A more affordable alternative is Croder's Spirit modular crankset.
@@yonglingng5640 thanks. My cranks have no signs of delamination and the axles look in good condition.
@@enigma1000 You're welcome. Switching to Rotor ALDHU isn't all that necessary unless you feel like shaving some weight and giving it modularity.
Is rotor any good then??
After watching your video I felt the need to yank my crank.
Excellent video! The question is, as always…why is almost nobody making a 24mm steel spindle crankset? Especially with carbon cranks they all go for aluminum…
because numbers sell...
Aside from Shimano, FSA still offer exactly that. It's just hidden under their "MegaExo" moniker and it's not as apparent as it should be IMO. The advantage is that they'll work as a straight swap into Shimano bottom brackets or those sized for the same spindle.
@@TypeVertigo Not all of them though, some measure 24.07 mm and my FSA Gossamer's spindle is exactly at that diameter. I can't fit it inside an Enduro BB made for 24 mm spindles.
Update: I got the carbon arms a few days ago and the increase in stiffness is pretty dramatic. The carbon arms feel like a crankset should, without the noodly flex of the aluminum arms.
Yeah, right. You put out 2,500 Watts and flex alumimum.
@@lastfm4477 I realize your trying to make a joke, but it's actually in high torque situations like starts and steep climbs, especially for a 200lb rider like myself
Hi Peak, Only cranks I have run on road bikes for years. :)
At 35nm on an aluminium axle and arm I feel like you'd bending things? Kinda like an olive in a hydro cable? (Says the non engineer) .
Its a steel axle on the 24mm.
Always replace a Shimano part. Replace it with some Campagnolo instead! Preferably replace all Shimano for campy:) Problem solved!