Negative vs. positive pedal feedback
Well, it turns out I have plenty of time right now to be an armchair suspension analyst.
I am interested in discussing the pedal feedback potential of the dw-link susp. vs. the blur.
They're both virtual pivot designs, and actually both have the pivots in the same location.
(At least if you compare the mojo to the blur, you'll see that the swing link is located
in the same location more or less on the seat tube, as well as the virtual pivot.)
The biggest difference appears to be in the length of the links, and the susp. path
on the blur allows a more upward path than on the dw-link, but they both rotate the rear triangle up and forward as as virtual pivot design does. My thinking is that the pedal feedback would be less on the dw-link because of the shorter links, but the lower link on the mojo is the same shape as on the blur, just shorter. This to me seems like the only real reason to assume off the bat that the mojo has less pedal feedback. (Which I'm assuming it does, based on what I've read from this forum) You have less potential for chain growth with shorter links.
The flipside to the shorter links issue is I'm questioning whether the effective travel of the
mojo is really 5.5. With those short links I'm wondering whether you can in reality get any
more than 4.5.
As far as positive vs. negative chain growth, the dw-link site has a
'positive chain growth' demo showing an arrow in the direction of the chain growth. Positive meaning the effective chainstay length getting longer.
You would have to assume that at some point during rebound, the laws of physics
take over and the chain growth would have to be negative. How and when this happens,
ie, during what part of the suspension cycle, can effect how much of the chain tension you're feeling through the pedals, but I cannot imagine how you could not feel any chain tension at all at some point while pedaling the mojo in the granny.
It might pedal very well, but if you are the kind of person
who pays attention to changes in pedaling tension, as I do, you should be able to notice
it at some point, as the chain tension is being used to acheive an anti-squat effect.
Yes, I know people who will claim they can't feel any pedal feedback
on the blur - perhaps their riding style is different, or they're not paying attention(more
likely the latter)
I do some real steep granny climbs -
if your pedaling cadence is slow you will notice more changes in pedaling tension than
if you are applying constant torque.
I think both bikes have a rearward rebounding path, the mojo might be more of a straight
rearward than the blur. But both bikes overall share similar design characteristics,
so I believe I would be correct in assuming they both share some level of pedal feedback,
the blur being more so because of the longer links. Any ideas on this ? Am I missing
some other consideration in my armchair evaluation ?
I am not going to get into this right now due to lack of time, but I need to point out a couple inaccuracies. You say that the Mojo and Blur share pivot locations and design characteristics. That could not be further from the truth. The Mojo and Blur pivot locations are not even close to each other, and the two bike share virtually no design characteristics other than the fact that they both use a triangulated swingarm and short links.
Originally Posted by le_buzz
You would have a hard time finding two suspension systems that are more dissimilar than dw-link and vpp, and two bikes that are more dissimilar than the mojo and a blur.
I have written huge volumes of information on this subject. Check out the archives here and on Ridemonkey and you will find what you are looking for.
That's fine, dw, thanks for the references. I did some research into some of the old threads and did find some good discussion on this. Sounds for the most part like the dw-link is the real deal. I found one reference to pedal feedback on a dw-link bike, but this seemed like it occurred in a relatively uncommon situation. There are quite a few technical climbs in my neck of the woods that might make this situation more common for me, but still, the pedaling feedback seems pretty minimal, and the overall pedaling efficiency would
Originally Posted by _dw
outweigh whatever minor pedal feedback issues there might be.
One thing I should make note of, though. after reading a lot of derby's old posts,
he makes numeruos references to the dw-link being the least pedaling feedback of all
suspension designs. Although derby's posts and theories seem right on about a lot of
aspects of suspension, some statements like this one seem a bit of an exaggeration,
ecpecially when you consider that a pedaling neutral horst link suspension system
would have no opportunity for pedal feedback to occur, whereas a suspension system
that uses chain torque, like the dw-link and others, would. Thats not to say dw-link
doesnt do a good jod of neutralizing it, but if you wanted 0 pedaling feedback in all
situations, the horst is clearly the only one that actually accomplishes this 100 % of the
time. Of course, then youre trading a potential pedaling issue for a bob issue, so it comes down to personal preference, whether you'd rather address the bob issue with chain tension or a platform shock. Most people, including me, would rather put up with some
limited pedal feedback, as long as its rare and minimal, than a lot of bob.
Since theres no way to actually demo a dw-link bike, maybe the
best thing to do would be to go down to the local Performance shop and have one of
the employees push up and down on one of the Iron Horses to activate the suspension while I watch the chain growth. Probably wont tell me much, but it wold be better than
doing a parking lot test, I think. The last time I checked out one of those bikes,
I remember the virtual pivot (bottom bracket pivot) to be very similar to the blurs as far as
looks and location, but now that I think about it, my statement about the upper swing link pivot also being in the same relative location is wrong. It is lower on the dw-link, the
rear triangle on the dw is much smaller and arcs differently as well. Its not always easy
to get the picture of how it things work from looking at demos, but yours is pretty good.
I wish it had shown the chain growth in the negative direction also, though.
Anyone, thanks for responding to my question. Some of the physics discussion was hard
to follow, but still was informative. If anything, it has complicated my buying decision,
but I have plenty of time to study up on the physics of it, as I wont be able to afford
to get anything for awhile, so its a all good.
It's the axle
I'm not sure my experience would mean anything to you. On my NRS I noticed pedal feedback when climbing rough stuff or rolling over big things at slow speeds. On the Mojo I have noticed zero pedal feedback. And I've been paying keen attention to see if I can notice it. I have riden about 150 miles on trails just this week alone. Today I rode over 30 miles on logging roads that had a lot of trees that recently fell and haven't been cleared. Some were small enough (up to a foot diam) that I just rode over them while pedaling. Zero feedback. Also I rode over some pretty deep ditches (sort of inverse water bars), and again had no feedback at all. I had an amazing day discovering a new park with a group I've never ridden with. And I was comfortably running up and down on these guys. It was cool. I'm so impressed with this bike. I can't find a fault. Hopefully my experience is worth something to your search.
Pedal feedback will be most noticable if you're climbing terrain that's rocky with lots of
sharp ups and downs in the granny ring. Thats when I most notice it on the blur.
Just climbing over obstacles doesnt produce pedal feedback. I can climb over ledges without
noticing it at all. The terrain has to be rapidly changing with lots of sharp rocks, and steep.
I live in a state where lots of the terrain is like this. If I can attack that terrain out of the
saddle and keep a reasonably steady spin, then I hardly notice any pedal feedback.
I'm wondering how many of the folks that say the dw-link has zero pedal feedback
actually ride this kind of terrain. Terrain that will bring out the worst traits of any
suspension system. I've seen only on dw-link on the trails here, an Iron Horse.
Of course that could just mean people here havent discovered it yet.
The classic horst link seems the smoothest on that kind of stuff, even though it doesnt pedal
Although marketing dollars may say otherwise, Horst link is in no way designed to be neutral to pedaling feedback, many rider's experiences, and kinematic analysis backs that up. Horst can be a good system when implemented correctly, but in current an past implementation, Horst link is in no way designed to specificallly target any performance parameter other than improved braking over a single pivot with the same main pivot location.
Originally Posted by le_buzz
dw-link does not use chain torque to manipulate any suspenion characteristics. Chain pull force is actually always the lesser of the two forces that manipulate the suspension. dw-link manipulates the driving force in the suspension in relation the the chain force vector to remain neutral to bumps and the effects of mass transfer.
Small amounts of pedal feedback still exists, but due to the small amount lack of abrupt changes in pedal feedback amount, your body has a very hard time sensing it, and it will not negatively affect performance. The same cannot be said for horst link or vpp.
You can read a little bit more about this on the ibis site where I wrote a short piece recently. http://www.ibiscycles.com/tech/dw_link/
I know you said that a test ride is tough, but in my experience that is the one thing that riders seem to universally understand. I urge you to do whatever it takes to get a ride on a bike that is set up for your size and weight.
Good luck and happy testing!
Relative Host link types
I should qualify my statement about dw_link feedback, because every bicycle has pedal feedback when hitting bumps, from hardtail design to 6-link Felt:
Originally Posted by le_buzz
Unlike every nearly every other design, the dw-Link has no UNDESIRABLE pedal feedback.
The only other design I’ve ridden that is close in smooth stable and efficient pedal performance to dw-link but only in the 4 to 5 inch range of travel is the Whyte design used by Marin’s Wolf Ridge.
Horst link type designs have a somewhat wide interpretation. Some say ICT is a horst link and it has no noticeable difference than a monopivot in the same seat tube position in pedaling performance, Kona “faux-bar” bike design using the “Dope” floating brake link is identical within 0.005% in performance to ICT except heavier. While the most classic Horst link designs done by Titus have a virtual pivot where the center of the round shaped path curvature is located a few inches behind the seat tube in a location impractical for a monopivot, but it has monopivot pedaling characteristics. The braking tensions are focused differently than monpivot due to the floating link the caliper is mount so that the floating brake design is the difference from Horst link from monopivot.
There are Horst type designs with less and more anti-squat designed by path geometry. Specialized has less anti-squat and results in more bob than the classic Host design by Titus (the ’07 models have returned to their earlier classic Hosrt designs so these new FSR’s should be closer to Titus in quality now). The Intense Horst links have a little more anti-squat and less bob than Titus but the pedal kickback becomes noticeable in the smallest granny gears. NRS has even more noticeable kickback and must be run topped out to avoid pedal jack induced bob.
The S-path VPP has a very smooth transition into pedal feedback. And the feedback is very noticeable as pedal induced jack (extension) when standing and climbing in the small and middle rings on a very smooth high traction surface such as pavement. Also while seated pedaling in deep repeated bumps there is very noticeable pedal cadence stall, severely interrupting momentum in bumps.
And of course we’ve all gone to full suspension to reduce with the desire to eliminate the kick in the pedals and butt from hardtails when pedaling through bumps.
Every design including dw-Link has feedback. The difference with dw-Link is every aspect of the feel of the pedals, from smooth to extremely rough, is most desirable and enhances the ability to accelerate, climb, and increase momentum.
Only the dw-Link eliminates pedal induced kickback, cadence stall, squat, bob, and wallow from 3.75 inch to 9+ inch travel.
Any other designer that claims this is not being honest or is very naïve.
Thanks for the demo and the explanation, Dave. That clears up some of the questions I
Originally Posted by _dw
have but adds a couple others:
In the demo you say that 2 forces that create anti-squat are chain pull and driving force.
Isn't chain pull to create anti squat another synonym for chain torque or chain tension,
which is what vpp uses for anti-squat. Not sure I see the difference here as far as that
aspect goes. I think the heart or your argument for dw-link superiority revolves around
position sensitive anti-squat.
Are you talking about rider position vis-a-vis the center of mass or shock position ?
Can you briefly explain what the 3 positions are ?
How would out of the saddle climbing affect the position
sensitive antisquat ? Does the linkage engineer some squat back in here for
extra compliance and reduced pedal feedback, as you refer to in the third stage, deeper
travel position ? Thanks.
You have some very interesting takes on suspension characteristics, derby.
Here are some comments and observations:
First of all, I don't consider hitting a bump to be pedal feedback, its a different
Originally Posted by derby
sensation entirely than chain tension produced pedal feedback. If chain tension
produced pedal feedback occurrs at the same time as when you're hitting a bump,
the actual pedal feedback sensed can be amplified further, in that respect I would
As far as the Wolf Ridge, this is as far as I know a single pivot, which would
presumably have quite a bit of pedal feedback, so how would that be a favorable
So you're saying the major performance difference between a monopivot and a Titus
Originally Posted by derby
horst link would be in braking ? I haven't ridden a classic monopivot, only a semi-
faux bar(Rocky Element) but have ridden a Motolite; I have a hard time believing
any monopivot would share similar characteristics to that classic horst design.
Can you explain how the Motolite has monopivot pedalinig characteristics ?
Are you referring to pedaling efficiency as in a faux bar monopivot (bob)?
My riding experience with the Motolite was that it would be quite different than a
Monopivot, ie, it has no pedal feedback, which a monopivot has, it has great climbing
traction, which a monopivot does not, it would have better small bump sensitivity
and also, as you mention, better braking. Certain well executed monopivots might
have better efficiency but at the expense of horrible pedal feedback.
I experienced no pedal feedback in extended rides on the Motolite and FSR and
terrain that would produce it, so I have to assume that horst's dont have
I won't nitpick your comments about the vpp pedal feedback, except to say that I notice
Originally Posted by derby
relatively little pedal feedback on high traction granny ring climbs, usually none on
middle ring climbs. I notice a fair amount on very rough, bumpy granny ring climbs.
Out of the saddle the sensation is somewhat reduced as you're pumping more
torque into the chain which seems to even out the pedaling tension fluctuation,
or perhaps it masks this out a bit.
The (current) Marin suspension system is not single pivot. It's a virtual pivot system called Quad-Link.
Now I've added that, I'll leave you to argue about these things whilst I go and ride my bike...
Until my dw-Link'ed Mojo, I'm coming off a very high quality Host link for 5 years the Intense Tracer, in my opinion the best balanced and most optimized and versitle Horst link type design ever produced. It’s a great bike, better overall than any other for XC/AM until the Mojo. I’m spoiled now and won’t ride the Tracer anymore and will sadly will sell ‘er soon only because I don’t have room to keep it.
Originally Posted by le_buzz
So I think my sensitivity to the standing pedal jack of S-path VPP is somewhat exaggerated. VPP's standing pedal jack with the Intense 5.5 and 6.6 is definitely worse than a Superlight high monopivot which I also rode for 2 year previous to my Tracer, maybe the longer travel VPP's than my Superlight exaggerates the VPP problem. And I could adapt and soften my natural standing cadence to reduce the VPP standing pedal jack, which I've never noticed in any gear of a high quality Horst link. I do prefer the soft feeling kickback of VPP over the noticeably sharp kickback of my Superlight, VPP is noticeably better than high monopivot.
Except the NRS, any Horst link designed so far could produce the same pedaling effects if a pivot was placed at the center of curvature. Most Horst links have a virtual pivot well behind the seat tube in a place impracticable for a mechanical pivot due to the tire or derailleur being in the way. Maybe we'll see single sided monopivot swingarms someday that don't flex too much, nor interfere with pedaling to be practical, that would produce the same path and mechanically netting the same pedaling performance as a Host link.
The NRS exception to Horst link types having a circular monopivot type path, is that the NRS path is mildly elliptical and the virtual pivot where the path radiuses intersect migrates from just above and near mid chainstay in position towards the seat stay a couple inches during compression travel. But we are entering the practical potentials of short links with the NRS.
BTW, the Wolf Ridge was changed from high monopivot to short-link multilink about 4 or 5 years ago. Also my hardtail and rigid bikes always kicked and stalled my cadence in sharp bumps. (Maybe you are setting the hardtail kick as a standard of acceptance). The dw-Link is much smoother pedaling in sharp rocks than any hardtail, so the fact is that hardtails have more noticable kickback or feedback compared to dw-Link.
Your questions are very good. But you will need more ride-time experience of a variety of designs now to progress much further in understanding. Theory alone will take anyone off into mythical dimensions of unreality.
Or you can just shortcut to the very best and buy a Mojo. After that you should continue to test many other designs. But you'll find that you are quickly spoiled for the highest quality performance and you won't want to stay on any other bike very long!
Here’s the most thorough published study in multilink pedaling theory with easy to understand examples, clear logic, and physics proofs of suspension analysis. It needs to be read a few times without preconceived opinions because it clearly disproves a few often repeated suspension design myths:
Keep seeking! Even when you believe you’ve found it.
It's the axle
Derby, thanks for the link. But it's not opening for me. I even tried the main url page, and nothing.
I should add that as a mechanical engineer who has also ridden mountain bikes since the mid 80's, and road, dirt, motorcycles since well before, I am not just some kid spouting off. I can say without a doubt that where there was pedal feedback on the NRS, there is none with the DW. Or maybe like was said before, none that is subtractive of the riding experience. It's smooth, if that means anything.
And I'll also add that even having taken vibrations and kinetics courses in mechanisms, I honestly do not understand what is going on with this linkage. Even after having tried to make sense of the patent claims. And like many of these linkages, it takes a computer printout of the force and acceleration and position in order to get a sense. Or have a good feel for the particular design. Or seat of the pants from a real situation. Sometimes a little of both is the only way to understand if something is right. But I think riding this bike will suffice. I might also say that I've never ridden a bike that didn't have me limping around with a sore butt after all the miles I've put on it.
Originally Posted by Gregg K
I don't know if you are old enough to have done graphical cam/mechanism design where one looks at intantaneous displacement, the deriviative of displacement (velocity), the derivative of velocity (acceleration), and the derivative of acceleration (jerk).
Dave has worked very hard to make sure there is not any jerk in the DW link pedal feedback.
I think the best way to make these parameters understandible is to plot displacement, velocity, acceleration and jerk as a stacked set of y axises(sp?:-) with a shared x axis labeled time. Start by drawing chainstay length as a function of time. Use dividers to find local d(chainstay growth)/dt to get velocity. After drawing that graph, echo what you did before to get acceleration and so on.
Looking at the DW link ingredients, the two easiest to fathom are leverage ratio(1/4th), and chainstay growth/pedal feedback (1/6th).
Patents exist to encourage inventors to teach their art so others can build on it and the world becomes a better place. For those skilled in the art of suspension design reading the patent should work. There may be less than 20 folks in the US skilled in that art...
You put it well: Designing these things is esoteric. To appreciate these things all you have to be able to do is ride a bike.
It's the axle
Clark, thanks for the post. Yes, I am old enough. But I was also in the courses at a time when we were just starting to use computers. As a result I missed a lot of learning due to the effort involved in programming. Actually, I'm not much of an engineer, as such. I'm more of a designer. In fact, I accomplish my engineering by two things that are not desireable. Luck and persistence. I used to joke that I should call myself Accidental Engineering. But that's why I never worked for someone else. OK, I've prostrated enough.
Hey, I even used a slide rule in my first attempt to go through college. I'll never forget when my friend in computer science (1974) got an actual four function calculator. Wow! I kind of liked my old circular slide rule. Pretty cool. I really should have deleted this post. I've said way too much. We're all kids at heart. Thus, mountain biking.
I haven't ridden the newer Quad link. The Wolf Ridge Tara-Quad is still in active production in the '07 Marin line and unchanged from the last couple years except the variable travel is not offered and travel is fixed at 5 inches (too bad! but maybe to reduce flex?)
Originally Posted by TJT
With just a quick visual look at the Quad-Link it appears to produce a somewhat similar IC set and migration map during travel as the Tara-Quad. If it was designed by Jon Whyte it is probably one of the very best. (Although I realy liked the Tara-Quad of the Wolf Ridge, I didn't like the ride of the previous XC design, I forget the name XC-Quad?, which the Mt. Vision used compared to my Horst link Tracer with the same travel.)
Whyte was a rare breed in bicycle suspension design, he was real mechanical design engineer previous to his career designing bicycle suspension and produced the suspension that began Michael Schumacher’s long residence at the top of the F1 podium. There are a few others with vehicle mechanics design background, Horst Leitner and Mert Lawwill had done a their suspension designs on motorcycles before bicycles. Felt comes from motorcycles.
DW is probably coming from the most advanced mechanical design education and background to bicycle suspension design, even compared to Whyte's F1 background in the ‘80’s and early ‘90’s.
Yes, this is true.
Originally Posted by derby
Thanks, derby for the link and the insightful comments. I tend to separate in my mind terrain induced feedback from drivetrain induced feedback, but I see your point as far
as the overall smoothnes of the ride.
I didn't have time to do more than scan through that website. That could be quite a resource, it seems a little dated (2001) but still worth going through at some point.
Does the Linkage software have a dw-link mock up , or did the dw come later ?
Anyway, while scanning through some of the pages there I thought I noticed some statements about the vertical path of the classic horst link not being prone to pedal feedback, while designs that have a variable length chainstay are, which is not intuitively
hard to guess. ( I may be summarizing or simplifying what was stated - there was much more than just that, and it was probably stated in an overall context)
I will definitely read more of the sections when I have more time - its a very worthwhile
read for an understanding of the physics behind suspension.
Youve evidently ridden many bikes, so your advice to 'take the shortcut to the best' is