When is a Stem Too Short?
It turns out that stems can be too short. Fabien Barel was on the ground floor of rider-forward geometry while he was riding for Mondraker back in 2012. I remember when he rolled out on a pre-production Foxy that had a massively long top tube, a steep-for-the-time seat tube angle and a crazy, 15mm-offset stem. I peppered Barel with questions about the stem while we rode together on his home trails near Les Gets, France.
When asked why Mondraker stopped at 15 instead of making a zero-offset stem, the former World Champion was candid: “We tried zero offset and the steering was too unstable. Fifteen is as far as we can go, and maybe twenty is better.”
Today, stem lengths have crept back from 50 to 35 millimeters (20 if you add Pacenti’s P-dent stem and bar combination). I was reminded of Barel’s discovery after I switched handlebars and experienced a similar instability. The new handlebar had similar numbers, but I would later discover that, in conjunction with the bike's 35 millimeter stem, my hands were in line with the steering axis – a position, I assumed, that approached Mondraker’s zero-offset stem. Curiosity led me to measure a number of stem and handlebar combinations and to the discovery that the brand of handlebar you choose, and the angle that you set it can alter the effective length of your stem by over 50 percent.
Kirk Pacenti's PDent bar and stem combination offers stems as short as 15 and 20 millimeters.
Working with the theory that a mountain bike’s steering is somehow stabilized when the rider’s hands lay ahead of the steering axis, I chose a handlebar with a different bend that did just that, and the bike’s steering returned to normal. The difference in hand position was only ten millimeters.
Oddly, I raced motorcycles in various disciplines, and almost every handlebar put my grips at or behind the steering axis. No instability issues there. The difference seemed to be that you grip a moto with your legs, which isolates your body from steering inputs. While riding a mountain bike, however, your contact points are most often the handlebar and pedals. Most of the time, there is forward pressure on the grips, so it stands to reason that having the grips in front of the steering axis would cause the steering to self-center, which would become a stabilizing force. Fabien Barel is a smart guy.
What the Handlebar Has to Do With This
Gemini's Kastor one piece carbon handlebar illustrates how the bar and stem operate as one simple lever. Gemini photo
Before we get any further into this discussion, think of the stem and handlebar to be one welded structure, instead of two components. In reality, your hand position, relative to the fork’s steerer tube are the only components in the equation that matter. Imagine three points of a triangle, two at the middle of each grip and one at the center of the stem’s headset adjustment bolt. That’s a very obtuse triangle, and now that 50 millimeter stems are considered to be long, there are less than 30 millimeters of wiggle room for rotational adjustments and sweep-back angles before those three lines converge into Barel’s danger zone. Geeking in the Basement Workshop
Just for fun, I amassed a number of handlebars for a test. I made a table/fixture that held a fork steerer tube at an imaginary 66-degree head tube angle. I mounted up a 45-millimeter stem and cobbled together some measuring devices to ensure that each bar was set at the same up-sweep angle of five degrees – the average angle that most bar makers (and riders) prefer. Some lock-on grips allowed me to measure each bar at a 780 millimeter effective width for consistency. I made reference marks at the center of each grip and the stem bolt, and then stretched a length of gear cable between them to ascertain the distance that the hands were positioned ahead of the steering axis (center of the stem's headset-adjustment bolt). Here’s what I discovered:
Rules: All handlebars were measured from the center of the grips (66mm inboard) with the outer ends of the grips at set at 780 millimeters wide. The upsweep for all bars was adjusted for the first part of this experiment at five degrees. The Race Face Atlas was recommended at 4 degrees. All others (where stated) were listed at five degrees. The Easton Havoc and Race Face Turbine 2 were chosen because they had 35mm clamp sections. A 40mm stem was used and their effective lengths have been noted in the chart. The manufacturer’s back-sweep is stated in the text. The numbers, rounded off to the nearest millimeter, are the distance forward from the center of the steering axis – that’s your effective stem length Handlebar-Adjusted Stem Length
*Denotes 40mm stem, which reduces effective length by 5mm
Syntace Vector Superlight 8
Syntace Vector Superlight 12
Race Face Atlas
Race Face Turbine 2
Gamut Cillos DH
Easton Havoc Aluminum
Effective Stem Length
The takeaway here is that handlebars with similar back and up-sweep numbers often do not position the rider in the same place. Where the bends occur in the span of the handlebar create different effective stem lengths. The differences are not huge, but they can be significant. If Fabian Barell’s observations are correct, and I believe they are, a handlebar-adjusted stem length around 20 millimeters would be as short as you’d want.
These numbers were derived using a 45-millimeter stem. So if you are running a 50, you’d need to add 5 millimeters to the effective stem lengths of each handlebar and conversely, subtract 10 millimeters if you were using a 35-millimeter stem. Using a stem that short would make it easy to arrive at a bar/stem combination that would put your grips behind the steering axis, and that's probably not good.
Using the “string” measurement technique that I demonstrate here can help you determine whether your handlebar and stem combination is in the ballpark, or if that combination warrants experimenting with a different length stem. You might be surprised to discover that your 50-millimeter stem is tinier than you believed it to be.Stem-length Adjusted for Handlebar Angle
How much you rotate your handlebar to suit your hand position affects the calculated length of your stem to a much greater degree than the particular bend configuration of your bar. Twenty-millimeter rise bars are popular these days, so I used the Race Face Turbine 2 to demonstrate. I rotated the bar to produce seven, five, and three degrees of up-sweep. Here's how it came out:
Handlebar: Race Face Turbine 2
8º back, 20mm rise, 5º up
Rotated to 7º up = 36mm
Rotated to 5º up = 22mm
Rotated to 3º up = 15mm
You can see that when you rotate your bar to achieve your sweet spot it can dramatically change the effective length of your stem. That change would be greater with a higher bar like the 30mm-rise Renthal or Gamut.
So, What Does All This Mean?
First and foremost, if you always run the same brand, model and width of handlebar, and are careful with your setup, then switching to a different length stem should provide instant and clear feedback. You’ll either like it, or you won’t, and you'll know the cause.
What this experiment suggests, is that the trend towards shorter stems has narrowed the range of rotational adjustment and hand positions, and that makes it easier to fall outside of stable parameters
and end up with undesirable steering qualities. Also, the particular bend of each handlebar maker’s design should be considered as a factor before you settle on a stem length. Shorter may not be better in some cases, and slightly longer might be a stabilizing factor.
Here is a simple tool that you can use to assess your starting point and, should you choose a different model of handlebar, or entertain a change in stem length or rotation, you can assess the ramifications in a minute or two with a piece of string and a ruler. If your set-up is golden, you’ll be able to reproduce it with any number of bar and stem combinations. If it's not, you can determine why and make positive adjustments towards a solution.