Fellow nerds,Since I'm a believer in pedal kickback now, I have a new problem to solve. There are too many pedal kickback devices to test, and...
Fellow nerds,
Since I'm a believer in pedal kickback now, I have a new problem to solve. There are too many pedal kickback devices to test, and it would be nice if there were a way to quantify what I'm feeling, or even better yet go around my feelings entirely and quantify what's happening at my rear hub, chain, and cranks whether I'm feeling it or not.
If you wanted to test a pedal kickback device accurately, how would you go about measuring the force of pedal kickback? I included a poll below, asking where on the bike and what force you would measure if you could only measure one thing (obviously, in a perfect would you'd measure all of these forces in all of these places).
I'm including this video because it's the only example I've seen of someone measuring a PK force on their bike. It's a pretty wild setup, because he's computing the equation for chain growth in the upper chain (you could do this using readily available pedal kickback graphs, but homie calculated it from scratch the hard way), the instantaneous position of the rear axle relative to the main triangle (using a potentiometer), and the rotational acceleration of the rear wheel (using a buncha magnets) so he can calculate the rear wheel impacts that we would expect to "chase" the freehub faster than the wheel is spinning and create PK force.
I'm not nearly smart enough to replicate this setup... so I'm wondering if any of you bright people have any suggestions for a better/simpler/easier to replicate method of measuring PK force. Lemme know below.
When considering the variables involved I don't think you're going to get constant effective measurement using just one data point. Using just one data value I suspect will be misleading. That said I think the hub idea is superior to doing it at the crank but how the heck knows??
Best way to isolate variables would be to place the bike on a treadmill (or just rear wheel) to have speed control over freehub. Mount a gauge to the frame that measures peak upward force at the pedal. Then use a ram to compress the bike through it's travel. This way you can control freehub speed, what gear you are in, and the ram speed (suspension compression speed). Eliminates force of chain bouncing and only measures pedal kickback as a result of hub engagement.
if you really wanna get good benchmark results, you need to isolate as many variables as possible. if you have the means, you might want to...
if you really wanna get good benchmark results, you need to isolate as many variables as possible. if you have the means, you might want to make a test apparatus to generate lab data, then compare it against the real world data.
This was my thought as well, a kickback dyno is the cleanest way to filter out the noise and compare characteristics. Get an old frame and mount it to a fixture table. Connect a strain gauge from the crank arm to the table. Replace the shock with an pneumatic actuator (for consistent compression velocity) and mount a speed variable motor to drive the cassette/hub.
I have never done data acquisition before, but this is where I would start. It seems that you will need two data sets to differentiate between what a kickback is and what a hit/impact is. Maybe you can do it with just a strain gauge/power meter. Then, I just looked to see if the data was smoother between the different runs, but my gut told me that it might not be enough if I wanted to compare one device to another.
A simple data acquisition setup like a Motion Instruments hooked up to the rear shock, coupled with the crank arm strain gauge/power meter. But something that you can download the data into Excel (or, if you are cheap, Google Sheets) and sync up with other data.
Take "Bike A" and start logging runs. Start each run with an "event," like hitting a simple drop to sync the data between the shock logger and the power meter. Maybe dropping off a curb or small jump will work, but you want to see a point where you can sync the data.
- Run 1 would be a standard run with the bike as it was intended.
- Run 2 would be a chainless run down the same track
Multiple runs might be needed, but you want to compare the data and see if you can identify what a "kickback event" looks like. Once you know that, you can install the devices and see if you can tell if a kickback event is lessened. The closer it looks to a chainless run, the better the device scores.
Here are some thoughts to consider if you can not clearly identify a kickback event...
- You also might want one on each crank arm.
- You might want to look at a wheel speed sensor and/or brake sensor because brake jacking (although opposite) might look like a similar event.
- Also, make sure you have your data synced up properly because, most likely, the two devices sample at different rates, so you will need to find a way to make sure they align. Maybe you want to do another finishing "event," like jumping off another curb at the end of the run before you turn off the logging.
Have we considered what type of suspension to apply this to? I can feel a big differences in the pedals on large fast compressions when riding my horst link high-pivot Shore vs the multi link virtual pivot on my Revel Rail. So much in fact I wouldn't brother with a kickback device on the Shore but would entertain it on my Rail.
For all these complicated questions, the correct, easiest & most cost-effective solution is to buy/ steal/ beg borrow all the products, test them all, then keep great records. This should have a descending and a technical climb component to it, as both are important for many of us.
Keep records and create a rating system. If I $30 magnet does 80% of what a $2200 set of wheels does with out the pedaling lag, well that's an easy choice. Test with a bike with good pedaling behavior so that it also has some average kickback behavior.
You have got the Sidekick (currently 25% off I think), with 3 different positions, you also have the Taurin hub, the O'chain, STFU, Jeff's magnet device and maybe a couple of others worth testing. Personally I could reduce that to just the rear Sidekick hub, the O-chain & the magnet. That's less than $1K invested in testing.
You could have everything in hand, test everything, and sell off what you don't want, all at a pretty reasonable cost.
The thing about testing race/ high performance set ups is that theoretical & lab tests are just a starting point, as you really don't know until you test and sometimes use stop watches. So either way, it's going to all need to be tested.
I could be wrong but you might be over-thinking it. A degree value of PBK is the comparative value across bikes, so bikes with a higher value will experience worse PBK feels. All conditions equal (wheel size and speed, bump force, gear combination, crank length, etc) a bike with 25 degrees of PBK will feel worse than one with 15 degrees.
I could be wrong but you might be over-thinking it. A degree value of PBK is the comparative value across bikes, so bikes with a higher...
I could be wrong but you might be over-thinking it. A degree value of PBK is the comparative value across bikes, so bikes with a higher value will experience worse PBK feels. All conditions equal (wheel size and speed, bump force, gear combination, crank length, etc) a bike with 25 degrees of PBK will feel worse than one with 15 degrees.
We’re not talking about testing different frames for PK symptoms, we’re talking about testing different PK products. So we would be changing PK products like the Ochain or Sidekick hub and holding all else equal (same frame, same rider, same track, etc).
Got it. I figured as much while thinking about this at lunch (nerd). I think you'd have to sample the chain itself so like someone mentioned previously some sort of sensor that measures force at the chain. We have that on our Dyname motor with a Hall Effect sensor hidden and some programming in order to get a readout on the chain forces (for purposes below, while pedaling but it is also affected by PBK while descending).
Got it. I figured as much while thinking about this at lunch (nerd). I think you'd have to sample the chain itself so like someone mentioned...
Got it. I figured as much while thinking about this at lunch (nerd). I think you'd have to sample the chain itself so like someone mentioned previously some sort of sensor that measures force at the chain. We have that on our Dyname motor with a Hall Effect sensor hidden and some programming in order to get a readout on the chain forces (for purposes below, while pedaling but it is also affected by PBK while descending).
If you really want to do this properly, hit me up, and I can help you do it, as I've written and run plenty of awkward tests like this for work.
The key will be to be able to correlate subjective impression of kick back with whatever data you're recording, before you start changing things. I'd be starting by calling power meter manufacturers like 4iiii to see if you can get the unfiltered signal from both cranks, and get a good way of logging at something like 100Hz. You need to identify conditions which repeatedly demonstrate kick back, whether that's a drop or a square edge impact. I'd be thinking about somewhere hard dry and repeatable, like a car park or loading dock you can drop off rather than doing it in the woods. Drop better than ramp as it's easier to keep the same drip with different wheel speeds. Video will be critical, as actually recording chain behaviour by sensors will be a nightmare. This also allows you to record sound, including rider comments. Ideally you'd attach a GoPro to the bike looking across, but it would probably wobble too much to be useful, so either you need one repeatable impact point that you can film from beside, or I like the treadmill idea, if it can be made to work. You'll need a way of synchronising multiple signals, so having a small impact some distance before the test that also makes a noise will allow you to do that. You'd need an accurate speed reading for the rider, perhaps better than typical GPS. If you're going to do chainless, you'll obviously either need downhill or room for an ebike tow.
Once you've got the broad setup, you'd want to test some simple variables to check the data correlates with rider feedback. I'm guessing you'd want something like 5 repeats each at a few speeds. Then review the data and confirm the signals are showing what your feet do. Do an easy test like removing the chain and check the signals reflect subjective feelings. Only then move onto more fiddly and subtle tests.
Wheel speed sensor and shock position sensor are not enough as you don't cover the chain slapping around. I really think the rotations should be covered.
That’s just to determine if a kick back event was possible. Kickback and chain slap are different. You won’t want to use a position sensor on...
That’s just to determine if a kick back event was possible. Kickback and chain slap are different. You won’t want to use a position sensor on the cranks because it’s possible to have zero crank movement and feel pedal kick back.
I didn't say on the crank, I said on the chainring. That would cover the movement of a device like Ochain. I know kickback is not the same as tugging, but looking at events where there are abrupt changes in the movement could likely mean a chain tug.
You have to measure at the crank as opposed to at the cassette because we aren't sure how much is just the chain flapping about and...
You have to measure at the crank as opposed to at the cassette because we aren't sure how much is just the chain flapping about and not PK at all.
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals. How would sensor teats people are discussing be able to delineate...
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
In theory the PKB values will be higher and proportional to the wheel position or speed, but even if they weren't the total number of events would hopefully be lower when comparing 2 different bikes or devices. Crank based measurement would be easier to test with different chainrings, but you would either need a dual sided meter or a single sided one on the side of your dominant foot
Either way the theory is torque on the drivetrain is the thing with negative consequences so thats the part you will want to measure. Trying to infer results from somewhere else is always less reliable
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals. How would sensor teats people are discussing be able to delineate...
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
I guess one of the key points I was making was that you have to check that whatever sensor you're using does correlate to the subjective impression from the rider. The theory is that something is pulling the crank backwards (or perhaps forwards). Firstly we want to measure that torque, and a spider based power meter would be easier, but harder to later test ochain etc. instead, you'd need a double sided crank power meter like a 4iiii. You can filter that signal to get the back-torque out. Then I'd suggest doing the same drop (to flat) at a few different speeds to see how it feels, and see if the sensor data agrees with your feet. If it doesn't, need to try something else! Only once you have that correlation can you go off and do some experiments to see what countermeasures improve things. Having a decent slowmo camera pointing at the drivetrain will help understand what's actually going on. In reality, it's going to be quite tedious, with lots of repeating, and taking care to have exactly the same speeds and conditions for each run
Fellow nerds,Since I'm a believer in pedal kickback now, I have a new problem to solve. There are too many pedal kickback devices to test, and...
Fellow nerds,
Since I'm a believer in pedal kickback now, I have a new problem to solve. There are too many pedal kickback devices to test, and it would be nice if there were a way to quantify what I'm feeling, or even better yet go around my feelings entirely and quantify what's happening at my rear hub, chain, and cranks whether I'm feeling it or not.
If you wanted to test a pedal kickback device accurately, how would you go about measuring the force of pedal kickback? I included a poll below, asking where on the bike and what force you would measure if you could only measure one thing (obviously, in a perfect would you'd measure all of these forces in all of these places).
I'm including this video because it's the only example I've seen of someone measuring a PK force on their bike. It's a pretty wild setup, because he's computing the equation for chain growth in the upper chain (you could do this using readily available pedal kickback graphs, but homie calculated it from scratch the hard way), the instantaneous position of the rear axle relative to the main triangle (using a potentiometer), and the rotational acceleration of the rear wheel (using a buncha magnets) so he can calculate the rear wheel impacts that we would expect to "chase" the freehub faster than the wheel is spinning and create PK force.
I'm not nearly smart enough to replicate this setup... so I'm wondering if any of you bright people have any suggestions for a better/simpler/easier to replicate method of measuring PK force. Lemme know below.
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you...
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you...
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike...
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate.
I’m thinking get data from normal control runs, and data from chainless runs, and those should be your upper and lower limits for suspension performance.
I’d expect to find that the better the pedal kick solution, the closer your suspension performance will match the chainless runs.
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you...
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike...
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate. I’m thinking get data from...
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate.
I’m thinking get data from normal control runs, and data from chainless runs, and those should be your upper and lower limits for suspension performance.
I’d expect to find that the better the pedal kick solution, the closer your suspension performance will match the chainless runs.
On the data Aquisition device it would be safe to say that the shaft speeds would increase when there is less pedal kickback/feedback or if you were just to go chainless.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike...
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate. I’m thinking get data from...
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate.
I’m thinking get data from normal control runs, and data from chainless runs, and those should be your upper and lower limits for suspension performance.
I’d expect to find that the better the pedal kick solution, the closer your suspension performance will match the chainless runs.
On the data Aquisition device it would be safe to say that the shaft speeds would increase when there is less pedal kickback/feedback or if you...
On the data Aquisition device it would be safe to say that the shaft speeds would increase when there is less pedal kickback/feedback or if you were just to go chainless.
Is that safe to say?
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals. How would sensor teats people are discussing be able to delineate...
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
I'm curious to hear more about the STFU chain damper and pedal feedback as I only ever see that in passing comments like yours. What bike did you have it on and notice that?
I am also curious if an Onyx hub would reduce pedal feedback, even though it is instant engagement, because it is a soft engagement. Rather than the harsh, abrupt, clank resonating through the pedals, would it be damped?
As for this experiment, would measuring the suspension only focus on PKB, when in reality, perhaps what a lot of people are feeling (and calling pkb) is just pedal feedback from the chain. It would seem prudent to test for all of the above.
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.But I think the trail needs to...
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
This makes a lot of sense. Too fast of a trail will likely produce too few measurable PK events. Ideally it'd be slow to medium and HARSH
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.But I think the trail needs to...
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
This makes a lot of sense. Too fast of a trail will likely produce too few measurable PK events. Ideally it'd be slow to medium and...
This makes a lot of sense. Too fast of a trail will likely produce too few measurable PK events. Ideally it'd be slow to medium and HARSH
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2 seconds. People tend to disregard high speed sections because they assume the hub will never engage, but that stop being true as soon as your wheel stops spinning freely.
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.But I think the trail needs to...
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2...
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2 seconds. People tend to disregard high speed sections because they assume the hub will never engage, but that stop being true as soon as your wheel stops spinning freely.
Hi speed super steeps where your brakes are on the edge of locking up but you're still cruising.
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.But I think the trail needs to...
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2...
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2 seconds. People tend to disregard high speed sections because they assume the hub will never engage, but that stop being true as soon as your wheel stops spinning freely.
So initally I was joking about that. I'm always down for a little bike-related torture.
...but now that I'm thinking about it:
Want to get a good look at pedal kickback? Take forward momentum out of the equation all together. All that you're left with is how much the kickback effects the kickback.
3ft high ledge, kick the cranks up against the hub engagement, go a little light on the suspension tune so you get absolute bottom-out, hold the rear brake, take plenty of pain killers, and hop slightly sideways off the ledge.
Be consistent with the amount of upwards and the amount of sideways you hop and it'll be the most repeatable "effect on suspension" test with the fewest "hard-to-eliminate-because-of-speed-and-inconsistencies-in-trail/line/surface" variables.
If nothing else, it'll help you define what kickback looks like in a data acquisition sort of way and will help you know what to look for when comparing to on-the-trail data.
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you...
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike...
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
Hey Charlie,
When we did DA testing with Sidekick that also incorporated crank sensors, we saw significantly more time being spent in the mid-stroke (~50%!) and notable reductions in crank acceleration forces (~20% less). We'll have a video in the near future that looks at the DA testing we did with Sidekick and dives a bit deeper.
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals. How would sensor teats people are discussing be able to delineate...
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
I'm curious to hear more about the STFU chain damper and pedal feedback as I only ever see that in passing comments like yours. What bike...
I'm curious to hear more about the STFU chain damper and pedal feedback as I only ever see that in passing comments like yours. What bike did you have it on and notice that?
I am also curious if an Onyx hub would reduce pedal feedback, even though it is instant engagement, because it is a soft engagement. Rather than the harsh, abrupt, clank resonating through the pedals, would it be damped?
As for this experiment, would measuring the suspension only focus on PKB, when in reality, perhaps what a lot of people are feeling (and calling pkb) is just pedal feedback from the chain. It would seem prudent to test for all of the above.
I ride a reeb sst with flat pedals and onyx hub. I notice the soft catch of the onyx by way of increased traction on technical half-ratchet climbs, were too much torque too quickly would spin out with a sudden pawl engagement. The soft catch is almost like a mechanical traction control for the first couple of degrees of engagement. This is coming from 10 degree hubs, not a high tooth count pawl hub. Headed downhill the sprag clutch may take some of the sting out of pkb, but it is still high engagement so you still feel the chain tug the pedals.
Putting on the STFU made the ride feel smoother, more refined and less busy. I was curious about a low degree ochain with the onyx but since I got the STFU I have not thought about it. So if you are interested in ochain, maybe start with the $40 and zero maintenance STFU then go from there.
When considering the variables involved I don't think you're going to get constant effective measurement using just one data point. Using just one data value I suspect will be misleading. That said I think the hub idea is superior to doing it at the crank but how the heck knows??
Best way to isolate variables would be to place the bike on a treadmill (or just rear wheel) to have speed control over freehub. Mount a gauge to the frame that measures peak upward force at the pedal. Then use a ram to compress the bike through it's travel. This way you can control freehub speed, what gear you are in, and the ram speed (suspension compression speed). Eliminates force of chain bouncing and only measures pedal kickback as a result of hub engagement.
This was my thought as well, a kickback dyno is the cleanest way to filter out the noise and compare characteristics. Get an old frame and mount it to a fixture table. Connect a strain gauge from the crank arm to the table. Replace the shock with an pneumatic actuator (for consistent compression velocity) and mount a speed variable motor to drive the cassette/hub.
Cool problem! I love this stuff.
I have never done data acquisition before, but this is where I would start. It seems that you will need two data sets to differentiate between what a kickback is and what a hit/impact is. Maybe you can do it with just a strain gauge/power meter. Then, I just looked to see if the data was smoother between the different runs, but my gut told me that it might not be enough if I wanted to compare one device to another.
A simple data acquisition setup like a Motion Instruments hooked up to the rear shock, coupled with the crank arm strain gauge/power meter. But something that you can download the data into Excel (or, if you are cheap, Google Sheets) and sync up with other data.
Take "Bike A" and start logging runs. Start each run with an "event," like hitting a simple drop to sync the data between the shock logger and the power meter. Maybe dropping off a curb or small jump will work, but you want to see a point where you can sync the data.
- Run 1 would be a standard run with the bike as it was intended.
- Run 2 would be a chainless run down the same track
Multiple runs might be needed, but you want to compare the data and see if you can identify what a "kickback event" looks like. Once you know that, you can install the devices and see if you can tell if a kickback event is lessened. The closer it looks to a chainless run, the better the device scores.
Here are some thoughts to consider if you can not clearly identify a kickback event...
- You also might want one on each crank arm.
- You might want to look at a wheel speed sensor and/or brake sensor because brake jacking (although opposite) might look like a similar event.
- Also, make sure you have your data synced up properly because, most likely, the two devices sample at different rates, so you will need to find a way to make sure they align. Maybe you want to do another finishing "event," like jumping off another curb at the end of the run before you turn off the logging.
Love to hear how you end up tackling this!
Have we considered what type of suspension to apply this to? I can feel a big differences in the pedals on large fast compressions when riding my horst link high-pivot Shore vs the multi link virtual pivot on my Revel Rail. So much in fact I wouldn't brother with a kickback device on the Shore but would entertain it on my Rail.
For all these complicated questions, the correct, easiest & most cost-effective solution is to buy/ steal/ beg borrow all the products, test them all, then keep great records. This should have a descending and a technical climb component to it, as both are important for many of us.
Keep records and create a rating system. If I $30 magnet does 80% of what a $2200 set of wheels does with out the pedaling lag, well that's an easy choice. Test with a bike with good pedaling behavior so that it also has some average kickback behavior.
You have got the Sidekick (currently 25% off I think), with 3 different positions, you also have the Taurin hub, the O'chain, STFU, Jeff's magnet device and maybe a couple of others worth testing. Personally I could reduce that to just the rear Sidekick hub, the O-chain & the magnet. That's less than $1K invested in testing.
You could have everything in hand, test everything, and sell off what you don't want, all at a pretty reasonable cost.
The thing about testing race/ high performance set ups is that theoretical & lab tests are just a starting point, as you really don't know until you test and sometimes use stop watches. So either way, it's going to all need to be tested.
I could be wrong but you might be over-thinking it. A degree value of PBK is the comparative value across bikes, so bikes with a higher value will experience worse PBK feels. All conditions equal (wheel size and speed, bump force, gear combination, crank length, etc) a bike with 25 degrees of PBK will feel worse than one with 15 degrees.
We’re not talking about testing different frames for PK symptoms, we’re talking about testing different PK products. So we would be changing PK products like the Ochain or Sidekick hub and holding all else equal (same frame, same rider, same track, etc).
Got it. I figured as much while thinking about this at lunch (nerd). I think you'd have to sample the chain itself so like someone mentioned previously some sort of sensor that measures force at the chain. We have that on our Dyname motor with a Hall Effect sensor hidden and some programming in order to get a readout on the chain forces (for purposes below, while pedaling but it is also affected by PBK while descending).
Please sell frame/ motor/ battery combos. Thank you!
All of it. It's all chain flapping around.
If you really want to do this properly, hit me up, and I can help you do it, as I've written and run plenty of awkward tests like this for work.
The key will be to be able to correlate subjective impression of kick back with whatever data you're recording, before you start changing things. I'd be starting by calling power meter manufacturers like 4iiii to see if you can get the unfiltered signal from both cranks, and get a good way of logging at something like 100Hz. You need to identify conditions which repeatedly demonstrate kick back, whether that's a drop or a square edge impact. I'd be thinking about somewhere hard dry and repeatable, like a car park or loading dock you can drop off rather than doing it in the woods. Drop better than ramp as it's easier to keep the same drip with different wheel speeds. Video will be critical, as actually recording chain behaviour by sensors will be a nightmare. This also allows you to record sound, including rider comments. Ideally you'd attach a GoPro to the bike looking across, but it would probably wobble too much to be useful, so either you need one repeatable impact point that you can film from beside, or I like the treadmill idea, if it can be made to work. You'll need a way of synchronising multiple signals, so having a small impact some distance before the test that also makes a noise will allow you to do that. You'd need an accurate speed reading for the rider, perhaps better than typical GPS. If you're going to do chainless, you'll obviously either need downhill or room for an ebike tow.
Once you've got the broad setup, you'd want to test some simple variables to check the data correlates with rider feedback. I'm guessing you'd want something like 5 repeats each at a few speeds. Then review the data and confirm the signals are showing what your feet do. Do an easy test like removing the chain and check the signals reflect subjective feelings. Only then move onto more fiddly and subtle tests.
Phew, that was long!
I didn't say on the crank, I said on the chainring. That would cover the movement of a device like Ochain. I know kickback is not the same as tugging, but looking at events where there are abrupt changes in the movement could likely mean a chain tug.
STFU chain damper is immediately noticeable, no brainer for anyone who notices feedback through your pedals.
How would sensor teats people are discussing be able to delineate pkb from chain slap? Tug is a tug, whether from chain oscillation or mismatched shaft/freehub speed, right?
In theory the PKB values will be higher and proportional to the wheel position or speed, but even if they weren't the total number of events would hopefully be lower when comparing 2 different bikes or devices. Crank based measurement would be easier to test with different chainrings, but you would either need a dual sided meter or a single sided one on the side of your dominant foot
Either way the theory is torque on the drivetrain is the thing with negative consequences so thats the part you will want to measure. Trying to infer results from somewhere else is always less reliable
I guess one of the key points I was making was that you have to check that whatever sensor you're using does correlate to the subjective impression from the rider. The theory is that something is pulling the crank backwards (or perhaps forwards). Firstly we want to measure that torque, and a spider based power meter would be easier, but harder to later test ochain etc. instead, you'd need a double sided crank power meter like a 4iiii. You can filter that signal to get the back-torque out. Then I'd suggest doing the same drop (to flat) at a few different speeds to see how it feels, and see if the sensor data agrees with your feet. If it doesn't, need to try something else! Only once you have that correlation can you go off and do some experiments to see what countermeasures improve things. Having a decent slowmo camera pointing at the drivetrain will help understand what's actually going on. In reality, it's going to be quite tedious, with lots of repeating, and taking care to have exactly the same speeds and conditions for each run
Hey Charlie, it just occurred to me that in practice, pedal kick does not actually manifest as pedal kick. It manifests as encumbered suspension. If you want to see the on trail effects of different pedal kick solutions, you should get some data acquisition for your rear suspension and ride a very repeatable section of trail. You’d be able to see the differences between Ochain and the sidekick hub etc by comparing your suspension usage.
Thanks Blake, this is exactly the discussion I'm hoping to have on here. What sort of difference would you expect to see in the data? Bike using more travel, using less, more force at the rear axle, or less?
My hunch is the bike with bad pedal kick will use less travel because it's getting hung up on the chain/anti-squat as it's trying to compress. (I think this would produce more measurable force at the pedals, too, because you're getting more force in your feet as opposed to absorbed by your rear spring).
My hunch is the same. Usually when people ride their bike chainless they report that they could go up in damping/spring rate.
I’m thinking get data from normal control runs, and data from chainless runs, and those should be your upper and lower limits for suspension performance.
I’d expect to find that the better the pedal kick solution, the closer your suspension performance will match the chainless runs.
On the data Aquisition device it would be safe to say that the shaft speeds would increase when there is less pedal kickback/feedback or if you were just to go chainless.
Is that safe to say?
I’d assume this is correct. But I’ve tested on a fast trail and didn’t see any changes in the data.
But I think the trail needs to be on the slow/medium side of speed. This way the POE on the hub engages to create pedal kick. If the trail is too fast and the shafts speeds aren’t fast enough, I don’t think it will happen. Or need to test with a high Poe hub like I9 or Onyx that is instant engagement.
I'm curious to hear more about the STFU chain damper and pedal feedback as I only ever see that in passing comments like yours. What bike did you have it on and notice that?
I am also curious if an Onyx hub would reduce pedal feedback, even though it is instant engagement, because it is a soft engagement. Rather than the harsh, abrupt, clank resonating through the pedals, would it be damped?
As for this experiment, would measuring the suspension only focus on PKB, when in reality, perhaps what a lot of people are feeling (and calling pkb) is just pedal feedback from the chain. It would seem prudent to test for all of the above.
This makes a lot of sense. Too fast of a trail will likely produce too few measurable PK events. Ideally it'd be slow to medium and HARSH
Ideally you’d get data on a slow chunky section, a fast chattery section, and a fast chattery section where you slam on the brakes for ~2 seconds. People tend to disregard high speed sections because they assume the hub will never engage, but that stop being true as soon as your wheel stops spinning freely.
Hi speed super steeps where your brakes are on the edge of locking up but you're still cruising.
HUCK TO FLAT
HUCK TO FLAT
HUCK TO FLAT
(say it with me everyone)
HUCK TO FLAT!
Yeah, I think for these types of trails all you guys listed, is the worst-case scenario or PK/feedback to feet.
So initally I was joking about that. I'm always down for a little bike-related torture.
...but now that I'm thinking about it:
Want to get a good look at pedal kickback? Take forward momentum out of the equation all together. All that you're left with is how much the kickback effects the kickback.
3ft high ledge, kick the cranks up against the hub engagement, go a little light on the suspension tune so you get absolute bottom-out, hold the rear brake, take plenty of pain killers, and hop slightly sideways off the ledge.
Be consistent with the amount of upwards and the amount of sideways you hop and it'll be the most repeatable "effect on suspension" test with the fewest "hard-to-eliminate-because-of-speed-and-inconsistencies-in-trail/line/surface" variables.
If nothing else, it'll help you define what kickback looks like in a data acquisition sort of way and will help you know what to look for when comparing to on-the-trail data.
Hey Charlie,
When we did DA testing with Sidekick that also incorporated crank sensors, we saw significantly more time being spent in the mid-stroke (~50%!) and notable reductions in crank acceleration forces (~20% less). We'll have a video in the near future that looks at the DA testing we did with Sidekick and dives a bit deeper.
I ride a reeb sst with flat pedals and onyx hub. I notice the soft catch of the onyx by way of increased traction on technical half-ratchet climbs, were too much torque too quickly would spin out with a sudden pawl engagement. The soft catch is almost like a mechanical traction control for the first couple of degrees of engagement. This is coming from 10 degree hubs, not a high tooth count pawl hub. Headed downhill the sprag clutch may take some of the sting out of pkb, but it is still high engagement so you still feel the chain tug the pedals.
Putting on the STFU made the ride feel smoother, more refined and less busy. I was curious about a low degree ochain with the onyx but since I got the STFU I have not thought about it. So if you are interested in ochain, maybe start with the $40 and zero maintenance STFU then go from there.
Post a reply to: How would you measure pedal kickback?