Does Pedal Kickback Actually Happen? I Use Data Acquisition to Find Out!

[@piotrekh7606]

Amazing! Please do more analyses like this! How many Points of Engagement does your hub have? Are you riding an oval or round sprocket?

[@daffydaz3329]

This is incredibly insightful, please do more! Most riders have no way of knowing if any of these terms are a real thing, and its great to be informed outside of the usual marketing hype. Thank you so much.

[@BRINKUK]

Such a lot of work went into this dude! Very impressed! Get an O-chain on there and lets see the data!!

[@clipless03]

If I'm reading the plots correctly, pedal kickback occurs when the angular velocity of the cranks goes above the angular velocity of the rear wheel. It would be interesting to see how those plots change if you removed your chain. Right now the spikes aren't that much higher than the rear wheel speed, but I think that's because the impedance of your weight on the pedals limits how quickly the shock can compress, and since you pedal kickback calculations depend on shock velocity, that will limit how high that peak can go. I bet if you took off your chain, you'd see the calculated pedal kickback angular velocities go up quite a bit. That might be the most telling bit of data. Thanks!

[@playgroundchooser]

First, this is without a doubt, the *best* video on this topic I've ever seen. I want a comment on here so when it blows up to a million views I can say I was there first. Yeah... I'm starting to think my entry level Shimano hub doesn't have enough engagement to really punish me with kickback.

[@mixmastamikal]

Would love to see the results when using a zero engagement hub like the Onyx. I think it would eliminate a lot of variability found with a lower engagement hub. Overall though I really appreciate the work you have put into this and that you made a video to share the results. I encourage you to please continue sharing your results. I am now subscribed and hope to see more.

[@steveedwards5280]

Really great to see someone actually applying some experimental method into this question. Most comments of this subject are either subjective riding or static maths. You mention to interesting points in your closing comments which both relate to how i ended up here. One is instant engagement (or the assumption of}. Im wondering if running an Onyx hub will lead to more experience of pedal kickback due to the instant engagement. Second, you mention the sound, which is interesting as i love the silence of the Onyx. I wonder ifthe silence and soft engagement might actually lead to less perceptible pedal kick back even though the zero degree engagement may lead to more events being transmitted to the pedals.

[@iamthecablecompany]

Neat project. Was wondering about this exact thing not too long ago. A couple of minor things. The upper right graph at 2:30 shows increasing kickback with lower rear sprocket size. Pretty sure that this is the opposite of what it should be. Probably just a ratio error that would be canceled out in the final results. Also I'm pretty sure the DB2 term would already be included in the recorded wheel speed. As it represents the slowing of the rear wheel due to wheel base growth. So by measuring angular velocity of the wheel instead of the forward of the rider you have already accounted for that term.

[@cassLabs]

Yes, you are right! I'm pretty sure this was my code mislabeling the legend. I'll dig into it and fix that bug. Thank you 🙂

[@Ian-qf5ny]

Very cool dude!! Wicked looking trails and unreal testing setup!! Ive always wondered about this! Ive been away from riding for a few years and this was never talked about before! I never gave it any thought. Though i always thought i was able to perceive brake jack and stiffening under hard pedaling. The false positives is a great teat pentameter! I have been tentative about high engagement hubs due to all this pedal kickback talk lately. In perfectly happy with classic low engagement hubs and whatever level of kickback i may currently be experiencing so im probably just gonna stay away from ultra fast engagement hubs and not worry anymore about it. Really glad ypu did this experiment though!! Very cool

[@KurtBob]

I'm like halfway though, and just wanna say, I love everything about this. All data driven, and, dude, ya went all the way with programming it and bustin out the callipers!

[@gucass]

I have only 3 paws on my wheel set so I hardly even feel it lol but on another bike with unlimited engagement I get 10x more kickback. All in all cheap wheelsets for the win hahah

[@ssjj9584]

I noticed higher engagement hubs have jerk back the pedals. Especially in higher gears on the brakes into bumpy rough sections in and out of turns until you get up to enough speed to not engage the hub when you pedal (depending on what gear you are in) if you are in a gear where you can pedal and engage the hub at a certain speed then hitting a bump will also engage pedal kick back. Landing on a drop in a certain gear, even though it will engage the hub as soon as you land you are propelled forward from the engagement so the energy goes to propelling you forward as the hub is engaged instead of kicking back the pedals. At least in my experience.

[@letsgoletsgoletsgoletsgoletsgo]

i run onyx hubs with infinite engagement , i haven't notice anything honestly , i will pay more attention next ride .

[@385949862]

Good stuff! Would love to see the spots in your trails where kickback happens, maybe from a 3rd person slo-mo view. Another interesting point would be to consider different frame geometries.

[@drejec1]

What's your hub engagement points?

[@JLBThreee]

I’m curious as well. 18 POE would the probably pick up less event than 54.

[@sweaty_techie]

This must be the first time i’ve seen a “bike class” in a KZbin video. Awesome 🤓 While no perceivable PK on a drop landing is expected as you feel a whole lot of force through the pedals from hitting the ground anyway and PK would just add to that, I do wonder if bikes with high PK are harder to ride through hard landings as not only is this “impact” force increased by having the PK force added, but all of that force is only through 1 foot (the front one) and throwing off your balance. Combining the measured PK value with gyroscope data that gives bike lean angle on landing across different bikes would be interesting, but the torque through each crank would also need to be measured to remove the rider’s input. The idea of significant PK is interesting too. Partly to allow for the biomechanics of what the rider naturally absorbs, and also for what value actually results in the pedal being “moved”. Filtering out for hub engagement should be straight forward as it’s a theoretically known number, although the chain tension will also come into play. With enough slack across the top of the chain and a limited enough degrees of engagement of the hub all theoretical KB events from suspension position may result in no impact at the cranks. I noticed the Fork class in there. Did you try measuring fork travel data too. I would expect that if the bike is “diving” forward then PK is also less perceivable, as the cranks rotate backwards relative to the bike but may stay fixed relative to the ground. Oh so many variables… Brake and cadence data to know if you were on the brakes are pedalling when a PK event occurred would be interesting if going down the path of understanding when PK occurs during a ride. Looking forward to seeing what else you throw together.

[@davec3568]

Good Job! I have a S4 Evo w/ WAO Unions and Hydra hubs, flat pedals and 5.10s. I notice PK in 2 circumstances... climbing: rear wheel extends then quickly compresses, descending: super chunky trail where I'm going a bit slower than intended due to a mistake. Both can be bad enough to bounce a foot off a pedal. I live in CO so have some really janky rock trails that can present worst-case scenarios, the trails in your vids are relatively smooth in comparison. One example with easy to find vids is the Bitterbrush Rock Garden at Hall Ranch in Lyons, CO. For me, it's a pretty minor issue, I notice it just a few times a year and generally as a result of not being smooth on the climb or making a mistake on the descent that forces me to slow down more than normal. With smoother climbing and maintaining over a few mph on the way down it's never an issue, but I do wonder if an O-Chain device would make things smoother... seems like you're all setup to test one! 😉

[@sryforpartiing]

Awesome video ! I was wondering the same thing last weekend at the Bikepark. I'd love to see you ride the trail and then map out in what instances the pedalkickback actually occurs. The drop is pretty obvious but knowing what section of trail provoked the kickback would be interesting

[@lucashenke6557]

Cool content, it would be cool to see if you could measure the magnitude of the kickback with pedal or crank power meters.

[@piotrekh7606]

Would you ever make a video on how to build a data acquisition system like you have?

[@cassLabs]

I will do a video explaining my system in more detail at some point! I am also planning on selling my data loggers at some point.

[@Ahaggah]

Good work man

[@donpalmera]

Would be interesting to see you do the same tests on a bunch of different rear suspension styles (single pivot, different linkages etc) to see if there is actually some non-marginal difference between them. I think a lot of reviews are bias by the fact that the reviewer sees a complex looking linkage, assumes that it has all of the good properties they have and then that feeds into the review. Like you said you experienced less kick back when you had your full face on because of the reduced noise.

[@bermroller]

Were there more events in certain cog sizes? Are you going to do this for multiple suspension types? Could the higher number of false positives indicate that some of the perceptible events might be coincidental and not be due to kickback?

[@garyewing8054]

Excellent work! Something I've been wondering about, is that since pedal kickback creates tension in the chain, it actually causes an acceleration of the rear wheel, so could theoretically make you faster! However if pedal kickback occurs more when you're on the brakes, then it's definitely a bad thing. Also, i feel that pedal kickback when climbing can cause the rear wheel to lose traction over slippy roots. Also also, I've been looking at pedal kickback on different bikes, just by holding the rear brake solid, compressing the suspension, and watching how far the cranks rotate backwards. This simple test seems valid to me, what do you think? Cheers, Gary

[@xphilippluburicx]

Awesome! That's some amazing content right there! Subscribed and like! Keep this stuff up 🙂 Greets from Germany

[@jason.miles31]

Would it be possible to plot total kickback in radians? As you mentioned it would be interesting to have a better understanding of the magnitude of the kickback. How much kickback did your cranks experience on the jungle drop? From the graph it looks like 4 radian/s per second for a pretty short duration.

[@Ahaggah]

Do you have this code available on github?

[@TheButlerNZ]

Excellent. All the "Kickback" videos I had seen all seem to base their calculations on a stationary rear wheel.. Something I have considered... IS KICKBACK ACTUALLY BAD? My understanding is, Kickback is loading the crank with negative rotation and the rear hub with positive rotation... Your legs act against the crank rotation moving the entire force to the rear hub... Positive rotation of the rear hub is acceleration... (albeit it miniscule)... "Every little helps, said the ant,"... (1590 G. MEURIER Deviz Familiers (via The Oxford Dictionary of Proverbs) ...So there mum and dad... It wasn't the Duchess!..

[@cherrymansk]

Wow! Great job!

[@Breal604]

Great work! 👍

[@yaacovr]

I'm going to start my comment with a disclaimer that my math and physics are not my strong points. My previous bike was the ibis Ripley LS v2. After about a year and a half of riding, I swapped out the 32t chainring for 30t. I then found that for whatever reason the pedal kickback became worse. I found myself on lower gears (obviously) and on these lower gears the kickback was much worse. I also found that while climbing, my shock was more compressed, though the change of chainring may not be connected. To a point where if I'm not mistaken, at the next service of my shock, I requested that a token be added. Does any of this make sense?

[@guymacarthur5161]

anti squat and pedal kickback change with chainring size. I would recommend getting linkage x3 and having a play around if yo want to know more. I don't believe a token in the shock will help though, probably would need more LSC to prevent pedal bob if anything

[@yaacovr]

@@guymacarthur5161 thanks. That was on my old bike that I sold already. I feel really validated right now so much appreciated!

[@mudders1]

Did you create this video because you were curious about the Evo's behaviour in particular? I ask as I've come off an Enduro to an Evo and gone from typical Specialized no anti squat to a lot. And over small chatter I get feedback that honestly feels like rapid shock top out or something is loose in the swat box. It's driving me nuts. Worse on the rear brake and in the bigger cogs. Absolutely fine on big stuff.

[@NE_RAW_MTB]

Great video!

[@vladimirrusev468]

For me the most important form the study is that you had more false positive than they are real PKevents. So even there is the perfect design without PK you will still feel one :) ... ok that is just the suspension not working like you would like it then. So that put at least for me the PK behind many other factors influencing the ride perception.

[@stephenlord599]

Brilliant … love the dedication, super interesting but I’m way too lazy (I’m going to just use age as an excuse)

[@klasadolfsson3213]

In general, things that you don't know if they happen or not is not a big problem.

[@borthwey]

This has gotten really silly as, let's say, an enduro bike gets designed with little anti-squat to avoid this alleged pedal kickback, and if the frame was designed for a 32t for instance, there is now the narrative that running a smaller chainring is "non-optimal" because it's not what the bike was designed for, so poor you, now you have to buy a larger cassette if you want a lower gear, you can't just install a smaller chainring, which is absolutely pathetic, because the higher anti-squat (that installing a smaller chainring provides) will give you a better pedalling performance, at the cost of this mythical increased pedal kickback, and you could actually install a much smaller chainring like a 26t and it's perfectly fine. I definitely would install a smaller chairing just for the better pedalling performance, were it not for the fact that I actually need the heavier gears.