X5SA Printer Calibration
The start of this is documented in the 6th November blog post and I am slowly working through reading the content of the Teaching Tech Calibration tool. The main problem has to be working out where to start and perhaps I should just follow the order of the list of steps on the site. Part of the problem with that is I need to use the machine to print parts to address areas that are identified. The first one is the cooling fan for the part at the nozzle. Claude even identified that the air used by the fan should be ducted from outside the machine as currently it's using the heated air rising from the bed. Something that does make sense, but something that seems common. I'm currently printing a new shroud for around the extruder nozzle to direct cooling at AT the part, which the current option does not do at all well. Given the number of designs on Thingyverse, that seems to be the first part replaced, although I have already addressed other areas such as the correct parallel paths for the belts. While a new option has been printing, I've been reading ahead and it would seem that it's not just cooling that affects stringing, but more likely the head temperature being too high, so I've just trimmed back to 200 degrees while it's printing and will see if the second half is cleaner than the first. I am also not entirely sure that the jets around the duct are actually pointing in the right direction. A few designs point a couple of jets directly at the nozzle and that may be preferable and then controlling the speed of the fan may have a more observable effect. Having lowered the temperature the print seems to be a lot cleaner so one wonders if the part is actually needed? It can't hurt to give it a go anyway.
In addition to part cooling, the retraction of the fibre in the Bowden tube is also flagged and given almost one metre of fibre between the head and the extruder drive, there is a lot of room for variation. Another modification has been to move the stepper drive to the centre of the back rail and halve the length of the tube, but another print on Thingiverse is an arm to simplify that and move part of motor and fibre out sensor above the centre of the back rail and even closer to the printer head. The next print after playing with the cooling duct. I may resort to a little hand tweaking of the nozzles on that before fitting it. Just clean out the holes at a steeper angle to that printed ... with care, but also cleaning any stringing in this first pass at the part.
The third step is yet again the bed levelling, which does seem a little hit and miss, and given that the whole bed assembly is balanced between the two z axis drive shafts, I do wonder if this could not be a little stiffer. I do have both an A3 router table and the Taig Mill, both of which have a very stable z axis. Originally I was actually looking to buy JUST the laser head to fit to the router, but a complete kit worked out cheaper than the suitable stand alone unit ... at the time. So I have the separate laser cutter option. Given the quality that I think is needed for some parts, I am sure that adding a 0.2mm nozzle extruder head to the Taig will give me something that can print high detail within its 30X15X15cm footprint, and the while the router footprint covers 30x43cm, it has a more limited z axis. Although even that could be addressed. As I said at the start too many options and no clear path forward.
Finally reached the page on testing the accuracy of the axis and while the guide on sizes of the printed models indicates that the step sizes hshould be changed to get the right result, I was not in agreement. Certainly the likes of the Taig mill has a very accurate set of lead screws and number of steps per inch ( they are imperial threads ;) ) is a accurate value. Michael at Teaching Tech does a nice job of explaining just why you do NOT measure the prints and has confirmed my own thoughts about the small discrepancies I am measuring. That the large part is only 0.25mm under size I think confirmes that the calculated values are the right ones, but I certainly need to address the belt tension problem which is currently a little slack!
That parts are coming off the machine that are usable is the main thing and I am more than happy with the accuracy so far.
The next steps are additional modifications and I also need to document previous upgrades just to complete the picture.
Steps to calibrate machine
X5SA Bed Leaveling