Installing a new Calidum bed heater on a Solidoodle 4... So, Printit Industries introduced a new mica bed heater recently, called a Calidum and I had to get one. I ordered it on Monday, May 16th at about 10:15am. I had a confirmation of order email shortly there after, and a confirmation of shipping at 10:30am! The package arrived in Wednesday's mail - just 2 days after ordering! I honestly wish more companies would use USPS for shipping an ditch FedEx, but that is a whole different can of worms... back to the bed heater... I was amazed at how thin and light this Calidum heater is! I just had to put the calipers on it... The main body of this heater is less than 1mm thick! Obviously, where the wiring and thermistor are attached is thicker, but the main heater is really that thin. The bed comes pre-wired with both heater connection wires and thermistor already in place - no soldering or taping required on the bed itself. It also comes with 3M adhesive applied - all you have to do is peel the backing off, align it, and stick it on - no messy glues or tapes required. The only thing you will be required to do is add the connectors you need to attach to your controller board. So I pulled the build plate out of the printer - pretty straight forward operation - disconnect the existing wiring from the controller board, remove the leveling screw wingnuts and lift the whole works out. In the distant past, I had replaced the stock build plate with a thicker aluminum plate due to the stock plate being severely warped, but reused the silicone heater pad from the stock setup. To attach it I had used some Ultra Copper automotive gasket maker (high temp). I had also installed cork as an insulation material to help with heating - also from the auto parts store (gasket cork for making your own gaskets). All of that had to be removed from the aluminum build plate - a putty knife is a very useful tool for doing this... Cork removed... note the original silicone heat pad is smaller than the actual build plate... the outer edges were always a problem when trying to print larger items in ABS - not enough heat out there... Cork and silicone heat pad removed, just needs a bit of cleaning up to remove the residue of the Ultra Copper. A single edge razor and Goof Off took care of that nicely. Once the bed was free of all residues I gave it a quick once over with some 150 grit sandpaper to smooth out any nicks in the surface that may have been created by all the scraping. I also gave it a wipe with an alcohol pad to make sure the surface was clean and free of any oily residues... then lined up the new Calidum bed (before removing the adhesive backing) to get it exactly where I wanted it, and marked the corners to make re-alignment easier, pulled the adhesive backing off, lined the Calidum back up and laid it into position. A quick rub across the surface to make sure it stuck down well, and it is attached. Just that simple. I reinstalled the bed leveling screws and at this point it is ready to put back into the printer. Also of note: there are pre-drilled holes for the 3 point leveling systems like the Solidoodle uses, and the corners are cut diagonally for use with 4 point leveling systems - so either way, installation is going to be easy. After getting the bed back into the printer, it was a simple matter of routing the wiring back out to the controller board and making 2 simple connections. I use a Rumba controller board, so the heater connections were super easy - one wire to each of the 2 screw terminals for the bed heat and done (there is no specific polarity on this heater, so it makes no difference which wire goes to which terminal). The thermistor wiring needed a connector installed - not terribly difficult, but I found that the wire gauge used here was a bit large and I had to trim a few strands of the wire off to get the terminals crimped on solidly. (I have been in contact with Printit Industries and discussed the issue with them. They say they will be looking at using a smaller gauge wiring for the thermistors on the next production run to make life a tad easier.) OK - physical installation is complete! Now, depending on what type of thermistor your old heater used, making an adjustment and flashing firmware may be required. This heater uses the same type of thermistor as the E3D hotends use, and as such, requires table 5 in the firmware for accurate temperature reporting. My old silicone heater used table 1, so I had to make the required change and flash my firmware. Not a terribly difficult thing to do. Now - on to the heating comparisons! Just how much better is this new heater? (if at all?) Well... With the stock 150x150mm silicone heat pad, my extra thick aluminum plate (4.6mm thick vs the stock 3mm) and a 3mm thick piece of glass on top of that, my bed heat times to 100°C were a bit long... approximately 23 minutes from around 23-24°C... This is a screenshot of the temperature graph during heat up for the stock heat pad... Notice how it starts off pretty well, but then begins to flatten out at about 45°C... and continues to flatten until it finally gets to the set temperature. Now, here is a screenshot of the new heater temperature graph from the initial test run... Note the much steeper climb from start to finish - no flattening out, and the time was about 13 minutes - a savings of 10 minutes!
I have since reduced the initial heat up time to right at 10 minutes, shaving 3 more minutes of that initial heating time. Remember, I have a LOT of thermal mass to heat - about 7.5mm if thickness in total, and I still have no insulation under this heater - yet. The heat is very consistent across the surface (no more cold edges!) and after doing a bed PID tune, the temperature is VERY stable - no more up & down temp variations. It is so stable and so consistent that I am actually able to reduce the set temperature and still get excellent adhesion. I have had this Calidum heater in service for several days now, and overall, I am very happy with the purchase. In fact, I am thinking I will pick up another one for the second SD4 I have. ~~~~~~~~~~~~~~~~~~~~~~~~~~~ Time to go melt some plastic! Update - 12/2/17 Still in love with the Calidum bed heaters! It has been a year and a half, and NO problems what so ever with the Calidum. I did purchase a 2nd one about a month later for my second Solidoodle 4. Installation was even easier on that one since I replaced the entire stock bed & heater all in one go (didn't have to scrape the old one off, lol) - #2 is now almost identical to #1 in terms of how it is set up. Even with some serious thermal mass, they both heat up quickly, and maintain even heating throughout printing. I would definitely purchase the Calidum bed heater again. Hoping that they will come out with other sizes soon... I have a large format that could use one of these. So not too long ago, a fellow 3d printer friend and I got into a discussion about calibration methods - specifically the feed & filament calibrations. I use one method, he uses another. He argued that it really doesn't matter what method you use (although, he insists that the method he uses is better because it is posted in the reprap wiki pages), the end results are what matters, and to a point, a can agree with that. My argument is that accuracy in the calibrations does matter.
The specific points of contention have to do with the feed rate and extrusion multipliers. The method of calibration he uses says to "get close to" feeding a specified amount of material, never touch the extrusion multiplier and fudge away any inconsistencies by adjusting the flow rate. The method I use says get the feed rate as close to exact as possible, adjust the extrusion multiplier to create the desired single wall thickness and there should be no need to mess with the flow rate. Since this discussion I remembered reading a forum post about the math involved in slicing a model for printing, and the calculations that the program must make in order to produce a recognizable object. The math backs me up. For those that are interested, the method my friend uses is Triffid Hunter's calibration guide. My method is outlined in my filament calibration posts found on this blog. The forum post that discusses the calculations of a slicer can be found here. You should read them and decide for yourself... ~~~~~~~~~~~~~~~ Time to go melt some plastic! All 3D printers and all hotends are not created equal.
We currently have four different 3D printers. Each one will run the exact same filament at a different temperature. One spool of filament - lets say white PLA, on each machine will print best at a completely different temperature on each machine. Machine 1 has an older hotend with the thermistor taped to the outside of the nozzle. Reported printing temp on this machine for the white PLA is around 165C Machine 2 has a generic J-head hotend with the thermistor installed in the heatblock. This one reports the printing temp of the same white PLA at 200C - 35 degrees more than the first machine. Machine 3 has an E3D lite6 hotend. Reported printing temp on this one, same filament, is about 215C - 15 degrees more than machine 2, and 50 degrees more than machine 1. Machine 4 has a full E3D v6. It and machine 3 are very close in temperatures (+/- a degree or 2 at most) because they share the same basic design. Fifty degrees difference from one end of the spectrum to the other, all with the exact same filament. So, when you enter the world of 3D printing, you need to be aware that there are no carved in stone rules about temperature. You have to find the temp that works best for your printer and hotend. Asking someone else what temperature they run a given filament at is, at best, a gamble. You might get lucky and they have the same type of hotend that you do and their temp range works for you, or, you may have complete and utter failure. ~~~~~~~~~~~~~~~~~~~~~ Time to go melt some plastic! So lately I have been seeing posts in Facebook groups of new folks trying to use layer heights that are too large, and some other common issues. There seems to be some misinformation floating around about these things and I would like to take this opportunity to hopefully clear them up
Knowing what size nozzle your printer has is very important. It is the nozzle diameter that several slicing settings are based on. Most 3D printers on the market today are using a 0.4mm nozzle size - it is a nice compromise between flow and detail. There may be a few out there that use slightly larger, or smaller diameters, so make sure you know what you have. Or, you may have an aftermarket hotend installed with a different nozzle diameter - again, knowing what you are using is a big factor in getting good prints. So, what are the settings that are affected by nozzle diameter?
For those that are interested, there is some great info posted over at soliforum.com about the math that slicing programs use (the write-up is based on Slic3r, but other slicing programs will have similar calculations): http://www.soliforum.com/topic/9551/soliprint-v11/page/2/ (look for the post by jagowilson about 1/3rd of the way down the page). Setting bed level is also critical to getting good prints. I see a lot of people wanting information about "auto-leveling" setups - which is the best one? The answer to that question is: The best "auto-leveling" system is the one that isn't needed. If you take the time to properly set up your printer (ie. calibration, calibration, calibration) then you should not have any need of "auto-leveling". Taking the time to do things right early on will save time in the long run. I rarely need to touch bed level on my favorite printer, even if I get a little impatient and pull parts off while the bed is still warm (very often seconds after they complete) the bed stays where it is set and I can turn the heater back on, load a new model, slice and print again within moments. Running an auto-level script for every print adds several minutes to print time, and frequently is not very accurate. In fact, I did away with the auto-leveling inductive sensor that came with my kossel kit. It was problematic from the start, and after adding the mirror tile (my typical print surface) to the bed surface, it became completely useless since it requires close proximity to metal to operate. Calibration of the effector is the critical component on a delta/kossel printer. Yes, it takes time and a lot of patience to do it right, but it is worth every second once it is dialed in correctly. So far, we have calibrated our feed and flow rates to get good extrusion, but some of the fine details in certain prints can be difficult to get good fill on. Well, here is where we fix that. In Slic3r configuration, in the Print Settings tab, there is a choice called Advanced. This is where we make magic happen Load up a model that has fine surface details - like lettering. Lettering is notoriously difficult to make look good with the various shapes of each letter. I will be using a small tag I created to print color samples in. This tag is 40mm wide x 38mm tall and the base is only 2mm thick with the letters being another 2mm - not very big. After loading our model, we go to the slicing tab in Repetier Host (RH) and click on the Configuration button and open Slic3r's configuration window and click on Advanced in the left menu. You should see something like this: If this is the first time you have ever used this tab, your numbers may be different than what is shown here, more likely to be almost all zeros with the exception of the First Layer setting which probably says 200%. The first thing you want to do is specify the Default extrusion width (top box), if you haven't already. That number, once set, should remain the same until you physically change nozzle size, and should be 120% of your nozzle diameter. I use primarily 0.4 nozzle size, so my default is 0.48 (0.4 x 120%). The other numbers shown above are all equal to the default extrusion width, which means everything will be at 100%. now lets slice our model with those settings and take a look at it. At first glance, it doesn't look too bad...but lets look closer. Notice the gaps in the areas the arrows are point to? These areas, when actually printed won't look very nice. So how do we fix it? Simple, we tweak extrusion width settings. Lets go back to Slic3r's Advanced settings and make some changes... First thing to notice is that we left the default setting alone. That one should not change unless you are changing nozzle sizes. Just about everything below that, however, has been tweaked. Whenever you make a change in the settings, be sure to click the save button, then re-slice your model with the new settings and see how they have affected the model. As you can see, there are differences between the original default settings and the new settings, but we still have a few gaps that need fixing, so back to the advanced settings we go and tweak a little more... Again, make sure you save the new settings, and re-slice your model. Lets take a look at what that did. The top layer of the base looks pretty good.... ....but the lettering still has some ugly gaps. To fix those, we continue tweaking the settings until we are satisfied with the overall look of the fill, keeping in mind that curved areas will never be 100% perfect. Just for fun, I went a little crazy with tweaking the extrusion widths and ended up with these settings... Resulting in this for a sliced top layer in the model... Please note that the Default extrusion width never changed throughout this process. For the record, yes, I have printed using similar settings as the last settings screen shot. It does work.
Slic3r does have more advanced settings in the stand alone version where you can set things like different layer heights for a single model, and use modifiers to set different infill rates for certain areas of a single model, but those are for another day. ~~~~ Time to go melt some plastic! |
AuthorA "Jack of all trades, master of none", I have dabbled in a lot of different things, but none have held my interest like these darn 3d printers do. Archives
May 2016
Categories |