Making sure that the bed is flat
-
Lykle and Stephen thanks for the reply,
It would be great to have some thick aluminum bed, but I have not found too many solutions for 310mm diameter bed that are ready to implement. McMaster has square plates http://www.mcmaster.com/#standard-aluminum-sheets/=14kdzw0 but they need to be machined, as they are at best have a tolerance of +- 0.127mm. A 300 mm bed could be found at http://www.ultibots.com/300mm-heat-bed-kit/ but it has a thin aluminum. 14 in Al disks here http://www.sandsmachine.com/alumweb.htm but seems like they are also have +-0.127 mm tolerance. While 1/16" plates are available, I would think that it will not stay straight on an uneven surface, or under an occasional pressure from a head crushing down
If you have a better solution, please share.
-
The two places I've ordered metal from online are Speedy Metals and Online Metals. Both have MIC6, but in square only. So, you'd need to make it round or modify the printer.
-
The two places I've ordered metal from online are Speedy Metals and Online Metals. Both have MIC6, but in square only. So, you'd need to make it round or modify the printer.
Or make it hexagonal - probably easier than making it round.
-
Thanks, other than the size matching, does the surface needs to be further machined to be completely flat? Speedy Metals also has steel and stainless plates, which they say could be cut to a specified size. Do you think using something like stainless would be a good idea?
-
Ordinary sheet metal is not usually very flat, but sheet glass usually is, because it's cast floating on a liquid. Tempered glass can acquire distortions, or thin glass can sag - and we're looking at 50-micron level tolerances. For large printers, getting flat enough plates can be expensive - metal ones need to be machined flat. This is why some firmwares (but not the one on the Duet) support mesh bed compensation - basically, use a sensor to measure the bed shape and adjust all Z heights accordingly (thereby propagating any bumps into your printed objects). Delta printers are a bit different - they have enough calibration parameters (three endstop heights, a tower radius, two angular corrections, and maybe a rod length) that you can compensate for a fairly complicated bed shape. Of course if your delta parameters are adjusting for a bendy bed, your object dimensions are going to come out distorted. So best to have a genuinely flat bed.
-
Most aluminum and steels are rolled to thickness and will warp when heated, due to stresses left behind by the rolling. The MIC6 AL plate is cast and milled flat, it'll stay flat when heated.
-
Thanks everyone for the advise. With the paper removed from under the glass, things look a lot flatter. I will be ordering 1/4" Mic6 Al, possibly cut as a hexagon or something.
-
So, I am ordering the Mic6 plate from http://www.sandsmachine.com/alumweb.htm
Although they have 1/4 inch Al that is 0.002 inch flat, they have suggested that 1/2 in Al could be even flatter. Also, I would think that 1/2in will be sturdier. However, do you think 1/2in is overkill? -
Half inch may take a long time to heat up because of the increased thermal mass. For 310mm diameter, I think 1/4 inch should be adequate.
-
So, I am ordering the Mic6 plate from http://www.sandsmachine.com/alumweb.htm
Although they have 1/4 inch Al that is 0.002 inch flat, they have suggested that 1/2 in Al could be even flatter. Also, I would think that 1/2in will be sturdier. However, do you think 1/2in is overkill?Yes I do (think it's overkill) - from personal experience. I bought 10mm thick as it wasn't much more expensive than 6mm. It'll be the first thing I change when I get chance.
Issue number 1. Generally the thermistor is fitted between the heater and the plate. - bad idea with thick plate - it takes a long time for the heat to "percolate" through to the top surface. So, the thermistor sees the temperature at the underside of the plate and turns off the heater until the temperature drops, but the top is still cold. It takes forever for the top surface to get hot as the heater keeps switching off because the underside gets hot quickly. Also, any change at the surface, like when print fans come on, won't be noticed at the underside of the thick plate where the thermistor is, so the heater won't react. I got round it a bit by drilling a hole 3mm dia by about 40mm deep into the edge if the plate and fitting the sensor (PT 100 in my case) in there. It's better but not ideal.
Issue number 2. Related to the above - even with the sensor fitted inside the edge of the plate, it takes about 6 minutes for the top surface to get to 50 degC. That's with an 800W 240V mains silicone heater (on a 400 x 400 mm plate). What's worse is that it takes about 45 minutes to cool back down. That's not too big a deal for me as I use glass on top of the plate and have 3 pieces that I can swap between but its still a pain - makes my study where I keep my printer like an oven too.
As I said, changing from 10 m to 6mm will be one of the first things I do. Oh, and of course your 1/2" is a tad over 12mm thick so even worse.
HTH
-
David and Ian, thanks for the quick reply. Yeah, although the temperature measurement could be solved by drilling a hole in Al plate, long heating and cooling times could be annoying. For now, I am switching to a 220W 24V heating pad from a 12V Onyx, which is an improvement but nowhere near the 800W.
