Musings and Experiments on the Art and Science of 3D Printing


Everything you wanted to know about the Zesty Nimble remote direct drive (RDD) extruder

By SublimeLayers Tuesday, May 30, 2017
I've been a "delta fanatic" for over 4 years. Like all things in life, delta printers have tradeoffs compared to their Cartesian counterparts. In exchange for very fast movements in all three dimensions, the extruder is mounted remotely and the filament delivered through a long PTFE tube called a Bowden tube, typically 2mm ID for 1.75mm filament.

Delta Musings
The early delta printers were fairly small (Mini Kossel) and the length of the Bowden was not too problematic. But as delta sizes increased - like the SeeMeCNC Rostock MAX and the more recent UltiBots D300VS - the length of the Bowden also increased. Long Bowden tubes have more friction for the extruder to overcome, especially if the extruder roughens the filament surface with its drive gear. Another issue is that long Bowden tubes introduce more hysteresis into filament retracts due to their inner diameter being much larger than the filament diameter. This results in print quality issues like excess stringing and blobbing.

Several years ago I recognized these problems and developed a solution to combat them with a combination of using 1.8mm ID PTFE tubing and Bondtech QR extruders. This combination greatly improved my print quality and is highly reliable (1000s of hours of printing without a single problem). The 1.8mm ID tubing has less issues with hysteresis, the primary contributor to print quality issues. But, any significant filament diameter issues or surface damage from the extruder drive gear results in jamming and filament starvation. This is where the Bondtech QR extruder comes to play. It's dual opposing drive gears exert much greater extrusion force with far less filament damage than the common single drive gear/pressure bearing extruders. The combination was very successful.

Then last year I added an UltiBots D300VS large format (300mm D x 445mm Z) delta printer to my stable. My intent was to build the D300VS stock and then replace the UltiBots Micro Direct Drive Extruder with my tried and true 1.8mm Bowden and Bondtech QR extruder. But after a couple of test prints I was so impressed with the results that I decided to put the extruder through its paces. This experience really punctuated the tradeoff we delta fanatics make with Bowden filament delivery - even with more advanced solutions like mine.

The Zesty Nimble

It was exactly at this time that I first learned of the Zesty Nimble remote direct driver (RDD) extruder. This is a different filament delivery mechanism that places the stepper motor (the primary contributor to extruder mass) remote and powers the drive mechanism via a rotating cable - also a type of Bowden. This has the advantage of removing significant mass from the delta effector while powering the filament delivery right at the entry to the hot end like a direct drive extruder system.

I backed Zesty's unsuccessful KickStarter campaign in October, 2016. KickStarters are hit and miss from a funding perspective and to Zesty's credit, Brian and Lykle persevered to bring their super light extruder to market. I kept in contact with the founders through late 2016 and ultimately they asked if I would be interested in testing the Nimble once they had a pre-production unit ready to test. I couldn't say no to that!

I received the pre-production Nimble on March 16, 2017 just a week before I had to leave to attend the 2017 Midwest Rep Rap Festival (MRRF) in Indiana. I was determined to get the Nimble installed and printing so I could bring it to show the crowd. Little did I know when I received it that mine was the very first (and only) Nimble in the wild.

I decided to replace the Bowden setup on a mid-size K250 delta printer powered by a Duet .8 controller and dc42 firmware (my gold standard delta firmware) with an E3D V6 hot end and standard .4mm nozzle. I have 1000s of hours of printing on this printer so I know it inside and out. Rather than print a new mag ball effector (which would actually have been the effector and all three carriages because I only used matched sets printed at the same time to eliminate dimensional issues) I modified the effector simply by drilling two holes for the Nimble mount.

The Nimble is remarkably small and light (appropriately named!) and mounting it was surprisingly easy as you can see here.
In addition to its diminutive size, the Nimble's design and engineering are remarkable. The extruder is ambidextrous - meaning that it can be mounted either left- or right- handed. And, it features a really convenient "breech loading" system - the red lever you see in the photo below is the breech lock.

And here's a photo of the remote stepper and drive shaft.
Before continuing, here are the basic stats for the production Zesty Nimble:
  • 23.5mm x 39mm x 28.5mm (WxDxH)
  • less than 28gm
  • 1.75mm filament standard, 3mm is possible
  • 30:1 gear ratio

And now for the part I'm sure you've been waiting for - testing!

My 3D printed fly fishing reel is a great torture test for hot ends, extruders and printers. There are 9 printed parts and each has unique challenges to print perfectly. The reel back plate with its 4 thin pillars and central shaft is a very difficult part to print without excessive stringing. I've spent an inordinate amount of time printing and studying this part and I've become very proficient in diagnosing printer issues with it. So, of course, it was the first part I printed with the Nimble.

My initial experience was a little mixed but understandable. With its 30:1 gear ratio, retracts require a little forethought. I typically retract PLA at 20mm/s but this speed stalled the stepper. A few back and forth emails with Zesty resulted in lowering retract to 10mm/s. The gears in my pre-production Nimble were 3D printed and the production gears would be much smoother and allow faster retracts. I didn't have high hopes that 10mm/s retracts would be sufficient to prevent stringing but I forged ahead. Here is the very first part I printed:
A very respectable part that many would be very happy to produce. But, as you can see at the red arrow, there was some stringing between the pillars. I've printed 100s of these with this PLA filament and know its extrusion characteristics inside and out. This was not a temperature issue, it was a retract issue due to the low retract speed. Otherwise the part was great with smooth, even and well registered layers. I printed a few of these in several PLA filaments with similar results.

For my next print I decided on a gear vase of my own design. I planned to print this at MRRF. It is not extremely challenging but stringing can be an issue for long Bowden delivery systems. This part turned out near perfect.

I printed in the neighborhood of 40 parts with this setup before, during and the week following MRRF. I gave a lot of thought to how to mitigate or improve the slow retract speed. The "solution", I thought, might be to use a Duet Wifi and decrease the micro steps from 16 to 2. This would reduce the steps/mm from 3100 (1.8° stepper) to 387.5 steps/mm - very much in line with a good extruder steps/mm range. During this time, Brian was experimenting and tweaking acceleration and jerk settings. He was getting much higher retract speeds - up to 80mm/s. He shared his test results and settings. Here's what he came up with:

extruder acceleration: typically 1000mm/s^2 - reduced to 50mm/s^2
jerk (speed changes): typically 600mm/min - reduced to 500mm/min

I made these changes in my config.g and loaded the reel back plate I use to test. This time I sliced with 20mm/s retract. I'm pleased to say that the Nimble performed perfectly with no stringing or other extruder related print artifacts. I continued to test prints and filaments including NinjaFlex (Nimble is excellent for this difficult filament), ABS, PETG, CF filled PLA, CF filled PETG, BronzeFill and WoodFill. Overall the Nimble performed very reliably and produced some excellent looking parts.

Filament changes are a breeze with the breech loading system. My only minor point on this is the breach lock (the red lever) does require a short break-in period. This won't be noticed by many users but if your delta printer is equipped with magnetic ball arms, you have to be careful in order not to pop off the joints. Not harmful but annoying. But after several dozen loadings, the parts break in nicely and I suppose the user develops a process that works without separating the mag joints.

During this time I've kept a relatively low profile about my Nimble results - not because it wasn't performing nicely but because 1) it was not a production unit and 2) I never endorse any product that I haven't personally put through the wringer with testing, testing and more testing. It's easy to develop a product that gives good results "most of the time", it's much more difficult to develop a product that produces great results with 100% reliability. I'm looking for products with 100% reliability to use on my 3D printers.

The Production Nimble
In late April, Zesty sent me a production Nimble to test - again the first one in the wild. There were a number of changes to the extruder body so a complete new kit arrived on my door step. It was a 10 minute job to change over to the new kit. I've printed over 100 of my fly reels (all 9 parts) and countless (well, at least 150) other models to test the Nimble under a variety of conditions and filaments. I am really happy with it's performance and can speak to its reliability over several months of continuous and heavy printing (5+ hours a day).

Here is a photo of the fly reel back plate printed in Hatchbox translucent blue PLA (one of the more stringy filaments I've come across) yesterday. This is an as-is photo still on the print bed untouched and un-retouched.

This is the level of perfection I seek in all of my 3D printed parts and the Zesty Nimble delivers admirably.

Results in a Nutshell
Nimble Pros:
  • very low mass (<28gms): reducing effector mass is critical for a delta printer. Mass contributes to "ringing" - overshooting corners results in imperfect layer alignment
  • very small (23.5mm x 39mm x 28.5mm (WxDxH)): effector space is typically very tight, especially on smaller delta printers. The petite Nimble can fit virtually anywhere.
  • simple filament loading: the breech loading system gives Nimble one of the easiest filament loading systems available.
  • ambidextrous mounting: the Nimble's design allows it to be mounted in either left or right hand orientations.
  • flexible filament: Nimble is the first really good extruder for printing flexible filaments on a delta printer that I've used. 
A Con or two:
  • removable breech lever: the red breech lever detaches easily from the extruder body. If you are careless, you might lose it or drop it during filament loading.
  • price: although not the most expensive extruder on the market, the Nimble comes in at €85.50 ($95.75 as of 5/30/2017). However, for this you get a very well engineered product that does the job and does it well. If you require low mass, small size and the ability to print flexible filaments, it is an affordable option.
Final Remarks
It's taken Zesty longer than many would like to finalize production of the Nimble. But to the company's credit, rather than rush an product not ready for prime time out the door they took their time to refine, test and deliver a well engineered and produced extruder. I give them a lot of credit for that - I know it was frustrating and expensive for them.

Along the way to production, Zesty has developed mounts/adapters for over a dozen printers for their single and dual Nimble extruders and are committed to developing adapters as needed. They've also refined their documentation for assembling, calibrating, tuning firmware and caring for the Nimble. 

The Zesty Nimble produces excellent results with a variety of filaments including flexibles. Flexible filaments are the bane of Bowden tube delivery so not many delta owners have been successful printing them. The Nimble will change that. 

Although my pre-production and production Nimbles were provided by Zesty, I remained unbiased throughout testing and provided critical feedback to the company. Frankly, I test so many products that I purchase or that are sent to me that I do not feel obligated to provide a good review if the product does not perform remarkably. I have very high standards for quality, reliability and my integrity and I evaluate everything I test to those standards. The Nimble met or beat all of my expectations. 

I am so impressed with the Nimble that I just purchased (full price) the Dual Nimble Upgrade kit so I can test it with the E3D Cyclops dual in-one out multi-filament hot end. 

24 comments to ''Everything you wanted to know about the Zesty Nimble remote direct drive (RDD) extruder"

  1. Really good to hear that the Nimble is the real deal - I have one being shipped to me. Question: How would you compare the Nimble results with your E3D Aero setup?

  2. I'll be posting reviews of the E3D Titan Aero and the Bondtech BGM direct drive extruders soon. These are all good extruders but each has it's place. You couldn't use the Aero or GGM on a mini Kossel for example. The Aero is very cost effective with its built-in hot end and the short path from the extruder to the hot end has some advantages. I really like the Aero. That said, I also really like the BGM with its dual opposing drive gears. It can push filament through a Cyclops or any other difficult situation. These are both large, beefy extruders. The Nimble is diminutive. You can put it places where the others won't go and might be a great solution for heated chambers since you can place the stepper outside the chamber. I don't think you can go wrong with any of them so it just depends on what printer you have, what (if any) hot end you have and perhaps personal preference. The good news is, we now have a few very good options to get rid of the Bowden!

  3. Very good write up as usual bud :)

  4. Richwap you wouldn't be able to compare it with a titan different type of extruder, if you wanted to compare it you would need to compare it against where the nimble actually derived from, also might be handy as a tip to use different colour of filament like black a colour renowned to show artifacts like ghosting, or anything that can show inertia in your prints, maybe some scales showing the true weight of the nimble on its own and with your effector, Please do't take this comment or response in a negative way the issue i have with the nimble and not you or anyone has ever given proper attribution to which the nimble has derived from , and thats from a open source view point

    1. Thanks Chris. I am aware of the hearsay about the origins of the Nimble but I have not talked to either party directly about it so I am not able to comment on that. I will say that the idea has been around for at least 6 years (,72925,73117)

      As for trying to capture ringing effects, I'll try to make some black prints and photograph them. My main goal is to find products that work for my workflow and not necessarily test them to their practical limits. I'd LOVE to have the time to do that but I don't.

      I don't use a lot of black as my customers like the wild translucent PLA colors. I can also weigh the effector assembly and report back on that.

  5. Great review and indeed, there are so many good options!
    I'm in the process of ditching my flying extruder setup and go for direct drive, but can't really decide now.

  6. The place to start is taking stock of what you have already and where you want to go. For instance, if you already have an E3D V6 hot end it makes sense to re-use it (unless you have another printer you can move it to). That takes the E3D Titan Aero off the table. Then, do you have a large format delta (>250mm build diameter) with stout mechanics? If not, then a remote direct drive option like the Nimble makes sense. If you aren't in a hurry, then wait a bit and see how things shake out with respect to reliability, robustness, etc.

  7. Ha! I have actually designed an adapter to place the Nimble inside the Titan Aero. Silly, but it does fit nicely.

  8. @Michael
    I'm a bit of an E3D fanboy, so I might just buy the Aero. The V5 was the first item I bought from them and since then they never failed me.
    I rarely print any faster than 60mm/s so the higher mass wont be an issue. The only drawback is that I was planing on adding heat to the enclosure, but I'm still considering it. I print ABS less and less now.
    Yes, right now I've a V6, but it could be donated to my MPCNC and try some insanely big prints lol

    I want to see that! :)

  9. Nothing wrong with being an E3D fanboy! I helped sort an issue with the V5 nozzle, it's on their blog:

    The Aero is an excellent direct drive integrated hot end extruder. I'll be posting a review of it in the next few weeks when I have some time. At this point it is my go-to extruder/hot end for my production fly fishing reels mounted on an UltiBots D300VS delta printer with mag ball arms. I print these slow so high speed issues (ringing, etc) are not a primary consideration. Ultra clean prints is.

  10. @Martin
    Yes sir, not a problem:
    Wrote a blog post for you. Thanks for triggering the need to write a blog. :-)

    Agreed that there is nothing wrong with being a fan of E3D. I like what they do and how they go about their business.

    1. Did you tried to combine with any small hotend like this one
      what would be the lightest option ?

    2. I really wish they'd make a 24v version of that hotend

  11. @steven - not yet but I am in the process of putting on a B3 Innovations Pico hot end which is ultra small. I'll post when I have that done and some prints under belt.

  12. I've currently got a Rostock Max V2 that I've upgraded to the injection carriages, Ball cup arms, Accelerometer board, E3Dv6, Tricklaser effector platform, and Gecko Tek Build plate. My plan is to eventually upgrade to the duet and FSR's (I plan to start printing and building a TAZ6 from scratch in the next few weeks and will do the duet upgrade when I get it close to done so I can use the Rambo in that build.) ;-)

    Anyway, I've been noticing some extruder clicking while printing Makergeeks Raptor PLA. Not that it's tough to print or anything, but if I go much faster than 50mm/sec it slips and clicks. I'm thinking about upgrading to the Bondtech QR (eventually want to print flex and carbon fiber at some point) but am now curious about the Nimble. Which would you recommend as the more solid upgrade? Any unique challenges in getting the Nimble to fit the E3Dv6/accelerometer board setup (even possible?)

    Thanks and keep up all the great info! I've learned so much about printing from this site and all your posts over at the SeemeCNC forum! :-)


  13. Harblar, Bondtech QR hands-down is the most reliable, solid, forceful extruder that I've tested. They are my defect standard extruder.

    AS for your V2 upgrade, depending on when you did this, there may be a simpler solution. After a few reports from me and a few others about extrusion pressure, SeeMeCNC redesigned the nozzle for the HE280 and did flow modeling to significantly improve it. I got one of the prototypes at the MRRF in March and I can say it makes all the difference. Get a new nozzle and give it a shot, I think you will be amazed at how much better it extrudes. If you were able to print at 50mm/s with the original nozzle you must have been in the 225°C range for PLA. That's not good!

    1. That's the way I was leaning. I also saw elsewhere where you recommended 1.8mm ID tubing so I'll probably go with that as well.

      I had the HE280 for a awhile, but I've been using the e3dV6 for the last several years and switched back to it. I upgraded to the copper alloy heating block and nozzles and added the silicon boot (If for no other reason, the boot makes it the better hotend, IMO!) :-)

      I'm still pretty much convinced my issue is the extruder. It's the original unmodded EZstruder. On the first layer or two I'll get clicking here and there. Just by putting a bit of manual pressure on the filament it stops clicking and extrudes fine. Speed isn't super critical for me as I tend to value quality over time, but if I could bump infill and supports to 60-70mm/s or better that alone would be great. I tend to leave inner perimeters at 35-40mm/s and outers at around 15mm/s.

      The filament I've been using lately is Maker Geeks Raptor PLA. It's a high temp PLA designed to extrude at around 245C. Personally I've had great prints with it at around 237-240. If you've never tried it, it's definitely worth checking out. It's super tough once printed and can be annealed making it even stronger. I've been printing a fullscale BB-8 droid dome and I've tried breaking a couple of the large pieces (that ended up scrapped for one reason or another) and I simply can't by hand. Standing on it finally will, but even then it's reluctant. Haven't tried annealing it yet to see how much tougher it gets, but they claim it boosts its Heat Deflection Tempt to over 290F. Great stuff!

      Anyway, thanks for the advice and keep up the great blog!

    2. If you do use 1.8mm ID tubing, make sure you have a capable extruder like the Bondtech QR. Normal extruders can't handle the extra friction reliably. The original ezStruder did have its share of minor issues but it works reasonably well for most things. But I need excellent reliability and performance so the Bondtech QR is my standard. I've NEVER heard a Bondtech skip unless the nozzle was literally blocked or plugged.

      I've heard of the Raptor PLA but not tried it yet. I'll give it a shot, thanks. I do use ProtoPasta's anneal able HTPLA ( which is anneal able too and love the stuff. My defacto standard filament for mechanical parts is Atomic PETG CF - I looovvvvvveeee this stuff. It prints like butter and is tough and rigid. Beautiful to look at too. No annealing required.


  14. Interesting read. I had sent Zesty an inquiry some time back and never got a response. Going to try again.
    I am interested to try the Nimble on my MK2 or my Haribo version I am building.
    Ultimately I would like to get one working on a Flashforge Creator Pro style machine to reduce gantry mass.

  15. Good Morning!

    I've been humming and hawing about the Zesty, and I feel like it's the right direction to go. Once I get the pesky Rostock pcb board out of my way the path to the direct drive holy land will be cleared. But, in the meantime, while I wait for budgets to catch up with my intent, I want to try a flying extruder.

    Do you have any experience with a flying extruder?

    1. Geoffrey, yes I have built a flying extruder in the past and also tested the one that Trick Laser sells (or used to sell). It does improve things - primarily by reducing the length of the Bowden tube. However, today I would recommend trying the Capricorn PTFE tubing as it significantly decreases Bowden friction and has a small internal diameter so there is less "slop" that results in stringing and blobbing.

      Also, Thank you very much for your donation! I will be sending you an invitation to my supporters Slack channel so please jump in and meet the group!

    2. Perfect thank you for your insight! I ordered some of that tubing so it should help.

      Should I post up on slack or the forum if I hit any snags with the upgrade?

  16. Nice post, I have to ask, you are aware zesty nimble is a bad copy of flex3drive?

    1. The Zesty nimble isn't a bad copy of the flex3drive, it's actually a good variation on the same concept. These types of systems were out in the wilds of reprap before flex3drive came about. Yes flex3drive had no commercial competition in this area for a long time but it certainly wasn't their original idea.