The development of the Jab derailleur began in early 2024. It was born out of lessons learned on our Missing Links for SRAM derailleurs, as well as issues we frequently encounter with derailleurs from SRAM and Shimano. At some point, we realized we could solve many of the issues by creating our own derailleur to empower riders and home mechanics to have more control over their equipment.

The name "Jab" is short for Jabberwocky - a legendary trail in our hometown of Ashland Oregon. Like many of the Ashland trails, the Jabberwocky name comes from the Alice's Adventures in Wonderland story. 

The development of the Jab has been an incredible journey, with plenty of trial, error, and hard-learned lessons. Not every idea has made the cut, and we've given everything we have to stay true to simplicity and durability.

We’ve put the Jab through its paces - testing, abusing, and improving it with every iteration. The result is a derailleur that delivers precision, performance, and serviceability like no other.

The derailleur parts are CNC machined by our manufacturer in Dongguan China. We've worked with this supplier for several years. They are high tech, have great communication and great quality.

The parts are inspected at the supplier before they ship out, and then again when they arrive in Ashland. Derailleurs are assembled in Ashland, and shipped out to customers from there.

Initially, we are focused on a CNC machined version of the Jab in order to make a precise, high end derailleur where we have flexibility to modify tolerances and shapes in order to refine the design. However, we also intend to launch lower cost, tooled versions as the company grows. CNC machining gives us the ability to compete with other high end cable actuated derailleurs like XX1 or XTR and deliver more benefits. We will always aim to strike a good balance of value and performance.

The core goals of the Jab are:

1. Rebuildable / serviceable - You can adjust, service, and rebuild it easily with common tools.
2. Compatible / modular - Mates up perfectly to different 1x drivetrains, MTB or drop bar, 11 or 12 speed, and allows you to convert it later.
3. Durable / precise - Confident, crisp shifts even after a lot of use. And when isn't dialed anymore, you can rebuild it.

The Jab is cable actuated (not electronic) for several reasons:

1. Reliability. While cable actuation is not immune to failures, dead batteries are pretty common, and who wants to be worried about battery charge levels in the derailleur and shifter during an important race or ride? 
2. Shifting speed. Fast and crisp, mechanical actuation allows you to push through multiple gears with your fingers, or click rapidly, and there is less delay/hesitation compared to the electronic options from the big drivetrain companies.
3. Feel. Cable actuation gives the rider intuitive feedback at your fingers so you know what is going on back at your derailleur and cassette. You can tell when the shifting is perfectly tuned, or has hesitation. You can push through multiple gears. This experience is similar to driving a manual transmission vehicle vs an automatic. You are one with your machine.
4. Development cost. To develop an electronic drivetrain that seriously competes with existing electronic drivetrains would require significant resources. Madrone Cycles is still a small company, so we are not ready for this yet.
5. Environmental impact. Electronic drivetrains have batteries and printed circuit boards in them. These contain hard to recycle materials and sub-assemblies. 

Version 1

Revealed at Sea Otter 2024, and tested March 2024 through July 2024. This version was made simply to see if there was any potential to make a competitive derailleur. It was clear we were on to something interesting, but it also became clear that derailleurs are deceptively complex, and need to be well tested to be reliable.

Tested:

• Hydraulic clutch: Eliminated due to complexity.
• Ball bearing pivots: These made it to 2nd version of the derailleur, but we improved how they were fixed into the links.
• Large cable pulley: Reduced diameter later because it was unnecessarily large.
• Carbon fiber cage: Eliminated due to non-recyclability of material.

Version 2

Revealed at the MADE Show 2024, and tested August 2024 through November 2024. A few units were made for testing, each being updated with revised components throughout testing. 

This version addressed things we learned on V1, and took a step forward in simplicity, and we made sure to steer clear of existing patents. 

We were pleased with the overall performance of the shifting and clutch. Setting the clutch up stiff is great for rough terrain. We found and addressed various small issues through testing, such as set screws coming loose, difficulty cleaning some areas, the cable pulley rubbing, and cam interference in the small cog. This version of the derailleur was overbuilt and heavy, but served its purpose well.

Tested:

• Flexible cage: Eliminated due to complexity and limited benefit. We tested several cage designs on these prototypes, all with various degrees of flexibility, and shapes.
• Ball bearing pivots: Improved over prior version, but the difficulty of replacement, as well as contamination and weight issues led us to drop the ball bearings and go with an advance bushing design for the 3rd version of the derailleur.
• Adjustable clutch: Worked great, but further simplification was possible.
• Ratchet clutch: Worked well.
• Nested linkage return springs: This makes the return force nice and high, even for the small cogs. This also allows us to design springs that will not fatigue and lose their effective stiffness.
• External stiffener plate mount: Worked great. Overall mounting bolt internals needed to be further simplified though.
• Multiple cam profiles: SRAM, Shimano, and Shimano shifter with e13 cassette.
  

Version 3

Revealed November 2024 in press releases, and tested through March 2025. Several units were made for final testing, tolerance refinement, user experience, and color evaluation. This is the version where we committed to produce the Jab for customers by opening the pre-order window from late November 2024 through early January 2025. We closed the pre-order window because the volume of orders was getting too high for us to manage.

This version was intended to be a final prototype, and was the culmination of all prior testing, distilled into the simplest form we could. We had a lot of communication with our manufacturer on tolerances, machining processes, anodizing, etc. The testing was more focused on refinement and final details, rather than exploring or learning new things. 

During the testing of this version, we encountered numerous unexpected failures in the design, initiating the phase of the development where it seemed like the universe was out to punish us for attempting to make a derailleur.

Tested:

• Tolerances: There are numerous critical fits, and all of them affect performance, we are refining each of them for good, repeatable performance, and tolerances that are manufacturable).
• Bushing pivots: We have a ton of experience and hard lessons learned on the Missing Link bushing pivots that was applied to the Jab. We have our own unique bushing design, which helps maintain tight tolerances and easy service with common tools. This also reduced weight, give longer life than ball bearings, and still allow smooth rotation. 
• 2x and 3x cages: Unfortunately the body of the derailleur will not support 2x and 3x, so that will need to be a different version of the Jab, which will be postponed until we catch up after the 1x launch in April. This was a very difficult wall to hit late in the development, but we had to set it aside and move forward with the 1x.
• More cams: We now have a cool in house cam development process, and are continually adding to the list.
• Ratchet clutch engagement: We got it down to 2 degrees using 2 pawls and wire EDM cutting of the ratchet and pawls for precise profiles.
• Bi-directional clutch friction: We found that another simpler clutch design works very well for most riding conditions, and promoted the bi-directional friction clutch to the standard clutch.
• Strength and weight: The weight of this version was a low as 270g at one point, but we ended up adding material back into it to improve the durability.