Improving 40mm Marzocchi Forks
When I first got my TR280i, I thought it's 40mm Marzocchi cartridge forks were a huge improvement over the damper-rod forks used on my older trials bikes. But after only one ride on my Montesa 4RT, I realized its 39mm Tech forks were better and wanted to make the Marzocchis behave similarly. Although I did not realize it at the time, the Marzocchis were beating up my wrists.
My first impression was that the 39mm Tech forks were quite “soft” compared to the Marzocchis and I assumed the spring rates were different. Although that is true, the difference was less than I had expected. The Montesa has about a 6% lower spring rate, but the oil is a lighter viscosity as well. Ultimately, I made several changes to the Marzocchis and am now completely satisfied with them.
Consider this general information and not necessarily specific recommendations. Where I do make mention of specifics, bear in mind I am relatively new to trials, weak, elderly, and weigh 135 pounds (61 kg).
There is no one-size-fits-all suspension setup. If it's not right for you, it's not right. But if you do not have much experience with trials, you may not know what is right. Personally, having gained experience tuning road race, supermoto, and motocross suspension, my trials bikes were initially considerably over-damped.
I will not go into the disassembly procedure here as the Hell Team has produced a very good document on the topic (although, there is one point I would suggest doing differently and will mention later).
Similarly, a thread on Advrider has some good information (but you will have to wade through it like everything else on the internet).
Initially, I thought it was possible the Tech forks had more travel than the Marzocchis and that is why they could be softer. But I now think modern trials bikes all probably exhibit about the same amount of travel, as that is partly what defines their geometry.
According to a GasGas specification sheet, the 40mm Marzocchi aluminum forks have 177 mm (6.97") of travel. Maybe that is true of some forks, but on my OSSAs the dust seals hit the bottom of the triple clamp after about 6.5" of travel.
When I measured the fork with the springs removed, I observed only 165mm (6.5") of travel (which is exactly the same as my 39mm Tech forks).
Note that there may be undocumented differences between the various model years. The Marzocchis on my 2011 bike are slightly different from those on my 2014 bikes.
You can learn a lot about your suspension by placing a zip-tie around one fork tube between the dust seal and the bottom of the triple clamp. Firstly, you can get a rough idea of the laden sag. Secondly, you can learn how much travel you have remaining at full compression (as measured from the zip-tie to the underside of the bottom triple clamp).
My zip-tie showed the Montesa was using more fork travel. After riding as I normally do, the OSSA measured 41mm from the zip-tie to the bottom of the triple clamp. For comparison, the Montesa had only 25mm of travel remaining unused. In a perfect world, you would use all your suspension travel (and nothing more) during a worst-case scenario. In an imperfect world, it is nice to have some travel in reserve for when things do not go right – bottoming your suspension can be quite painful!
While you have the forks off the bike, it is customary to replace the zip-tie with a classy O-ring.
I use an O-ring on each fork tube – I will call them long-term and short-term. The long-term O-ring allows me to record the maximum fork travel observed for a practice session or event. The short-term O-ring is useful to see how much travel any given obstacle or section requires. This O-ring can also be used for measuring sag.
Laden Sag & Preload
Laden sag is how much the suspension compresses with just your body weight (wearing riding gear) on the bike.
Preload is the amount the fork spring is initially compressed when installed in the fork tube. For a given spring rate, the more preload you use, the less the suspension will sag. Preload does nothing more than change the ride height of the bike. It is used primarily to accommodate riders of differing weights, but it only provides a limited range of adjustments. If you can not achieve the desired sag, it is an indication that a different spring rate is required.
A change in preload (and thus laden sag) can be affected by altering the length of a solid spacer above the spring, and/or via a threaded adjuster in the fork cap.
Measuring sag is at least a 2-person job. Paul Thede's (Race Tech) method is the most accurate I have seen as it accounts for friction in the fork seals (or linkage if you are measuring the rear).
Sag is necessary to deal with dips in the terrain. Forks compress as the result of bumps and extend as the result of dips. Sag is generally set to about 25% of suspension travel for off-road bikes. But this is no hard/fast rule, as it is partly affected by rider preference and riding style. Twenty-five percent of 165mm is 41mm.
Note that the rear sag on a trials bike is different from other types of bikes. Öhlins recommends 85 – 90mm. This equates to 45 to 50 percent of the rear travel on a typical trials bike. Personally, I find this works pretty well.
A small (a few millimeters) change in sag can have a pronounced effect because it also changes the bike's geometry which can affect traction. See the section entitled Rear Axle Position for some related information.
I use a Baxley front wheel chock to hold the bike upright while measuring sag. This eliminates the need for a 3rd person and may lead to more consistent results. Because this type of stand holds the front wheel off the ground slightly, you should elevate the rear wheel by the same amount to keep the attitude of the bike constant.
Because body position affects the sag numbers, strive to keep your body in the same (and relatively neutral) position for best repeatability.
Only the right-side Marzocchi fork (red cap) contains a preload adjuster. It requires a 22mm tool to change the setting. The preload adjuster has a range of 6mm. The thread pitch is 1mm, so each full turn changes the preload by 1mm.
If you unscrew the fork cap using the 22mm hex adjuster, the preload gets set to the minimum. At minimum preload, I had less than 25mm of static sag with the stock setup.
The stock plastic Marzocchi preload spacer is 96mm long. It has an OD of 29.84mm and an ID 25mm. Because the stock fork setup yielded insufficient static sag for me, the first thing I tried was to shorten the compression-side preload spacer by 10mm to an overall length of 86mm. This was an improvement, but I knew I could do better by changing the spring rate.
Two-part Marzocchi preload adjusting fork cap
The stock Marzocchi fork springs are part number 5141418/P. I could not find any OE alternate rates or any aftermarket supplier of suitable springs. Although it is possible to have custom springs wound, I did not really know what rate I wanted and at a cost of $175+ per try, I sought a different solution.
The Marzocchi forks have a spring in each leg, whereas the Tech forks use a single spring. Since I have never seen a spring rate published for either of these forks, I was forced to measure the dimensions of the springs and calculated their rates with a spreadsheet.
The design formula for a wire spring is: k = G*d^4 / 8*D^3*N
k = spring constant
d = diameter of the wire
G = a constant equal to the shear modulus of the material (depends on metric or Imperial units)
D = average coil diameter
N = number of active coils
Being raised to the 4th power, the diameter of the wire has the strongest influence on the spring rate. (A manufacturer of springs once told me that they measure the diameter of each batch of incoming material to 0.0005" when calculating the number of turns required.)
When reverse-engineering an existing spring, D = OD - wire diameter. But when winding a spring on a mandrel, D = ID + wire diameter. This value is also quite influential as it is raised to the 3rd power.
Being a linear term in the equation, the number of active coils is less critical. When calculating a spring's rate, the number of active coils is generally taken to be about 2 fewer than the total number of coils (including partial turns).
Note that, although it may be counter-intuitive, the more turns a spring has, the softer its rate. This will be important in the next section.
The stock Marzocchi fork springs have a 0.149" (3.785mm) wire diameter, OD of 1.2135" (30.8mm), comprising 26.75 turns, with an overall length of 303mm. This calculates to 23.26 pounds-force/inch or 0.415 kgf/mm or 4.07 N/mm. We need to double this value to account for both springs (which gives 0.83 kgf/mm).
I did the same calculation for the Tech forks (which only have a single spring) and got 43.7 pounds-force per inch or 0.78 kgf/mm. This was my target rate.
Unfortunately, the OD of the Marzocchi spring is 31mm which is very uncommon when looking for a substitute spring from another bike. In fact, the only readily-available spring I could identify with that OD is the 2018-2019 YZ65. However, a wide range (2002 – 2011) of KTM 65s use a slightly smaller 30mm OD spring. And, just over the limit are YZ85 / KX100 springs with a nominal 32mm OD.
The ID of the Marzocchi tube is 31.75mm which gives a bit of latitude. Due to a stroke of great luck, I had a pair of springs from a YZ85 among my spares which showed an OD of 31.4mm and a rate of 0.32 kgf/mm. I used one for the experiment described below.
Springs in Series
My solution to soften the front end was to place two springs in series – the stock Marzocchi spring plus a small section I cut from the aforementioned YZ85 spring. The easiest thing to do was simply replace the stock preload spacer with a spring on one side.
By my calculations, this was slightly softer than my target spring rate (that of the Montesa 4RT) but it was easy to do, and I could always shorten the spring (which would increase its rate) and make up the length shortfall with solid preload spacer. However, my first attempt worked really well and I had no desire to change it!
The formula for springs in series is, “the reciprocal of the sum of the reciprocals.” I remember this because it is the same as the formula for electrical resistors in parallel. When working with only two springs, the equation simplifies to: combined rate = (rate1 * rate2) / (rate1 + rate2).
Series spring adjacent to stock preload spacer
The viscosity of the fork oil has a significant effect on damping. If you are used to tuning suspension for other motorsports, you may be surprised how little rebound damping is needed for trials.
GasGas says to use an SAE 7.5-weight oil in their Marzocchi forks. Unfortunately, there is a lot of variability between manufacturers in what they consider to be a 7.5-weight oil. I strongly suggest that you select fork oil based on its kinematic viscosity (which is measured in centiStokes) instead of by its rated weight. Peter Verdone has a fabulous table of many readily-available fork oils along with their kinematic viscosity. Forks run cool, so all I care about is the kinematic viscosity at 40 degrees C.
For my first fluid change, I was told by the US OSSA tech support guy to use Golden Spectro Cartridge Fork Fluid (125/150, Very Light). Peter Verdone's chart rates this oil at 26.1 cSt @ 40C.
My Montesa manual says to use “KYB 01” oil in its Tech forks. This oil has a kinematic viscosity of 15.6 cSt @ 40C. This is significantly lighter than 26.1 cSt. I wanted to try lighter fork oil in the Marzocchis, but the KYB 01 is difficult to obtain so I substituted Maxima 85/150 (classified as a 5 weight) which Peter's chart lists as 15.9 cSt @ 40C.
The apparatus I use to characterize oils is shown in the photo. It really is only intended to give a relative comparison of two oils at the same temperature. It is often useful to compare an unknown oil against other oils I have on hand.
Timed flow tester at fixed ambient temperature
The air gap (from the fork oil level to the top of collapsed fork tube) determines the air-spring rate. The air spring is in parallel with the wire spring. It is nonlinear and has the greatest effect near the end of the stroke (when the air is compressed the most).
The standard air gap (as specified by GasGas) is 160mm. This is measured with the fork fully collapsed and no spring installed. It is impossible to have more than about 235mm of air gap.
I suggest that you try riding with the air bleed screws (in fork caps) removed (initially one, and then both) to understand the maximum effect the air spring can have. Be forewarned, this can be messy. The compression (left) side will spray more oil than the rebound (right) side.
An example chart from Öhlins is pictured. It has nothing to do with trials forks, but does show the general trend of decreasing air spring force with an increase in air-gap.
Credit: Öhlins. Typical air spring behavior, example only
Seals & Stiction
Seal stiction can be determined by the Race Tech method of measuring sag. Having the minimal possible seal stiction is desirable. SKF's aftermarket seals are claimed to minimize stiction. They are green in color. The part number for the 40mm Marzocchi fork is KITG-40M.
The SKF oil seals are marked 40 x 52.25 x 9.8
The SKF dust seals are marked 40 x 52.45
The Marzocchi clickers have 6 detents per revolution. The maximum number of clicks varies somewhat depending on how the fork cap was assembled but is in the neighborhood of 22 clicks.
Rebound damping “controls” (dissipates) the energy stored in the spring. A heavier rider will require more rebound damping than a lighter rider because more energy is stored in the spring that must be dissipated before another bump is encountered.
Compression damping affects how much travel is used when hitting an obstacle or landing after an obstacle. I use the sliding O-rings previously mentioned to see how much fork travel any given obstacle requires. If you feel too much travel is being used, try increasing compression damping. If the terrain is beating you up badly, try decreasing compression damping.
Minimal rebound damping is required for hopping. Conversely, you may not want a pogo-stick front end for some obstacles and a compromise may be needed. My test section is a stair-step rocky climb. Too little rebound damping will cause the front wheel to ricochet off each step, which makes the bike difficult to control.
In general, you should use the least possible damping (consistent with the other aforementioned considerations).
Fork Tube Position in Triple Clamps
Sliding the forks up in the triple clamps affects handling, turning, and traction. There is an approximately 10mm range for this adjustment. Having the caps flush with the top of the clamp is generally considered the maximum extension. And, you must always verify that you have at least the fork travel (165mm in my case) between the dust seals and the underside of the bottom triple clamp.
Rear Axle Position
The farther the rear axle is moved rearward, the more the load is shifted to the front of the bike.
I have noticed something curious about my OSSA regarding gearing. I prefer riding in 3rd gear with a 2-tooth smaller than standard countershaft sprocket. Curiously, this is a nearly identical (within 1%) overall gear ratio compared to the standard front sprocket in 2nd gear. Now, a 1% gearing change is nearly impossible to feel on a trials bike, and yet the bike behaves differently. I attribute this to the position of the rear axle, which is moved rearward with the smaller front sprocket.
Rear axle position can have a cascading effect on other tuning variables. Moving the rear wheel rearward lengthens the lever arm acting on the shock. This in turn makes the shock spring feel softer to the rider. (Because my bike “works better” in this configuration, I decided to install a softer rear spring.)
Without a change in preload, a softer rear spring increases rear sag which changes the bike's fore/aft weight distribution (and therefore traction). These are subtle, but noticeable, changes and often a rider will subconsciously compensate with a change in body position.
There is no need to tape over the small hole in the rebound cartridge (as described by the Hell Team) when refilling with fork fluid.
Additionally, there's no need to fill the cartridge more than about 50mm from the top as there is that much stuff (floating piston, spring, etc.) that needs to go on afterward.
Below is an exploded-view diagram of the Marzocchis for reference.
Suspension Education Program
The Suspension Education Program is a concept attributed to Andy Ibbott who is an instructor for Keith Code's California Superbike School. The idea is to make test laps with your suspension totally screwed up (but in a systematic and controlled way) and then make notes (mental or otherwise) about what the suspension feels like. Then, when you get a specific bad feeling on the track, you will know which knob to twist to fix it.
I think this concept is applicable to all forms of motorsports, including trials. A simple place to start is with extremes on the front suspension clickers. Ride with caution, as the changes can have a significant effect on how the bike behaves! Try the following:
Minimum compression and minimum rebound damping
Maximum compression and minimum rebound damping
Minimum compression and maximum rebound damping
Half compression and half rebound damping
Of course, you could do the same with the rear suspension as well.
Stanchion tubes protrude through the top triple clamp about 2mm.
9 turns of a 0.32 kg/mm YZ85 spring in place of stock preload spacer (this gives a combined rate of 0.38 kg/mm for each side)
Maxima 85/150 fluid with a 160 mm air gap, 7 mm preload spacer
Compression clicker: 13 clicks out from fully seated
Maxima 85/150 fluid
170 mm air gap, stock preload spacer
Rebound clicker: 20 clicks out from fully seated
Minimal (zero turns from no preload)
This gives about 45mm of sag in front (which could be decreased via the preload adjuster) and about 25mm of unused fork travel under normal riding conditions. The rear sag is set at 91mm.
Increasing a Spring's Rate
I needed a lower spring rate, but what if you are a heavier rider and need to increase the spring rate? Simple – just remove coils. If you want to make the spring 10% stiffer, remove 10% of the active coils and make up the difference in length with additional solid preload spacer length. However, you must also consider the problem of “coil-bind.” That is, you must not shorten the spring so much that the space between the coils is less than the fork's travel.
Another important point is that because the wire diameter has such a large effect (4th power) on spring rate, do not measure over a painted spring!
I have never modified the rear spring from a monoshock bike as they are pretty large and difficult to work with – but the principles are the same. I would not hesitate to shorten a rear spring from a twin-shock bike, however.
My technique for shortening springs is as follows: I use an abrasive cutoff wheel to cut the spring. An oxy-acetylene cutting torch is another option. A hacksaw would be my last resort. Then, I heat the end coils with a MAPP-gas torch and collapse them with large pliers and sometimes the aid of a vise. A bench grinder is useful to smooth the cut end. Finally, a belt sander is used to finish the cut end of the spring.
Italian Quality Control
I would like to think this is an isolated incident, but I doubt that it is, based on my experience with Aprilia. From the factory, there was an extra floating piston in one of my rebound cartridges! The damping action of that cartridge alone (out of the fork) was not at all smooth and it proved difficult to disassemble.
Marzocchi assembled this rebound cartridge with an extra floating piston.