Hall-Theta (Hall-O) Z-Probe Bed Leveling Mount for Rostock MAX by 626Pilot
Work in progress - follow for updates. Update - 3-6-15 The previous "mount mounter" has been replaced with upper and lower mounters. This provides two benefits. First, the lower mount has projecting "feet" that will absorb the impact should there be a bed crash. This will prevent the probe's mounts from being sheared off in the event your firmware fails to respond to a probe hit, either because of loose wiring or a firmware crash. Second, the new mounts stiffen up the probe immensely. I found that the probe's wiring could pull at the top of the probe a little bit as the platform moves around, and the new mounts constrain things enough to counter this effect. I got a slightly better (less error, more repeatability) as a result. Note: The probe itself hasn't changed since the 1-31-15 update, but if your probe was printed from a version older than that one, you'll need to download and print a new one. (Old changelogs at bottom) Who is this for, and what do I need? The Rostock MAX is a great printer, but it really needs help printing in some places at the edge of the build surface. Due to the vagaries of delta printer kinematics, whenever the carriage must travel "outside the triangle" formed by the three towers, positioning errors become highly likely. If all you ever print is little stuff that fits inside the triangle, you probably don't need this. However, if you DO print things that go outside of the triangle, this may be the only way to get decent prints. If you want to use this probe, you probably need to get a controller board that runs Smoothieware, such as the Smoothieboard or Azteeg X5. We have a user with two different boards (an X5 and a Smoothieboard) who has used this probe to calibrate a Rostock MAX and another delta printer, and it worked fine. Why not just use the auto-calibrate from Repetier or Marlin? We tried - the results were worse than manual calibration, and were useless for printing. That might change by the time you read this, so you might want to check on it before investing in a new board. However, the ARM-based processors you get with a Smoothie controller are exponentially more capable than the Arduino-based controller your printer probably has now, and the Azteeg X5 is quite affordable compared to an older board like the RAMBo, so it's not a bad buy. (Personally, I recommend the Smoothieboard 5X, which has big beefy power FETs - my Azteeg X5 lost its heater FET last year after only a couple months.) Whatever board you get needs to have either two Z endstop connectors (min and max) or a Z probe connector (basically the same thing, just marked differently.) Probe Repeatability This was tested using G29 on my Smoothie fork: [PR] Repeatability test: 10 samples (S) [PR] Acceleration (A): 90.0 [PR] Debounce count (B): 0 [PR] Smooth decel (D0|D1): true [PR] Eccentricity test (E): on [PR] Probe smoothing (P): 1 [PR] Probe priming (Q): 5 [PR] Feedrates: Fast (U) = 70.000, Slow (V) = [PR] 1 step = 0.00938 mm. [PR] Priming probe 5 times. [PR] Test 1 of 10: Measured 1136 steps (10.650 mm) [PR] Test 2 of 10: Measured 1136 steps (10.650 mm) [PR] Test 3 of 10: Measured 1137 steps (10.659 mm) [PR] Test 4 of 10: Measured 1136 steps (10.650 mm) [PR] Test 5 of 10: Measured 1136 steps (10.650 mm) [PR] Test 6 of 10: Measured 1136 steps (10.650 mm) [PR] Test 7 of 10: Measured 1136 steps (10.650 mm) [PR] Test 8 of 10: Measured 1135 steps (10.641 mm) [PR] Test 9 of 10: Measured 1136 steps (10.650 mm) [PR] Test 10 of 10: Measured 1136 steps (10.650 mm) [PR] Stats: [PR] range: 2 steps (0.0188 mm) [PR] mu: 1136.000 steps (10.650 mm) [PR] sigma: 0.447 steps (0.004 mm) [PR] Repeatability: 0.0188 (add a little to be sure) Most G29s I run return anywhere from 0 to 0.02mm (20 microns) of repeatability. For reference, we want 30 microns or better to calibrate for an acceptable 1st layer at a height of 0.1mm, so 18.8 microns is fine. Origin of the idea This is a reinterpretation of Marco Antonini's Rostock Z-Probe, done in Sketchup. His probe doesn't fit to a Rostock MAX platform, so I decided to design a new mount that would, as well as accepting neodymium magnets you can get from Amazon and a metal rod you can get from McMaster-Carr (both of which are ideal if you're in North America.) I also thought the original probe didn't look threatening enough, so I embellished that a little. At the moment, this part (and the instructions) should get you up and running with a basic Z-probe. . OLD CHANGELOGS: Update - 1-31-15 A "mount mounter" has been added. This is a yoke that you can bolt to your effector, and the main mount can be snapped in place and fastened with a screw (although it fits tightly enough for me that I haven't needed to bother). The main mount has been modified to fit properly with the yoke, so you'll need to download and print it again! The advantage of this is that you don't have to use probe offsets, because the probe will be perfectly centered. If you're using this to do an auto calibration, you really should avoid probe offsets because they suck and they are terrible and they stink and they will degrade the quality of your calibration. I know this from experience! Update - 10-22-14 Board holder has been resized a little to make fitting the board easier. Also, the side mount has been lengthened. If you mount the probe with the side mount, inside the effector, it should be right in the middle! This is the best possible place for a probe to be mounted. Update - 7-8-14 I added a platform mount to the side of the probe, about 2/3 of the way up. This allows the probe to work with an E3D mounted under the platform. I've also added a calibration piece, which contains only the probe rod passage holes. You may need to adjust your filament flow to get these just right: enough pressure to keep it steady, but not enough to stop it from moving freely.