By sharing my own (frustrating) experience, I hope to help out others who might be having bike-fit issues, knee-pain, etc. I’ll follow up in an appropriate time once I see how I am progressing.
I spent ~5 hours at 3DBikeFit.com two days ago, working with Kevin Bailey, who gets very precise and picky on the fit.
We used video extensively, checking on both knee tracking and centering of the body on the bike. The use of video (including before/after side-by-side video) is persuasive in showing the knees wobbling or diving in, etc (black stick-on dots help make this easy to see).
With before/after video recording, a 1mm change can be made can be made to say, pedal spaces, and then before/after compared: sometimes with no discernible difference and at other times surprisingly significant changes.
We also found with careful measurement that the main bike had a 5mm narrower stance (Q-factor) in addition to being strongly asymmetric (5-6mm). By using a long-spindle (+4mm) DuraAce pedal plus a ~1mm pedal washer,
I decided not to attempt original 2012 fit: the pre-changes 2012 video shows a knee diving in, body way off center over the bike, hunched back, etc. Looking at it, I don’t know how I rode that way, but it’s clear that I had to contort and warp various body parts to make it work, which after a while surely had a “baked in” effect.
As an aside, I am also not so sure that the non-Shimano cranks are entirely circular (might have some wobble in the vertical plane around the ring side).
The stance width / knee tracking finding provides a plausible explanation with visual evidence why riding my 2nd bike (the Look 595 Ultra) with its wider and symmetric crank** has felt better on my knees (well, almost symmetric: 1mm difference, which we perfected with a ~1mm pedal washer).
Knee tracking is mainly a function of:
- (1) being centered (symmetric) over the bike, and
- (2) a pedal spacing / Q-factor that is proper for one’s knees.
- (3) cleat position (centered and straight on the shoe generally)
Putting together on the bike feel and the video evidence, I make the following tentative conclusions:
- A symmetric crankset is essential to demystifying bike fit (compensating for asymmetry gets confusing). Check an actual installed crankset carefully for centering on your particular bike.
- As my mountain and cyclocross bikes have symmetric cranksets (and a wider Q-factor), it makes zero sense to do otherwise on my road bike.
- Q-factor and pedal spacing can be key. If something is working for you with excellent knee tracking, exercise extreme caution in changing a crankset or buying a new bike that might have a different Q-factor.
And if I went "back", I am changing the crankset in ~2 weeks anyway, which would essentially put me where I am now anyway.
So we determined that the old bike I mentioned had a wider stance by 4-5mm and this seems better. Using long-spindle pedal and a washer, we got the Moots within 2mm of that bike and the left knee tracks beautifully (which is no guarantee of no pain, but surely a wobble in/out cannot be a good thing).
Ironically, right leg does not track as well, but it has not been an issue.
I am persuaded that the old fit just had my body all twisted... and it remains slightly so as the current video shows; this will have to work out over time, but I think it's best for the long term.
So I'm going to ride my "daily" and just give it some time and assume it will work its way out.
* Stance width is the distance between the pedal. It is actually more than just “Q Factor” (crank-arm width) since it includes the shoes and cleat position.
** Some care is needed when measuring pedal offset because one cannot assume centering of bolt holes in the frame, or even symmetric tubing (especially with carbon fiber frames near the crank area). Also, do NOT assume that quoted nominal Q-factor is accurate to the mm.
Matching the stance width
We found with careful measurement that the main bike (Moots Vamoots RSL + Cannondale Hollowgram) had a ~5mm narrower stance (Q-factor) in addition to being strongly asymmetric (5-6mm).
By using a long-spindle (+4mm) DuraAce pedal plus a ~1mm pedal washer on the left side, it was possible to closely match the stance width and symmetry to the DuraAce crankset on the 2nd bike (Look 595 Ultra, the one that felt better on my left knee).
The two bikes are now both symmetric for pedals as well as within 2mm of stance width. Note that stance width on mountain bikes is on the order of ~15mm wider, so the +2mm difference on the Vamoots RSL is presumably favorable.
With this change, the knee tracking is so close that it’s hard to clearly say that one is better than the other.
Measurements here are to specific places on the pedal (for ease of measurement) which might be slightly different than official from/to figures for Q-factor.
Vamoots RSL with Cannondale Hollowgram SRM crankset:
145mm stance width
+4mm longer pedal spindle
+ 1.5mm left washer
Look 595 Ultra with Shimano DuraAce SRM crankset
147.4mm stance width
+ 1mm left washer
Will this setup help resolve the knee-pain issue? It had better, because what can one go on other than seeing fluidity of movement through a range of motion. But I have to keep my training levels low (for me) as such things still take time for the body to adapt, to “untwist”, etc. My fingers are crossed.