The Dragonfly's forks are manufactured by Metric 4 in Italy. The name M4 Racing Parts is also used in the maintenance manual. (Eventually, I noticed the apostrophe in their name and henceforth will write it that way.)
Forks allow a lot of room for creativity among manufacturers. This fork has some idiosyncrasies. For example:
Two different viscosity fluids are used.
The fastener at the bottom of the fork legs is a left-hand thread.
A pin spanner is required to remove the fork caps and also to disassemble the forks for seal replacement.
A puller is required to remove the fork bushings. The usual “slide hammer” technique won't work.
There does not appear to be a top-out spring. That function may be controlled by hydraulic action.
Surprisingly, the M'4 forks are about 10 mm longer (axle center to top of fork cap) than Tech's aluminum forks. Because the Dragonfly is so diminutive, I expected the M'4 forks would be shorter, if anything.
Only the left side fork tube contains a spring (M'4 part number 402A30275). No optional spring rates are mentioned in the user manual. This separation of springing and damping duties allows the damping fluid to remain much cleaner.
For reference, another fork produced by M'4 uses an air spring on the left side instead of the wire spring. This fork was used by EM in 2020 and is described under the Epure Race's Mechanical section heading on this website.
Although the M'4 Maintenance Manual is the definitive guide, some additional information is summarized below.
Two different fluid viscosities are specified in the service manual. The majority of the fork fluid is specified as an SAE 5-weight. The M'4 part number for this is 402A0500.
Additionally, a small amount (~30 ml) of SAE 20-weight fluid is introduced separately on the damping (right) side via the bottom into a separate sealed chamber. No part number is given for this fluid.
Fork seals are part number 402A5239. (Previously, I reported hearing that Tech fork seals fit, but they will not.)
Upper fork bushings are part number 402A4339.
Lower fork bushings are part number 402A4038.
A left-hand thread used at the bottom of both forks. It requires a 10 mm hex key.
Based on the Hell Team's excellent Dragonfly video, a Park Tool SPA-2 (red) pin spanner works for removing both the fork caps and fork seal retaining ring.
Left side, Preload adjuster
The complete left side fork assembly weighs 2270 grams (estimated because I removed about 100 g of fluid from mine).
The Preload adjustment provides a range of about 15.5 mm (1 mm per turn). It requires a 6 mm hex key.
The specified fill is approximately 325 ml of SAE 5-weight fork oil. I don't think this viscosity is at all critical, since no damping is performed in this fork.
The standard oil level is 75 mm, but this provides a very stiff air spring. See the Air Spring section below for more information.
The complete right side fork assembly weighs 1883 grams.
The specified fill is approximately 275 ml of SAE 5-weight fork oil. The oil level is specified as 140 mm. The kinematic viscosity of this fluid is important, and I will discuss it later.
The Compression adjustment at the top of the fork has a range of about 4 turns (5 clicks per turn). It requires a 3 mm hex key.
The Rebound adjustment at the bottom of the fork has a range of about 6 turns (10 clicks per turn). It requires a 3 mm hex key.
The axle clamp screw requires a 5 mm hex key.
Right side top, Compression adjuster (3 mm hex key)
Right side bottom, Rebound adjuster (red 3 mm hex key)
M'4 specifies a 5-weight fluid. Since there is considerable variation from one manufacturer to another regarding what exactly constitutes a 5-weigh oil, I prefer to go by kinematic viscosity.
But when I don't know the kinematic viscosity of the original fork fluid, I must make a relative comparison with a known fluid. This comparison is only valid if both oils are at the same temperature.
The bottom line is that the fluid I got out of the M'4 right fork was identical to Maxima's 85-150 (5-weight) which has a kinematic viscosity of 15.9 cSt @ 40° C.
The kinematic viscosity of the used fluid on the left (spring) side was slightly higher, and much darker. Both changes are consistent with the presence of debris in the fluid.
Note that a small amount of 20-weight fluid is additionally specified for the right side, but is held in a separate chamber filled via the bottom (see M4's manual for instructions).
Timing flow to judge relative viscosity
I'm calling this the “Bottoming Chamber” because it's not as sophisticated as a typical hydro-stop. There is no valving per se. Its purpose is to prevent metal to metal contact at the bottom of the stroke.
The manual says you will get 25 ml of fluid out, and to refill with 30 ml of SAE 20-weight fork oil. I got 40 ml of fluid out. The viscosity may not be critical, and I discuss this below.
Note that the photo below shows the left-hand thread that must be undone to completely disassemble either fork. This would be required to replace the seals and/or bushings. But if you are at all hesitant about this step, I'd say it's not critical for routine maintenance. The O-ring sealing surface is somewhat difficult to clean, and the fluid I got out was largely uncontaminated. It is completely separate from the 5-weigh damping fluid. Because fork fluid will escape if you use gravity to assist starting the left-hand thread, I recommend using a wooden dowel inside the fork tube as a pusher. It may be helpful to practice this when there is no fluid in the fork.
The most important fluid to change is the right-side 5-weight, which influences compression and rebound damping.
M18 x 1.25 LEFT-HAND thread (requires 10 mm hex key)
Torque to 20 Nm (14.75 lbf-ft)
“Bottoming Chamber” at bottom of right fork tube
A spiral lock ring holds chamber into fork tube
Because M'4 does not provide a part number or detailed specification for the 20-weigh fork fluid, we may assume it is not critical. Below is a list of the kinematic viscosities (at 40° C) in centistokes for some 20-weight fork oils.
PJ1 61.7 cSt; Spectro 65 (elsewhere 72.2) cSt; Silkolene 67.7 cSt; Castrol 68 cSt; Maxima 71.1 cSt; Bel-Ray 73.5 cSt; Elf 99 cSt; Ohlins 99 cSt; Putoline HPX 99.4 cSt; Agip 104 cSt.
The list shows a wide variation. In fact, several 15-weight fluids fall inside that range and one 30-weight fluid is very close to the high end.
The fluid that came out of the bottom of my forks took 114 seconds to flow through the test burette (shown above) at 65° F.
The most similar fluid I had on hand was Spectro 20-weight (purple tint). It took 110 seconds to flow through the test burette at 65° F.
For reference, the most viscous fork fluid I have is PJ1, 30-weight. It is rated 116.9 cSt at 40° C. It took 168 seconds to flow through the test burette at 65° F.
So my recommendation for the 20-weight is to pick something close to the lighter end of the list above.
M'4 Fork Seal measures 39 × 52 × 7.5 / 12.5
The fork seals are marked “Metric 4 ROLF Z-2 SFRD 39 52 7,5 / 12,5” Rolf is a German company with a production facility in Italy.
The seals presumably utilize two springs (I have not yet removed a seal to verify). No dust seal is used.
The 52 mm OD is different from the Tech's 51 mm dimension. The seal height is specified as 7.5 mm and 12.5 mm (a 5 mm step). I measured the step height as 5.75 mm.
The red retaining ring (just visible in the background) holds the seal in place. For reference, its thread is 53 mm x 0.8 pitch.
If you pull a fork tube completely out of a fork leg (even without replacing the seal), it's probably a good idea to re-grease the seal. I use Race Tech's USSG 01 (ultra slick seal grease), but other greases may well be satisfactory.
My fork spring measuring apparatus is shown below.
It comprises a bathroom scale and a drill press. The drill press is capable of moving 4.75 inches (120mm) so I can only measure the spring's initial rate.
However, the results validated my calculations.
Setup to measure a fork spring's rate
Mecatecno says the M'4 forks have 170 mm of travel. With the forks still mounted to the bike and the spring removed, I could only achieve 158.5 mm of travel. I used 92 mm of fork travel while doing a conservative test ride. I'm certain my wife used even less travel. Compression and preload adjusters were at minimum for the test.
So, as received, the forks are extremely stiff. It's no wonder they work for the US importer, who is a big guy, weighing 220 pounds (ca. 100 kg). For riders who weigh the same (or less) than the Dragonfly itself, the forks can be improved.
I had heard that Tech springs will fit in the M'4 forks. They will, but won't help a lightweight rider. I was quite surprised by my spring-rate calculations and validation measurements.
For reference, the M'4 spring is progressively wound and has an OD of 1.225 inches, a wire diameter of 0.178 inches, and a total of 41.50 turns. Sorry for the US customary units, but that's how my spreadsheet is set up. This gives an initial rate of 31.4 lbf/inch (5.6 N/mm) and a final rate of 40 lbf/inch (7.1 N/mm).
By “initial rate” I mean the rate before any coils go into coilbind. Progressive springs provide at least two different spring rates. As the coils that are wound close together touch each other, they are no longer a spring but form a solid spacer instead. This allows the remaining coils (spaced wider apart) to provide a higher spring rate because there are now fewer active coils. Remember, with other variables held constant, the fewer the turns a spring has, the higher its rate.
This is exactly the behavior we want in a trials fork. The soft initial rate handles small bumps and keeps the fork action compliant for optimal control. But big obstacles create large impact forces and a higher spring rate is necessary to deal with them.
Although my testing apparatus does not have enough travel to measure it, the final spring rate calculates to 40 lbf/inch (7.1 N/mm).
Since Tech springs have an initial calculated rate of about 42 lbf/inch (7.5 N/mm) they would not help a lightweight rider.
As an aside, I have been provided the following information on Tech fork springs that I'll leave here for reference:
Tech No. 090473902 (Prog. 1) 7.2 – 8.8 N/mm
Tech No. 090473903 (Prog. 2) 7.2 – 10.2 N/mm
Tech No. 090474151 (Prog. 3) 8.0 – 12.7 N/mm
Upper spring, Tech “standard” used in EM Epure Race. Lower spring, M'4 used in Mecatecno Dragonfly.
So I turned my attention to the air spring. All forks possess an inherent air spring (by virtue of the trapped volume of air being compressed inside the fork). This air spring acts in parallel with (thus adding to) the wire spring's force. The air spring's rate is tunable via changes in oil volume. More oil yields a stiffer spring. But it's highly non-linear, and a small volume change can produce a large change in force near the fork's maximum travel in compression.
The M'4 manual says the spring-side fork tube contains 325 ml of SAE 5-weight fluid. I never go by volume. I use level instead. The level is specified as 75 mm.
I decreased the level (to soften the spring effect) to an educated guess of 165 mm. It's easier to add more oil if the spring is too soft than it is to remove oil if it's too stiff. This measurement is made from the oil inside to the top of the tube (with the fork tube fully collapsed). Also, remember that the compression damping is adjustable on the M'4 fork, so that gives another tuning variable.
Even after decreasing the fluid level again to 194 mm, the air spring was still too stiff. I'm now trying a fluid level of 250 mm.
I concentrated on the air spring in the left fork tube, but eventually wondered if the right side had an air spring that could be adjusted as well. The short answer is no. The spring function (both air and mechanical) is strictly on the left side, and the damping function is strictly on the right side - a nice separation of duties.
Although the M'4 service manual leaves much to be desired, it eventually dawned on me that the oil level is specified (but in an unfamiliar manner). When describing the fill volume, they have parenthetical remarks specifying “75mm chamber” for the left side and “140mm chamber” for the right side. As a quick review, oil level is measured in millimeters with the spring removed and the fork tube completely collapsed. The distance is measured from the top of the fork to the liquid level inside.
Now here's the surprising part, my oil level measured 142.5 mm with the fork tube fully extended. The adjacent photo shows where this oil level would be relative to the compression valving assembly.
I wondered how this could possibly work, since the chamber above the piston would fill with oil and hydraulic lock before reaching the specified 170 mm travel.
It turns out, the right-side oil level is fairly constant throughout the range of the fork's travel.
The level only changes by the displacement of the 12 mm diameter cartridge rod. So, over the full 170 mm of fork travel, this equates to a volume of 19.3 ml. A 19.3 ml change inside the 32 mm ID fork tube equates to a difference in level of about 24 mm.
Compression valving assembly with a 142.5mm “air gap” shown for reference
Close-up of M'4 compression piston with needle valve for adjustment
Note that there's a steel shim both above and below the compression valving piston. These shims cause fluid to flow slowly through the valve when the forks are compressed, yet bend to provide little resistance to flow when the forks are extending (due to the availability of much bigger holes in the piston during extension).
The adjustment clicker at the top of the fork opens and closes the needle valve, thereby changing the resistance to fluid flow.
With the clicker fully clockwise, the flow is maximally impeded, thus providing the slowest fork movement during compression.
If you don't ride like Oriol Noguera, the fork may be too stiff for you.
I use a zip-tie around one fork tube to measure the maximum travel while riding. Despite the 170mm specification for fork travel, there appears to be only ~160mm available. If you are a lightweight rider and feel the fork is too stiff on compression, remove oil from the spring side.
The ID of fork tube is 32mm. This yields an internal volume of 0.8 cubic millimeters per millimeter of level change. This is equivalent to 1.25mm of level change per milliliter for fluid change.
The adjacent box is an embedded link to my other trials-related website (which is about EFI OSSAs). You will find information that's specific to the tuning of 40mm Marzocchi forks, but there is general information there as well.