We’ve been working hard to collect accurate and detailed information across all of our departments to give you this single January update. To start with, here is a summary of our news. On the up side, we have completed vetting and testing about 95% of the Pro’s components at production volume. The basic Pro firmware and software have been put in place, and we have seen the Pro complete good prints. This may sound simple, but in fact represents the culmination of all of the integration needed between the software, electronics, engineering, sourcing, and assembly teams. We now see that the core Pro engineering and electronics design are fundamentally finalized for full-scale production. This is despite us not having the all-important autocorrection feature enabled yet, which was in the plans to be finished around the time of the market tier release. We believe the Pro results will show themselves at their best in 2-3 months when the printer has most of its systems fully enabled and optimized.
On the flip side, we see room for certain improvements that with a moderate delay will add significant value to the product for the Jan-March tiers. We wish we could have announced this sooner, but as a matter of principle, we don’t want to speculate about potential delays and will only announce them if we are 100% sure they’re a fact. As we will explain in detail below, we’re not just being perfectionists - this delay allows us to make a final round of minor corrections (i.e. 0.1 mm refinements) on several parts to provide better precision, assembly, and reliability. We could have rushed the product, but based on overwhelming support from our backers and contributors, we see that you would prefer to wait for the more refined version of the Pro. To be exact, our new expectation is that delivery will be delayed by about 2 months up through the February Tier, and the March tier will be delayed by about 1 month.
Early-release tiers (Nov/Dec Backers) should expect to receive their backer survey in the next 2 weeks. We will have further information about the survey and options in the update below. We are excited to hear your experiences from the in-the-field use and are grateful for your support in this endeavor!
Our Progress & Successes
The biggest news is that we have eight Pros in various stages of testing that are printing using production parts and electronics. This represents the culmination of 1000’s of features and lines of code coming together successfully. The team has put in heroic amounts of time and effort into this project, it’s a real labor of love for us. Through their teamwork and a lot of inspiration we are seeing our dream of a revolutionary printer come to life.
Even more than that, these Pros are printing using the new M3D GUI (Client Side Software). What that means is these are the first prints from a full M3D Pro system all working together. Seeing the work from all departments come together in one machine is momentous, because the printer has officially come to life. As a comparison, reaching the same stage with the Micro took twice as long.
With respect to the firmware, it’s important to reiterate that only basic capabilities and features of the Pro firmware are fully finished. These features include firmware that understands standard G-Codes, a single USB device that is recognized immediately by the host computer, a bootloader to allow updates to both ARM processors, and all of the control algorithms required to successfully print. This is roughly equivalent to the feature set of the Micro, with the exception that the Pro is running motors with acceleration and the Micro does not. Meanwhile the Micro has multi-point bed height calibration and we’re implementing this over the next few weeks on the Pro.
There is also a key distinction between the Pro firmware as a platform and what runs on the Micro, and that is software maintainability. The Pro was architected to a solid plan and design, and each system implemented has been quite a bit easier to integrate without worrying that other subsystems might start misbehaving. These systems get unit tested and when released are far more ready for prime time than some of our prior releases. This also means that the pace of development for upcoming additions of heated bed, untethered printing and power recovery features - leveraging risk reduction tests communicated in earlier updates - is expected to pick up. We're looking forward to an extremely capable printer in the very near future.
We want to give you an idea of how far along the Pro platform is. This print was done using only the most basic software.
The correction guidance, backlash, and bed leveling systems are not yet enabled for this print and yet the printer executed the file with no problem. This is very exciting for us to see the Pro at such an untuned state already comparable to the fully developed Micro. With the rest of the systems in place, an optimized Pro will shift the user experience towards focusing on printing for their application rather than tinkering with their printer, and this puts us in great shape to change the game.
Again, what you see is limited until we get those last production parts to arrive. In the meantime we’ve tested many iterations and will continue to tune the Pro as seen in these prints from the Pro. For those of you with a discerning eye: We did notice poor z-wobble and that our extrusion settings were a bit low and we are currently working on improving these issues. There are certainly many areas which will need continual refinement over the next few months. However, we’re also seeing phenomenal retraction and very good overhang printing (80 degree).
Here is another test using one of the more intricate prints requested. The layers are joined wonderfully and this was with no supports while printing the overhangs.
Here is the early example of 50 micron layers produced by a Pro. As some may realize, printing at fine layer resolution appears to be a frivolous battle to wow consumers with technical specs since it takes a very long time to print. However, the reality is that printing at very fine resolutions indicates fantastic temperature control and gantry rigidity, and that's why we aim for these specs.
Over the holidays we were hard at work and made a number of improvements to several minor systems based on initial tests. One really significant system was the new internal feed port. Tests show a significant improvement and promises an elegant space saving option without the worry of significant increase in drag on the filament. By comparison to the Micro, the Pro internal port system has been overhauled; it benefits hugely from a larger spool, which puts far less force on the system; a bearing-based filament holder (accessory add on) that further improves consistency; significant force increases on the extruder side; improved cooling internally; and the auto-correction on the extruder motor itself. In combination rigid filaments will work in this improved internal port system very well.
We've upgraded the DC jack to handle twice the maximum load that the Micro PCB could and to provide a larger style plug to reduce the risk of damaging the electronics between these different models.
The print head has undergone an upgrade and gained a few style points as well.
Above you can see a comparison of the prototype print head vs. the delivered production parts. A couple of paragraphs below you will be able to see that the print head is slightly transparent to allow for viewing of the internal workings. eig In summary, we have several Pros that are successfully printing. Two printers are going out to CE for testing, and the rest are being used by the engineering and software teams. Another 12 printers are currently being built by the production team and being used for training. One of the reasons progress takes a long time right now is because the few parts we are waiting on have to be made by hand, which is painstakingly slow compared to a production process. These are the printers that will go out to the earliest-tier backers in about 2-3 weeks.
Pleasant & Unexpected Surprises
Heated bed overhaul: The heated bed system was completely redone to improve the electronics systems. The first benefit here is printbed detection, allowing us to stop the user from printing unless a bed is installed. We’ve added short (+) and ground detection which allows us to detect that a user is trying to install a bed upside down. We’ve added a safety interlock, which detects shorts and shuts down power within 2 microseconds. Additionally, because the Arm-core processors are so sensitive, we’ve added secondary circuit to protect the master chip.
Extruder LED port: We managed to squeeze in an LED port for the extruder cover. At this time there is no plan to include an LED in the first 2500 units, but the option is there and we’re certain users would come up with interesting ways to light up their units. We may add a kit at a later time.
Special edition stickers: We are looking at having a limited run of stickers made to celebrate your status as a backer.
Semi-transparent cover: We created a deep black transparent cover, which allows us to get great visual effects without having to make an expensive and easily-damaged painted part. A near result of this is that with enough lighting, you can easily see inside the printhead, show off the brass, or monitor the filament path.
Things to Refine
Right now, of the 1000s of component features that make the Pro we feel good to have confident knowledge of the few small refinements left to complete. With 95% of the project on projected course we have some housekeeping on some small components we still need to refine. We don’t see much else that is likely to be added to the list given the amount of time we’ve had to experiment with the production parts. The rest of this update highlights the things that we know need improvement.
BOX frame: Came in with light polish to make it shiny, it isn’t perfect looking at a few assembly seams / Bases are slightly non-flat depending on processing (this is hard to notice but still not up to our ideal). In a few months they will be coated with a expensive UV cured clear hardcoat and assembly mold lines will be refined and will be more seamless.
(by selecting the option to defer your delivery, you may receive one of the later cases that will receive the upgraded coating, as seen below)
Motor pulley: Pro’s are going to be a few microns more wobbly at this stage, and we found the set screw loosen giving a slight increase in backlash (the correction systems can detect this but we want to eliminate it as much as possible). This will be refined to have near perfect centering / set screw will be beefed up and coupled to matching flat to guarantee it never slips.
Heater control board: came out bad and was non-usable even for prototypes. We’re rush ordering new ones and it's a good thing too because it forced us to design a much more aggressively designed board that uses 3 times less power (about 0.3 W waste instead of 1 W) and has multiple hard and soft logic lockouts that prevent overloading of the system or ARM processor. This is one area where the world wide shipping slowdown caused by the chinese new year is leading to us needing to delay delivery so we can have enough boards for all units.
Heated bed: came out nice (see below) but can be moved by hand with significant effort side to side about 1.5 mm. This means it could short to ground with enough effort, a condition the printer can detect and is harmless, and can let you know it's pushed too far. In the future (2500 units in), slight refinements will eliminate this issue.
In the last month we also drastically beefed up the bed quality, paying more than double to get super-strong glass. As impact testing of the heated bed shows, without a doubt, the durability of the glass. Everyone will benefit from this improvement today.
Threaded rod height: current rods sit on the bearings and have a bit more friction than we planned. That may not affect the actual print quality or print speed but may reduce the highest Z speeds that are mainly used to get back to home after a print.
Flex wire soldering: 11 flex wire termination points on 7 PCBS were done using surface soldering to save space. Although we made several hundred the manufacturers had to put their best people on them and it isn’t sustainable long-term. It also causes them to be a little fragile. In the next iteration (about 1000 units cut-in) we will redesign them to be a bit beefier and have through-hole components only.
Sliders: The sliders came out a little less perfect than expected. They can be made to work perfectly fine but need far more TLC than is desireable for full-scale production. For instance sleeve bearings need to be glued and hand reamed and belt installation can be tricky. For the most part this just slows down our production a lot and requires more learning. To iterate on this design we’re making a new mold for the sleeve bearings that eliminates uncertainty in positioning the parts/production times.
The gantries came out passable, however, internally we made some changes to the bearing stacks that we thought would benefit things but made making prototypes nearly impossible the last 2 months. So we made two new designs. One by CNC and one by MIM (injection molded stainless steel). As a result, we have two good sets, and need some tweaking to make production faster in the next iterations. For example we have to retap the ends to make sure they are accurate, an issue we dealt with on the Micro before.
Test of the threaded rod on a corrected threaded front.
We found the square rods were not to our quality standards and had to be sent back to the supplier to make adjustments then re-deliver them within our specs so they are concentric and uniform. Unfortunately we ran out of time to get these rods re-made before the chinese new year.
Milestones for software continue to be reached. Below you can see our software running both a Pro and a new Micro+ on the same computer. Even more exciting is you can see the multi model support feature in action. Save time and print multiple objects in the same job without needing to stitch them together in an editor suite.
The Bootloader is done. This all important/critical step for the Pro’s firmware allows us to send printers out and remote-update them, and activate additional features. Thus saving on shipping costs and downtime when new features are ready to update.
Please take a moment to see the Pro inaction working on a multi model print job, via the link below.
Now for the question you have all been asking…
We are releasing the backer surveys in the next 2 to 3 weeks for the first tiers and the rest shortly after. This survey will need to be completed in order to have your rewards delivered and will have additional options for shipping information changes and leaving us any special instructions. The survey is your opportunity to also order accessories and additional filaments.
Above is the proposed external spool holder for the 500 ft spools. (please note that these are concept renders and the final product may vary slightly).
(a closer view)
For a quick look ahead: Here you can see a prototype for an SD extender. Down the line with software development this will give alternate access loading to the internal SD memory system.
YOU CAN DEFER YOUR TIER!!!
In your backer survey, there will also be an option to defer your delivery date to a later time if you wish to receive a more developed Pro based on the details we've posted here. You can also change this option back at any time.
By deferring your Pro delivery into the May-June timeframe, what you would most likely get is the following improvements: The most refined version of the product in terms of hardware, firmware, and software; Internal port bearing kit for free; Extruder light kit for free; UV-clear coated box frame; Refined box frame with less obvious assembly lines; Refined extruder cable, with finer mesh; access to all planned accessories for delivery in a single combined shipment.
Please see the updated delivery schedule for the reward and pre-order tiers:
M3D Pro, Early-release tier: Previous delivery estimate Nov 2016 -> New delivery estimate February
M3D Pro, Early-release tier: Previous delivery estimate Dec 2016 -> New delivery estimate February
M3D Pro, Early Production tier: Previous delivery estimate Jan 2017 -> New delivery estimate Late March
M3D Pro, Early Production tier: Previous delivery estimate Feb 2017 -> New delivery estimate Early April
M3D Pro, Market ready production tier: Previous delivery estimate Mar 2017 -> New delivery estimate April
M3D Pro, Market ready (Deep Discount): Previous delivery estimate Mar 2017 -> New delivery estimate April
M3D Pro, Market ready (Super Discount): Previous delivery estimate Mar 2017 -> New delivery estimate April
M3D Pro, Market ready production tier: Previous delivery estimate Apr 2017 -> No change, but expected Late April
M3D Pro, Market ready production tier: Previous delivery estimate May 2017 -> No change at this time
- M3D Pro, Market ready production tier: Previous delivery estimate Jun 2017 -> No change at this time
Thank you everyone for your continued patience and support! If you have any questions not yet addressed in our Pro updates, please send them to [email protected] and we will be happy to get them answered.