We assemble the Altronics Core I3 printer and put it through its paces.
3D printers are becoming a lot more accessible to the home hobbyist, so when we saw that Altronics had released their Core I3 we had to get our hands onto one to test it out.
On slicing the tape and opening the box, the first thing noticed was the sturdy packaging. Everything in the box was securely housed in closed-cell foam, with a spot for every piece, and a protective foam cover as well. Several layers allowed for all the different parts, while the entire lot was bound with shipping wrap.
After laying out all the pieces, being careful not to open the bags of small hardware for fear of mixing up the parts, a check revealed that all the parts were present and undamaged. The protective plastic film on the acrylic parts was retained until later. It is worth noting two identical-looking pieces on page 3 of the instructions. One is acrylic, and one is fibreglass, but the outline is the same.
Two different sized fans are used, and four stepper motors are present: one with the filament drive attached, two with toothed drive pulleys loosely installed, and one with a bare shaft.
The parts are divided into three foam trays, one of which contains five bags of small hardware. Unlike some flat-packed furniture and other products on the market in flat-packed or kit form, these trays and bags are not divided by the construction stage unfortunately.
In retrospect, checking the parts list against each individual bag of small hardware may have saved a couple of headaches, as detailed later, but these turned out to be due to errors in the instructions rather than the incorrect number of screws.
Several tools are provided. Some small hex keys with ball ends, a spanner, and a pair of scissors. Although the hex keys did the job, in some cases the tolerance was too tight to fit the screws in question, or too loose. If you have your own set, use them instead. The same goes for the spanner: It is meant for the M3 and M4 nuts but fits neither very well. As these are all nylon locking nuts, the sample was assembled with the aid of small multi-grips and a shifting spanner that were on hand.
Additional tools that could be used to make assembly easier are a small pair of long-nosed pliers, and small side cutters. Calipers or a ruler will help in verifying the screw sizes, as some are only 2mm different.
Finally, also in the box is an SD card and a USB card reader. The card contains separate PDF documents for the assembly instructions, operating manual, and a connector map for the main board, along with a video each for the assembly process and operation. The card has a copy of Cura slicing software as well.
With no prior experience assembling CNC machinery or kits of any kind, but with extensive experience with circuit board kits and flat pack furniture, this build was always going to be interesting. For the sake of authenticity, it was attempted without watching the videos.
Assembly begins with the base frame, which contains the stepper motor and drive belt for the moving print bed, the Y-axis. Assembly of this section was fairly straightforward. Pay careful attention to the materials list, however. The aluminum profile used is the same cross section in multiple areas of the build, while the length varies. The same is true of guide rods and drive belts. The material list describes which one to use with phrases like ‘The longer ones’. Where parts listed are already assembled or bundled, the materials list also calls this out.
In yellow boxes are important points, such as cautions and screw sizes.
The build starts with the installation of the slide bearings to the underside of the build plate bracket. Note that each has four screw holes, while only two screws are used. Take care not to tighten these up too far yet: The bearings need to be able to move by hand.The belt clip goes on next. This is direction-sensitive, so ensure it is installed in the correct direction.
When installing the limit switches, referred to as ‘stroke switches’ in the manual, long-nose pliers are useful in guiding the locking nuts into place. Pay attention to the yellow highlighted instructions here, and don’t tighten the nuts too far yet. The same goes for the T-nuts installed in this step - they need to only just hold on.
When assembling any of the toothed idler pulleys, care must be taken to assemble them correctly. They have three studs on the matching faces of each, which fit into recesses on their opposite number. However, they have to be rotated until the teeth on the outer face line up, which was found to be only in one of the three possible positions.
When assembling the idler pulley holder, note that, as elsewhere in the manual, a size is given for the nylon spacers. This is the length, not the inner or outer diameter. Note also the assembly order - the bracket attaches to the frame first, then the spacer/pulley assembly goes in. Pliers are helpful here for inserting the spacers.
The next step is mounting the motor. A couple of points are easy to miss so take your time. The motor mounts to its bracket with four M3 x 6mm screws, which do not get tightened yet. The bracket allows for motor movement while the screws are loose, which is used later. Also easy to miss is the direction that the control cable connector is mounted, and the direction that the motor attaches to the plastic frame section, which at first appears symmetrical, but in fact is not.
Assembling the next section of the frame had some challenges. This is where the first T-nuts are used. Normally, T-nuts are slotted into the channel, and they begin to rotate when the screw has enough friction. They are designed to twist through 90° and then catch on the side, allowing the screw to be tightened with the nut in place. Two of the ones used here did not catch, and had to be held with a hex key. Additionally, care must be taken with the screws which hold the feet on. These longer screws also end in T-nuts but have to pass through two holes in the frame with do not align easily.
The guide rods are the next item to install, and as before, the material list describes which ones. This step will be difficult if the slide bearings are too tightly secured. The other plastic frame section mounts now, and once the entire assembly is firm and the bed plate slid back and forth a few times, the bearing screws can be tightened further, but not completely.
Installing the drive belt wasn’t as easy as it could have been, but it is certainly achievable with care. The orientation is critical, as is making sure the M3x6mm screws holding the motor on are loose enough to allow movement. The belt passes around the motor and idler pulleys, with the join on what is the underside as seen from the assembly perspective (although this is actually the top when the printer is assembled and working). In other words, the join goes closest to the build plate. This may be confusing to read but makes sense when you are looking at the build.
The easiest installation method is to loop two cable ties around one end of the belt but only tighten them enough to bite and not pop undone. The free end is then slid under the belt clip, over the bracket, and under the cable ties. Only about 15mm was available as slack on the sample. With the cable ties tightened, the belt passes around the idler; over to, and around the motor pulley; and back to the belt clip. What is left of the belt is now secured in the same way as the beginning, as shown here.
Once this is secure, slide the motor toward the outside of the frame until there is sufficient tension on the belt, then secure the motor screws firmly.
The final step to the lower assembly is to secure the acrylic covers. They are not the same shape, and one mounts the limit switch. Both mount with the same number of screws: Two M4x8 button heads with T-nuts in the channel, and two M4x25mm button heads with lock nuts through the plastic frame.
With a final check for movement, the bearing screws on the bed plate can be firmed up.
The instructions call for a test to see if the plate triggers the limit switch. On the sample, the plate reached the switch but did not trigger it. The plate was contacting the PCB holding the switch before the switch was actuated. Bending the switch legs forward on the PCB by about 2mm fixed this.
With the lower frame complete, attention turns to the gantry. This section relates to both the X and Z planes. Note that the lead screws are different lengths, and the guide rods are the shorter ones but without the shoulder at one end. Note also the toothed pulleys: Two have large shaft diameters for the lead screws, and one has a small shaft diameter for the stepper motor.
Read all of the yellow boxes before assembling anything in this area (and throughout the manual). Several important points are flagged, such as the non-symmetrical belt clip, and the idler plate, which has two holes on one edge and only one on the other. Small pliers will again be useful in locating the locking nuts here. The self-tapping screws holding the limit switch on, are easily overtightened and stripped. Note also the different lengths of the lead screws. The longer one has bearings both ends, while the shorter one ends up in the motor coupling later with a bearing at the other end.
As before, don’t fully tighten the motor screws - movement is needed later.
Section three is the bulk of the vertical frame. Look at the remarks column in the material list carefully, as explanations are provided with good reason.
Be careful to mount the limit switch to the plate made of fibreglass - it is identical in footprint to an acrylic one used on the other side. The locking nuts were difficult to insert here and hold while the screws went in. Eventually, some masking tape solved this problem. Additionally, the M3x6mm cup head screws called for here did not seem to be long enough, only engaging by a thread or two. They were substituted for 12mm screws from the workshop supply, however, at the end of the build, enough 10mm screws were left over that these could easily be used. The mounting plate is sloped on one side, and this is a mirror for the one on the other side, so follow the image in the instructions closely. As before, note the direction of the connector on the stepper motor when mounting it, and keep the screws loose for now.
When mounting the hardware described in the instructions, page 35, pay attention to the direction of the slope on the mounting plate, and be sure that the acrylic sheet is being used. Additionally, this step appeared to have another screw size mistake: M3x18mm cup screws are called for. Not only did these seem too long, but there did not appear to be enough left. 12mm units were substituted from workshop supply. In retrospect, this assembly mounts in a position where the 18mm screws would have plenty of room, but the length is redundant and there are still not enough 18mm screws. At the end of the build, only one 18mm screw was left, but enough 10mm screws remained to suggest that in fact, these are the correct screw for this step.
When mounting the belt idler assembly to the Z transmission plate, take care with the orientation of all parts. The far left of the picture depicts two extra holes, which are important as a cable tie point later. If the Z transmission plate, which is otherwise symmetrical, is upside down, the cable tie point will end up in the wrong place when the time comes to anchor the extruder head cable bundle. Have a close look at all of the diagrams below.
Installing the shorter lengths of aluminium profile in their vertical position was fairly straightforward. Care must be taken to align each on both sides to the distance called out in the instructions. Note the orientation of the base frame: Measurement is from the end without the stepper motor. The instructions suggest using the glass plate plus any acrylic part to get the measurement, and this is indeed a helpful way of getting a square alignment.
When installing the Z motor assembly and its opposing bearing support bracket, be careful to mount them on the correct side of the profile, as per the instructions. The stepper motor in the base frame and the one being installed should be on the same side of the profile. Secure both with their T-nuts so that the plates are level and resting on the top of the plastic T-pieces that hold the profiles together.
Be very careful when adding the plastic bracket at the top which connects the two profiles. Follow the instructions, page 42, to place the part in the correct orientation. The same goes for the acrylic sheet that braces this assembly. It can only mount on one side of the bracket.
After this, the lead screw assembly from earlier is mounted to the frame. Some finessing is needed to get the ends of the guide rods and lead screws into their home positions, before checking the motion of the X axis slider. If it does not move freely, rotate one of the lead screws to align the axis, before installing the closed-loop belt. This step is not clearly highlighted in the instructions, though the images show it. With the belt in place, the transmission plate with the idler pulley and cable tie holes is next. Again, a little finessing is required to seat this, but nothing that should not be expected with any kit.
Assembling the extruder head is straightforward but requires care and attention to detail. Points to note are the size and direction of fans, the small metal tubular throat that fits into the heatsink block, and the way the heater and heat sensor install into the nozzle block. The heater element and temperature sensor are retained by a hex head screw, which should only be tightened enough to hold position, and not damage the silicone insulation of the sensor.
There is a small set screw inside a recess on the heatsink block for retaining the nozzle assembly. Tighten this before the final fan is installed to the block, after which the print head is installed to onto the X axis slider.
Attaching the X axis drive belt is relatively pain-free, although long-nose pliers help when feeding through some of the tunnel-like spaces involved. The belt mounting process is familiar from the lower frame assembly, although the one in the sample had much more spare length. It was tempting to check if they had been swapped around but measurement proved that the instructions do indeed suggest the correct ones for each place. The diagrams make clear exactly what has to happen, and mistakes can occur if the diagrams are not consulted.
Nearing the end of the assembly, it is time to mount the build plate. Four M4x35mm screws mount through countersunk holes in the heated build plate base, with springs slid over the the screws before they pass into the baseplate carrier. Adding these one at a time proved successful. Pay attention to the orientation here. The silver side faces up, black side down, and the connector underneath toward the stepper motor in the base frame. The glass plate sits on top and is held in place with four foldback clips.
The final task is cable routing to the control unit. Material list #6 describes the layout of each wire in the tray, and where each goes. The mainboard end of each is colour-coded. The white end goes to the destination, such as the X axis stepper motor or limit switch, while the opposite, colour coded end goes to the appropriate socket on the control board. The next few pages of the instructions contain location drawings to highlight the connections. There is one white header on the board which does not receive a connection.
With that complete, the instructions do not describe the next step at all. Included in the materials are two lengths of spiral binding. Using the pictures as a guide, these are wrapped around the cables from the extruder head, all the way to the main board, then a separate bundle from the motor and plate heater, around to pick up the Z axis motor, and then to the control board. Folding back excess cable length results in a neater job at the end.
OUR THOUGHTS ON THE BUILD
All in all, the build is quite easy, as long as care and attention to detail are applied liberally. Notwithstanding a few screw size errors in the instructions, they are clear and easy to follow, with ample pictures and diagrams for clarity. If these are read before attempting assembly, the build should be problem-free. The process took around two and a half hours, and wasn’t a serious struggle to someone who has never built a machine from a kit before.
There is more to do before we can print our favourite object, however. Next up, the machine has to be initially calibrated, and a test print performed.
Initial calibration turned out to be very straightforward indeed. Once powered up, the instructions describe using the rotary encoder/pushbutton in the front panel to navigate to a menu for adjusting the Z axis. The menu is a little buried, but just scroll down until it pops into view. Once selected, the Z axis will move up and down for three complete cycles. Toward the end of this time, the screws holding the Z axis to the frame are tightened gradually until all are firm. After this, as far as the instructions are concerned, setup is complete.
The menu does allow for a great many more parameters to be modified and adjusted, however, this is an out-of-the-box review. As such, it is now time to perform a test print with no further changes.
The unit is not provided with a filament holder, and this will become many users’ first print. For the sample, a little aluminium extrusion and a bracket were drawn from workshop supply and added to the side of the frame. This is fine for the test print, however, the reel is too close for larger prints and will interfere. A longer term solution is needed.
The test print selected is a D6 dice model. In anticipation of the grammar purist onslaught, die is no longer considered the singular noun in Standard Australian English, nor in most other modern English languages. The model was made quickly using free online software, and little attention was paid to accuracy of the pip placement. The point was to gain a model with some concave and convex curves as well as straight planes in X, Y, and Z axes.
The SD card bundled with the printer comes with Cura slicing software as an executable file. Not only was the sample install file in unrecognisable characters, it was also outdated. Instead, a fresh install was downloaded. The Core I3 was not on the list of available printers. However, because the Core I3 is based on the Prusa model, the Prusa I3 was selected as the machine type.
The dice model was fed through Cura using mainly default settings: 0.1mm thickness, generic PLA, 20% infill as default, and bed adhesion turned off. By default, bed adhesion is ‘on’ and set to print a brim. This didn’t really suit the print, and was less of a test.
The SD card was inserted, and the print selected. The bed and hot end were preheated, for which the ‘prepare’ menu has to be accessed, and the temperatures dialled in. The bed was set to 60°C, and all appeared well. However, it became apparent that the Core I3 struggles to reach 60°, and this took quite some time. Once it did, however, the print was started.
The first print failed spectacularly and quickly. However, the printer appeared to be functioning well, so the glass print surface was cleaned thoroughly with isopropyl alcohol and a microfibre cloth. This removed any surface contaminants from the manufacturing and shipping process.
The second print seemed to be going much better, however at approximately one third of the way through, the print lost adhesion to the bed. The print bed temperature seemed to be below the set 60°C, dipping as far as 52°C. This should never be the sole factor in bed adhesion, however, so plan B was called into action.
With blue painters’ masking tape in the appropriate place on the bed, the third attempt at printing the dice was initiated. All seemed to be going well, so it was an appropriate time to have a closer look at the machine in action. The blue back-lit LCD display on the control unit gives readings for the X, Y, and Z planes. The readings appear to show millimetres away from the origin. Also displayed are nozzle and bed temperatures, time elapsed, and build completion as a percentage.
The idler pulley for the X axis also has an issue which was visible during operation: Although the instructions clearly show the 2mm nylon spacer between the plate and the pulley, this takes the belt out of alignment and causes it to rub on the bearing assembly for the Z axis. This could be fixed by using some nylon flat washers in place of the supplied spacer to reduce the spacing but still keep the pulley spaced away from the plate.
Fast forward an hour, and the print is complete. This time, with the masking tape for bed adhesion, all went well. With direct-to-glass printing earlier, the first layer had seemed coarse and a little messy. The model was deliberately printed without supports of any sort for the concave curves of the pips. On the base of the model, the direct-to-glass prints had struggled with this, although the layers above became progressively neater.
On the successful print, the first layer was a similar story, being a little messy and coarse. This is a theme for a great many 3D printers, even some expensive pre-built ones. The rest of the print was very clean indeed, with minimal aberrations in the layers and well formed concave surfaces in the pips. The photos will attest to the quality of the print. The top layer was clean and well formed, with no gaps and tight bonds between the extrusions.
With this in mind, the printer was put to use making a bracket for the development of an upcoming project. This bracket is a simple structure to hold a push button arcade switch. This print yielded some minor problems, including some deformations in places, but none that are out of line for a 3D printer and on par with the units we use here to produce parts for our builds.
All up, the Core i3 is an impressive printer for its price that was successfully assembled by a novice and was able to achieve quality results out of the box. The menus available on the controller are worth exploring, and the provided manuals and explanatory videos are helpful. With some fiddling with settings and trimmings, the result is sure to improve further.
Grab it online from Altronics:
- K8400 $599 (Check website for up-to-date price)