Separate names with a comma.
Some features disabled for guests. Register Today.
Discussion in '3D printers' started by antonyking, Feb 3, 2017.
Large scale True 500x500x500 build volume 3D printer.
antonyking published a new build:
Read more about this build...
Looking forward to watching your creativity unfold on this machine.
This one should be good
should be cool!!!
Thanks Guys the build is a little slow ATM while parts dribble in. I hope to get into it full on over the next few weeks.
I have decided on a few of the electronics and I have 3x Nema 23 0.9deg(400 steps/rev) 127.5oz.in arriving tomorrow. I am going to start with a MKS Gen v1.4 board and drive the Nema 23's with TB6560 Drivers The Nema 17's on the X axis and extruder will have DRV8825 drivers.
Everything I have done to date, as in yesterday has been added to the build page. Hmm maybe I should do it here?
Anyway the build is mocked up with the 3D printed test plates while await for the aluminum plates to be machined.
I am printing off some temp 3d parts to make a simple "one motor to belt drive 2 leadscrews" idea as a test and because I only have one TB6560 driver for the Z at this time. Plus it would be interesting to see it go.
I will start wiring the motors. One good thing about the MKS Gen 1.4 is that it has pins for external drivers already and I don't have to make up a way to pin into the normal Driver socket.
The wiring has been a snack to sort out but I didn't think about the length of all the wires. So I will have to go to JayCar and buy some meters of wire.
The MKS board makes like so easy with it's layout.
I setup the board with Marlin 1.1.0-RC8 which is what I am running in the below short movie of the homing. I also did a firmware setup of the Stable 1.0.2-2 as a backup.
The advantages of the new firmware makes it all work very nicely, so it would be great if there was an official release of 1.1 soon. One of the new features is the BLtouch is now a one comment line in the new firmware with new Mesh Autoleveling. You can see I have a Geeetech 3DTouch which a BLTouch ripoff and I have one setup on my LulzBot Custom and they do work very well.
The first run.
After many small issues like duff touch sensor x2 that seems to have an weird fault causing it to go into alarm sometimes and sometimes working. I have a new one coming.
I replaced the Y axis linear drive and replaced with a belt setup. Instantly I got better Y axis. In my discovery though I fount the TB6560 drivers may be part of the issue and that was after I had to return 2 which had the switches fall apart. I replaced the Y axis with a Nema 23 0.9deg motor same as the 2x Z axis and I am driving it from the MKS Geb 1.4 with DRV8825 drivers.
I had a eureka moment when trying to get the TB6560 drivers to run the Z axis smooth when I realised the motors are low power, they are 12v but only 0.38A and given the 8825 are 2.2A I figure I'd give it a go... bugger me not only do they driver the Nema 23's smoother they will do it at 1/32 step.
Where I was missing steps with the TB6560 I can push the speed till I start jumping teeth on the belt. I have a 10mm wide GT2 belt and pulleys coming but I may have to got to GT3 belt for the Y axis just to help cope with the mass.
Very nice build !
Looking good. How's the ringing on the part now that you have gone to belts?
I really like your Vslot "nail" Do you have an stl file of it you are willing to share?
OpenBuilds V Slot Joiner by ant0ny
On the calibration cube it's not as bad as I expected. I will see if I can get a pic.
I had a situation where it would just miss a few steps at the very top of the calibration print, so I thought it was temp related with either the drivers or motors but I did a random longer print of a hinge part and it didn't miss a beat.
I am printing with a 0.4 nozzle which would be way to small if you where to utilize the size of the printer... so I need to get the volcano setup with something like a 0.8 nozzle to see what it's like for what it would be used for.
Thanks for the stl file. Much appreciated.
i know i posted this elsewhere, but just for good measure.
try using pir foam, 'sink' the extruder 0.5 mm into it and start printing.
no more precision leveling required (foam is very forgiving) and also no more heating and all electronics related.
Ok thanks I might try this.
Would it be feasible to use all 3D printed parts for the build instead of just for the mockup? Would lower the cost for me considerably.
Well yes it could. I intend to build a smaller version of this same printer with the already printed components... not as big but 300x300 or 400x400 maybe.
The only 3 parts I would not print, just for reliability would be the 3 gantry plates the wheels bolt to.. so the 2x C-Beam Gantry plates and the 1x X axis Gantry plate.
I'll do a write up the smaller one when i build after this one is delivered.
"but what has been my biggest issue is the non printing moves"
You are moving a LOT of mass on the Y axis. Especially if a layer ends with the nozzle at the front of the object and starts the next at the back of the object (or visa-versa). The mass builds up momentum on a move like that that can jerk the belt, and motor, while trying to stop. I print all of the parts for my printer with 200mm/s infill and 150mm/sec perimeters and was having Y shift problems until I dropped the non-print moves back to 70mm/s. With your mass, you may need to go a lot slower on those long moves to compensate for the momentum. Also, try setting the Y motor to remain engaged when not moving (#define DISABLE_Y false) Doing that would mean you could not move the Y axis when not printing until you hit the Emergency Stop to release it, but it keeps the motor from freewheeeling when trying to stop that mass from moving.
Keith, Good tips I'll look into this.
I received my Ally plates which will stiffen the frame a little as well and re calibrate.
BTW, you can also add a M999 gcode to the End G-Code in Slic3r to automatically cause that reset to release the Y axis if you keep it engaged at rest.
I believe you have a 3DTouch from Geeetech. Did you run a repeatability test on your printer (M48 gcode command) ? If yes, what are your results ?
I ran the command so it would probe ten times (M48 p10 v4), and here are the results:
Mean: 0.093375 Min: -0.162 Max: 0.175 Range: 0.337
Standard Deviation: 0.108110
As you can see, the deviation is not very good, and the range is even worse !...
This is the result of one M48 test, but all the M48 I tried are in that range.
Any solution to solve that issue ?
I tried to increase the number of microsteps from 16 to 32. Did not help.
I tried deactivating the lcd display (takes a lot of processor). Did not help.
My BLTouch mount seems plenty rigid.
I did not disassembled the probe yet to clean it. But i don't like to do it...
No I haven't done this. I did see Toms Guide where he compared.
For me it's the ease and it works OK but the fact that it's even in the ball park is better than screwing a bolt and micro switch etc like the old days... lol what... like 18 month a ago.
I have both BLTouch and 3DTouch so I should to a comparison. I use 3DTouch for myself but I only sell on BLTouch with some printers I build. 3DTouch are not as good by a long shot... I have 5 3DTouch's and 2 work consistently and 3 are ... well spares.
Thank you for your reply !
I wonder if my accuracy problem does not come from a side-to-side play of the pin when extended. My 3DTouch has quite a lot (+/-0.5mm at the tip, maybe more).
For example, if the angle of the pin varies when probing, it could explain why the accuracy is not super good. When you try to get a 0.01-0.02mm accuracy (accuracy you can expect with an optical sensor), a slight difference in the pin angle can make a big difference, I guess.
If you have currently a 3DTouch and a real BLTouch in your workshop, can you try to compare them ? See if there is side-to-side play with the real BLTouch ?
I tried to fix the side-to-side play, but was not able to get a smooth translation of the pin, so I went back to the original pin. Maybe when (if) my 3D printer will work, I will print some sort of sleeve for the sensor that will reduce the play. I don't know.
Or I buy a real BLTouch...