After looking at many of the entry level FDM style 3D printers on the market, Ben Roberts chose to invest in a Leapfrog Creatr 3D Printer. He has written an in depth review of his first experiences with the machine and was kind enough to share his experience with us.
For a person who has a fascination with electromechanical devices and CAD, a 3D printer must be the most desirable thing ever invented. It certainly is for me! My interest in precision electromechanical devices comes form years as an Engineer in commercial radio broadcasting. I began in the 80’s when the industry ran on tape and the machines demanded constant fettling and repair. Injected into this background was a three year stint in 3D animation and graphics.
Many years later, I find myself with this mix of skills and interests, trying to select a 3D printer to buy. With so many to choose from now, it was a steep research curve. From the outset, I found the RepRap and MakerBot styles unappealing. I didn’t like the spindly look of the RepRap or the laser cut wood chassis on the MakerBot. I was looking for a machine that appealed to my mechanical sensibilities, which was proving difficult without spending a lot of money.
One site I found very useful in the decision making process was www.3Ders.org. They have a page dedicated to a comparison listing of all the currently known printers. My shortlist of desires were: decent construction, big build area and heated bed. Eventually, the Leapfrog Creatr became the only candidate. The company appeared to be solid and they had begun shipping the single extruder version to customers. However, another temptation was lurking on the site, a dual extruder version for a bit more cost. How could I resist? Here was a professional looking machine with the biggest build area (at the time) and dual extruders which would enable me to make more complex models. I created an account on the www.lpfrg.com and clicked through a purchase. Then I waited….
And waited. There was a caveat on the site to expect around a six week delivery so when six weeks was up I contacted the company and received a reply the following day. The company representative informed me that I would have to wait a little longer. Eventually, I received a call from UPS asking me to pay the GST (clean forgot about that!) and the port handling fees. Two days later the crate was in my spare room waiting to be unpacked. The unit was very well protected and survived the trip from the Netherlands with only one mark that I could see which was a scuff on the inside of one of the side panels from the extruders rubbing on it. The factory had locked the axes down with zip ties but the extruders had still managed to come partly loose from their carriage on the Y axis.
My first impression of the machine was that it was very nicely constructed with quality components. Looking inside, the stout nature of everything from the brackets to the drive belts is impressive. It was important to me to have a machine that would get the best possible quality from the FDM process.
The chassis of the machine is primarily made from “T” slot extrusions. The verticals have a radius on the corner which gives the cabinet its smooth outer look. Internally there are square section “T” slot extrusions that make up the horizontal elements of the chassis. The open spaces on the sides of the enclosure are filled in with powder coated aluminum panels, which complete the very finished look of the machine. In my opinion, this is the best looking machine on the market and equals the MakerBot2 .
The linear motion components are just as impressive as the chassis. The X axis runs on 12mm unsupported rods and is driven on both sides to avoid racking. The drive is from twin NEMA 17 stepper motors through a substantial 10mm belt on 1:1 pulleys. Lots to like here.
The Y axis uses identical belt hardware but a single stepper the carriage runs on 8mm unsupported rails with cylindrical bearings. The extruder carriage assembly is made up from laser cut 3mm aluminum plate components and have a tab/slot arrangement for assembly. This uses less fasteners and is dimensionally accurate.
The extruders themselves feature 0.5mm nozzles and are heated by resistors rather then nichrome coils. This is a much neater arrangement and involves no thermal cement or Kapton tape.
The Z axis is very substantial . In line with current practice it is suspended/driven at three points. Three point drive greatly simplifies bed leveling as there is less interaction between adjustment points during calibration. The drive is composed of a single NEMA 17 stepper driving a loop belt that engages three 1:1 pulleys mounted on threaded rod. The rods spin and raise and lower the bed on plain nuts on the rods. The nuts are held captive on the brackets connected to the driven plate. The driven plate of the Z axis is 3mm aluminum. The bed itself is borosilicate glass attached at four points with 4mm cap screws and springs. The bed heater is glued to the underside of the glass sheet and has a mat of waste cotton insulation backing it.
These neat turntables for the filament reels are a nice inclusion. I’m not using them here as the filament I ordered came in bundles.
Under the electronics cover inside the printer are the motor drivers, power supply, Z-Axis motor and the Arduino clone microprocessor. I was pleased to see that real motor drives are being used and not the Polou type.
The printer ships with a couple of spare sheets of bed material that assists the adhesion of the part to the bed and a single sheet of A4 paper which directs you to the leapfrog website to begin the machine setup. The instructions on the website look a bit “light-on”. However, the instructions were laid out well and adequately covered the basic machine setup.
First, the software had to be downloaded and installed. Pronterface is the bespoke machine controller for the Creatr. It unpacks with a copy of the open source Slic3r. The website recommends you hit the Slic3r website to get the latest version. With the latest Slic3r unpacked and copied into the Pronterface folder I just needed to install the Arduino USB drivers. With all of these procedures completed I was ready to start Pronterface. The first thing to do with Pronterface open, is to configure Slic3r. Handily Slic3r has a wizard and you just scroll through the instructions on the website and fill in the boxes in the Slic3r wizard.
Slic3r is a credit to the developers. It looks and behaves like a mature piece of commercial software. It is integrated into Pronterface as a plug-in but is actually a front end for processing your model into G-Code. You will actually do most of your settings and processing in Slic3r.
The makers recommended not having the bed heat on for the initial print. I did not download the suggested model for printing, opting instead, to print an open sided cube I’d made in SolidWorks. Being unfamiliar with the software I assumed it was unable to display my model as a short, square tube. It looked like a blank square. Undeterred, I pressed on. The website directs you to a Vimeo clip to demonstrate how to load the filament and it’s quite straightforward. Using the “Extrude” control in Pronterface, you feed the filament up the tube and then give it a bit more help going into the feed roller and pinch bearing. It’s important to be patient with this because the feed stepper turns the feed roller very slowly. Don’t forget, there is 1.75mm filament being melted and squeezed through an 0.5 mm nozzle. Like an oncoming train, the moment to push the “print” button was growing ever nearer.
When the printer was delivered, there was a neat little gearbox already printed out on the bed. I had noted the nice crisp edges and lack of runout on the model. I had loaded a blue ABS filament so I was a bit surprised to see orange filament coming out! Ha! It was just some leftover from the gearbox print. My blue filament eventually arrived and came out in a nice even squiggle. “This is it!!” I thought, and hit the “print” button.
It’s difficult to actually observe the print with the extruders obscuring the bed and after a couple of passes, I wasn’t seeing anything cube-like appearing. For the noob, this is a defining moment. How on earth do you figure out why it’s not printing? I homed all the axes in Pronterface and moved the bed down 30mm or so. I hit the “Extrude” button again and watched as it produced a good looking stream of ABS. At this point I threw away years of training and experience in fault finding. I made several alterations to the Slic3r settings all at once. This time when I hit “Print” I was rewarded with an actual object! The machine began printing my cube and got about three layers down before the part lifted off the bed and got munched up. I was really pleased! Maybe I hadn’t achieved my cube but I had seen the FDM process working right before my eyes. What followed was a descent into chaos. I didn’t get another thing out of the Creatr. I fooled around with so many variables, that I utterly lost track and the machine seemed to be fighting me. I called it a day.
The following evening I only had a couple of hours to experiment and I really didn’t progress at all. My final attempt was to go through the Slic3r wizard again and put all of those variables back to the standard settings. I still didn’t get a print.
Saturday dawned, heralding a full weekend of printer tomfoolery. With a bit of time to cool off and reflect, I attacked the machine with a bit more subtlety. One setting I had made was a Z-axis offset to put the bed closer to the extruder on the first pass. Inadvertently, this had the effect of blocking the extruder and it had actually gone hydraulic. No wonder it was making a sort of morse-code instead of a bead of ABS! I eased the offset by 0.2mm and magically, the machine began to print! . I turned on the bed heat for this attempt and set it at 60degrees. I printed my cube and realized why it had looked like a blank on the Slic3r. I’d created the cube in SolidWorks in the top view, with the hollow in the Z axis. When I loaded it into Slic3r, it had interpreted it as having the hollow in the Y axis. As the cube printed out, it printed a 20 by 20 mm square and then began to print the 3 by 20 mm sides. When it arrived at the top, it tried to print into thin air and the final side of the cube was a bit of a mess. It was a completed print though!
Since then, I’ve attempted more complex and larger prints. I’m very happy with the quality of the prints but am still having some noob problems with settings. This Raspberry Pi case printed well but had some layer separation issues. This is an incomplete print but shows the 40% infill is very even.
Pros and Cons.
- Solid construction.
- Quality hardware.
- Uses open source software .
- Electronics are standard RepRap configuration.
- Generous build envelope.
- Turntables for filament reels.
- Waste cotton insulation under build table is fraying.
- Plastic wrap on cabling untidy.
- Poor access to extruder nozzles to clean off waste.
Review written by