September 19, 2019 at 9:00 am #114701
I have had the thoughts of using my Lowrider 2 as a 3DP as well. Building a Print head and mounting it to the 611 plate will be a matter of modeling and printing it. The part I would need to figure out is the heat bed. My idea is to use my 4’X8′ table and partition off 2’X3′ as the only area to print on. I then want to figure out how to take (6) 12″X12″ heat beds and control them all at once with 6 thermisters. Is that even possible. The print surface would be a piece of glass from my work (As we have a scrap piece of plate glass and a bunch of tempered glass they said I could have). Is this something possible, or would it be better to build a separate MP3DP machine but build it to have a print size of 2’X3′. The reason for this specific print size it is matches a “D” Size piece of paper and thus anything I could print on a “D” Size sheet of paper I could also 3d Print (Like House Floor Plans)September 19, 2019 at 9:19 am #114706
That is pretty awesome and crazy. Are you an experienced 3D printer user? Even if you are, this is going to be a real challenge. But it is possible.
You can wire up several heaters (they will need external mosfets and a properly spec’ed PSU) and put a piece of glass on top. To start, use a standard 8″x8″ bed and get things rolling. The glass will hopefully be very flat, but that’s not your only problem. The Z will probably not be perfectly level to the plate.
The first layer needs to be very accurate. If the first layer is 0.2mm and you’re off by 0.05mm, that can lead to adhesion issues. 0.05mm is a much smaller Z tolerance than we need in milling. This is going make a lot of your prints fail due to warping or just completely coming off the bed.
The prints are going to take a really long time. If you go from an 8″ X to a 24″ X, that’s triple the time. If you go from an 8″ Y to a 36″ Y that’s 4.5x the time. So now you’re talking 14x the time. If I print something that covers the whole bed (which is very nerve racking) on my MP3DP, it takes an hour at least to do 10mm high. So each 10mm high you print is going to take about 14 hours (this is a very rough estimate, you can try slicing something to see what slic3r thinks it will take). Not to mention, to get a good finish, you’ll probably have to slow down a lot, since you’re moving the whole gantry. That could add 2-3x more to that first 10mm.
If you got a huge nozzle, like 1mm and fed it with 3mm filament, you might be able to get that down to reasonable times again (a 1mm nozzle has 6x the area of a 0.4mm nozzle). Back to 2 hours or so for 10mms. That will take a lot of heat to melt that much plastic, so be prepared for those issues too.
I’m not sure how big the Z could reliably go. 2-4″ is the norm for the LR around here. Printing at 8″ (200mm) would be really neat (but several days, with a 1mm nozzle).
Add to that the fact that as you print an hour longer, those random, occasional errors are going to be just as prevalent. So if you try to print a 7 day marathon print, it could fail on day 6, several times. Just a painful thought.
The real trick is going to be leveling the bed. A sensor and mesh leveling is the only way I would attempt this. I would be really interested to see what the numbers are coming out of that.
I would love to see this. It would be amazing. I loved watching Dui’s MPCNC in China print huge parts (and it’s got a huge nozzle and is liquid cooled). I know Dui could do this. So could a few others here who know their stuff. I don’t want to discourage you, but it’s not going to be easy.September 19, 2019 at 11:00 am #114736
As far as 3d printing I am pretty well versed. I print floor plans to make scale Architectural models right now. I have to take and make a ton of parts and pieces to make a complete floor plan and with the different options of added an option of a 2′ bump or something adds to this. The thought is how much time in the overall it might save me on some of these. I know the print time for a single print is huge, but I would think the print time for one large print would be the same as printing multiple models and then taking the time to assemble them.
The exterior finish is no big deal as I am using a different product on the outside to simulate siding and it is done in sections just like real siding is done. This would actually probably end up working out to being about the same time, but I would need to spend more time watching the printer (While I sit out there and draft the next project) rather than spending the time modeling in the house. I am guessing a lot of my time in initial setup would be bridging for the openings. I make most of my operable windows on my printer right now and it could continue to make them, and then I could clean up the opening and install prior to cladding my exterior and doing the trim.
It is worth the thought. I have a couple of projects in front of this ideal, including getting my MPCNC Lowrider to doing its job first, and then I will put this into thought. Right now the board is clean, and for a little bit I have no models coming up, but I have a couple of jobs in the future that could turn into 7 to 10 house models at 2 weeks or more each just for printing so it is worth thinking about.
Thank you for the info, and it gives me something to think about.September 19, 2019 at 11:05 am #114737
I had a friend do something similar.
I think in the end, he ended up only using a thermistor on the hot plate in the middle of the print area. What he found is that the glass or aluminum bed did a great job of dispersing the heat to even it out across all the hot plates. He only monitored the one plate, but then used separate mosfets all controlled from the same pin to turn all the plates on/off at the same time.
I think he also ran a custom 5 minute sleep/wait time for after the bed getting to temperature before starting his print. It allowed the heat to even out across the build plate.
One of his plates tended to run hotter than the other, but most PLA and ABS have a decent 5 degree variance that they can take on the heated bed.
The machine he built was a massive Core XY type build using 2040 extrusion.
Good luck and share some pictures as you go.September 19, 2019 at 11:05 am #114738
It sounds like you’re the perfect person to try it. If you do, please post back here. If it works or you get stuck. I’m sure several people will be interested in it.September 19, 2019 at 11:18 am #114743
As I get my Lowrider up and running, and the next couple of projects off the board, I will put some serious thought into this and update as I go. If I run into any problems I will be sure to ask for help.
Thank you all for the input.September 19, 2019 at 1:00 pm #114776
Yeah, you really don’t want to use multiple thermistors unless you have some way to calibrate them. It’s not unusual at all for the normal 3D printer thermistor to be off by 10°C in either direction. With a single thermistor and a single bed heater it doesn’t matter, you just adjust the heat to match what works best, not to match a specific ‘temperature’. If you did calibrate though, you could run separate PIDs on each heater and run the whole bed at a single value. Remember that a standard 8″x8″ bed uses around 11A of 12VDC, so your 24″x36″ will take 13.5 times that, or a 150A supply. In actuality if you are using standard 200xmm x 200mm heat beds you’ll use 15 of them, or else you’ll come up short of the 36″ side, and that’s 165A, or three 60A supplies. Note that that means you’re not running off a standard 110V 15A circuit, you’ll want to have the newer 20A standard or go with 220V.
You must be logged in to reply to this topic.