Once I had the printer and the AMS set up, I set to work printing our test models, starting with the 3D Benchy, a boat. I printed it with Bambu Lab orange PLA, and it took a little over 15 minutes. The overall quality of this print was fantastic, with nearly no vertical fine artifacts (VFAs) on the side walls and no problems with the overhangs on the front hull or the bridging for the roof or doorway.
(Credit: Michael Lydick)
I measured the overall width of the Benchy as 59.96mm, with 60mm being ideal. The print measured 22.94mm across the roof, with 23mm being ideal. Overall horizontal width was 30.96mm, with 31mm being ideal. This puts the average deviation at approximately 0.15%—well within acceptable tolerances in a near-perfect class of accuracy.
Next, I printed a hexagonal cube organizer to observe bridging and retraction. Normally, if the retraction or cooling isn’t working properly, we’ll see any amount of stringing or warping along the bridges between the hexagonal patterns. In this instance, the cube was perfect. I printed this at the 0.1mm layer height with the pre-tuned Bambu Lab PLA settings, and could not see the layer lines.
(Credit: Michael Lydick)
I watched the printer actually bridge the top slot, and was amazed when it crossed nearly a full inch in thin air without any sagging. This print quality appears injection molded, and is superior to the Prusa Core One test print of the same file, in that I could easily see the Prusa layer lines with some minor stringing in the hexagonal boxes.
I noticed while printing the hex cube how quiet the machine was during the actual printing process. I had to walk into the room to see if it was still printing, in fact—a first for me. It never exceeded 50 decibels. However, the pre-print process is quite loud. While going through its startup, the filament cutting sounds like knocking the handle of a butter knife on a table top. I grew accustomed to it, but I questioned the need for it when everything else about the H2S is so well thought out.
Next, the tolerance ring test, which has a rounded print-in-place fit for 0.35mm, 0.3mm, 0.25mm, 0.2mm, 0.15mm, and 0.1mm. All six required no effort to deflect and spin around in their respective orbits. I didn’t have to break any of the balls free; they were loose straight out of the printer.
(Credit: Michael Lydick)
This was also printed at the 0.1mm layer height with the pre-tuned Bambu Lab PLA material settings as read from the RFID chip in the spool.
Next up was the Autodesk/Kickstarter test, a benchmark print with a scoring rubric in which a perfect score is 30 points. We award or subtract points based on the accuracy of each part of the print, which includes spires, a stacked circle tower, and pins that are intended to fall out of their respective holes. We also award or subtract points based on quality metrics, such as VFAs and bridging of each level. The H2S achieved a total of 27.5 points, which is exceptional.
(Credit: Michael Lydick)
Our screw thread test was also a success, with gentle to moderate finger pressure required to screw the metric stainless steel screws in from 8mm to 3mm without any issues or noticeable interferences. Full threading was achieved with no noticeable variation in threading pressure from the 8mm to the 3mm screw holes.
(Credit: Michael Lydick)
Finally, I printed our geometry detail test, which is designed to get a feel for how certain types of geometry features might print at certain sizes. Printing with Bambu Lab PLA at a 0.1mm layer height, the thinnest wall thicknesses were visible across the boxes and cylinders. However, I saw noticeable degradation in the last “Test” font on the top, with some of the letters not printing legibly.
(Credit: Michael Lydick)
