CR-10 S5 Bed Modifications

thingiverse

This is a solution to my CR10S5 bed leveling issue. When I purchased it, the bed was out of level >3mm. After I finished with the changes I have around .5mm variance in the build plate when warmed up. One of the pictures has the readings. The readings were measured from the top of the x rail. After initially trying to adjust the bed level, I had the left two bed leveling springs tight and the right two loose. After heating up the center was bowed up >1mm. I about returned the printer but decided to keep it. It is too big for my workspace but I already had some items ordered from TH3D to put on it. (Magnetic build plate and EZABL) Anyway I thought I would share the experience and the details are below. 1) I removed the bed to figure out where the issue was. The first thing I noticed was a 300 mm heated build plate mounted on a 500 mm plate. Surrounding that 300 mm heated build plate was an aluminum structure which looked like a way to keep the plate somewhat level. It dawned on me that this would restrain the 500 mm plate during thermal expansion and the plate had no were to go but up after heating up. (hence the >1 mm bow) In addition this looked like it would create a heat sink around the 300 mm heated bed and prevent the transfer of heat to the rest of the plate. So in my mind it had to go and I removed it. (Not too mention it was bolted together with mismatched heights.) 2) I measured the 4 corners of the sub frame plate (with the v wheels and belt drive) down to the perimeter frame and found >2 mm variance here. After looking it over it appeared the plate was bent. I pulled it apart and removed the v wheels and the plate layed flat on the table. I measured the eccentrics and spacers thicknesses. I had >.5 mm variance in them and the idler spacers weren't the same thickness as the eccentrics. There was no way this was going to get the v rollers at the same height in the v rail which would bend the plate. I also noticed the the eccentrics and idlers were mixed on both sides of the v rail. In my mind there was no way to actually get all the wheels touching correctly. I needed all the eccentrics on one side and the idlers on the other. This would cause another problem with the fixed v rail bolted to the perimeter frame. You would need to adjust the v rails to get the idlers preloaded correctly. I also needed the same height eccentrics and idler spacers. After pricing these out and thinking about the other modifications needed I decided to go a different route. 3) I opted for 20 mm linear rails and lm20uu bearings. The rails are 27.5" long and I cut them from a 55" shaft. I printed the 20 mm rail clamps to keep the same height as the 2040 linear v rail. This didn't affect the belt drive at all. I printed bearing holders to use the existing holes in the sub frame plate and to put the bearing flush against the plate. (Trying to maintain the same z build height). There are two different bearing holders. One is slightly oversized (which I used) and one is the exact size of the bearing. I wanted it slightly oversized for any misalignment and to prevent flexing the sub frame plate. The y limit bolts on the back frame rail instead of on the v rail. The bearing holders have 5 mm heat set inserts from Amazon. I used 5 mm x 16 mm pan head screws to hold the rails and bearing holders. https://www.ebay.com/itm/20mm-Diameter-Shaft-55-inch-Long-Hardened-Rod-Linear-Motion-CNC-Routers-/131046475200?hash=item1e82fa85c0 https://www.ebay.com/itm/US-Stock-4pcs-LM20UU-20mm-CNC-Linear-Ball-Bearing-Bushing-20-x-32-x-42-mm/391680100800?_trkparms=ispr%3D1&hash=item5b31f409c0:g:Q58AAOSw4WRbH4Pd&enc=AQAEAAAB0BPxNw%2BVj6nta7CKEs3N0qW3sTvQCS7%2F%2BzEyD6lbyOvl9%2FSGeep0hvwrCr2ePwzukH94JBkxrfrKe44GLgBrTo0O%2FWlPJB0pI7f6QQ3wyT9Qeb9y4xckYBxOeluNsQys8xYEkKOo574bMaB035Webb3Vt1VDAz1jgxxb8s5Ykngf2rMbRA9H2axDjg8NO7UQ0T%2F9meQLK5ey3xh9DKvxxgROLULUE3NbfO0aDNYfWOSsQTqJ6wXHbCXMWnooYK%2BOWRyj%2FrGGhypxd2S0r7%2FPXDN34fMEhnjFktIMfhXQIWCWpx3Etg0y%2B2ZFhSG1ITT2LSTsRr%2Bk5bN5N4%2F4Vk%2FvyYNyGqV7x9byMICZdNrf%2FQhT0Zw4DWUhLeiaAeWpVidwmEZwD9Vrnk4gW%2FiWfHq0yfugWOJ8OVf6L8kiqG152tdpQ0q00gV8gkfNU5JOv6vsBjQKUX4lWYTJ%2BHc4QMlpizJd683xg363ajNLYfn42fUKfCR%2BNyBLegprl1X9gPHFgJ6u4GWFYdgRZyrJwAsh3V%2Bmrn18f0w4EecKYpCC1eKqqFEgNgh35ST1vDdBmbzfREppbgfXNCl8zzdx5PNnld%2BgDPovyMvKhob%2BEprMzJqb&checksum=391680100800a60a955c447342128e0331f2c707fe9d https://www.amazon.com/a16041800ux0827-Cylindrical-Knurled-Threaded-Embedded/dp/B01IZ1RN0A/ref=sr_1_4?keywords=5mm+heatset&qid=1568245302&s=gateway&sr=8-4 4) I used solid 3/4" od by 3/4" tall garolite spacers from McMaster instead of die springs. I also used AA shims ( from McMaster) to level the bed. I used the original bolts and knobs to hold the bed down. You could probably just print some spacers if you wanted. https://www.mcmaster.com/92273a154 https://www.mcmaster.com/97235k808 5) After leveling and heating it up I have around .5 mm variance in the bed with the stock glass it came with. I feel it is within range of Auto Bed Leveling or printing with a thick bottom layer. 6) I did use the adjustable z limit mount from https://www.thingiverse.com/thing:2349068 Kudos and thanks. If you need any other details or have questions let me know.